Actual wagers across synced sportsbooks. Separate from the frozen model picks.
Source update 2026-10-28 13:14 UTC · Chicago dates. Year uses settlement date, or placement date when settlement is unknown.
2026 calendar-year summary
237 of 999 all-history tickets · All sportsbooks. Search does not change summary totals.
Source net profit
-$208.37
Settled stakes $3,429.22
Our ticket outcomes · W–L
85–90
Pikkit aggregate · W–L–other
Not verified
All-stakes ROI
-5.50%
Settled-stake yield
-6.08%
25 pending · $360.00 staked · 0 confirmed outcomes use placement dates because settlement dates are unknown. Counts are tickets, not individual parlay legs.
Tickets · 237
2026 · All sportsbooks · All tickets in the selected scope.
Pending tickets · 25
Kickoffs in Chicago time, verified from Pikkit ticket details. Unverified times are listed last within each event date.
Event date unknown · 25 pending · $360.00 staked
Baylor +1.5 Spread Baylor @ UCF
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-27
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.901
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Duke -7.5 Spread Boston College @ Duke
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-27
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
2.050
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Fresno State +5.5 Spread Oregon State @ Fresno State
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.935
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Indiana -19.5 Spread Minnesota @ Indiana
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.877
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Iowa State +8.5 Spread Oklahoma State @ Iowa State
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.909
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
James Madison -20.5 Spread Troy @ James Madison
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.870
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Kansas State -2.5 Spread Kansas State @ Colorado
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.877
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Kennesaw State +2.5 Spread Kennesaw State @ Middle Tennessee
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.952
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Louisville -22.5 Spread Stanford @ Louisville
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.901
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Mississippi State +20.5 Spread Mississippi State @ Texas
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.926
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
North Dakota State -24.5 Spread UTEP @ North Dakota State
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.893
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
North Texas -13.5 Spread Florida Atlantic @ North Texas
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.885
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
North Texas -7.5 Spread Florida Atlantic @ North Texas | James Madison -14.5 Spread Troy @ James Madison
Bet365 · round robin 2 · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
Not supplied
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
North Texas -7.5 Spread Florida Atlantic @ North Texas
James Madison -14.5 Spread Troy @ James Madison
Ohio State -6.5 Spread Ohio State @ USC
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-27
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.885
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Old Dominion +2.5 Spread Marshall @ Old Dominion
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.990
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Over 46.5 Total Kent State @ Sacramento State
Bet365 · straight · Stake $10.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.909
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Over 71.5 Total Florida Atlantic @ North Texas | Under 54.5 Total Troy @ James Madison
Bet365 · round robin 2 · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
Not supplied
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
Over 71.5 Total Florida Atlantic @ North Texas
Under 54.5 Total Troy @ James Madison
Penn State -7.5 Spread Purdue @ Penn State
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.893
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Pittsburgh -7.5 Spread Georgia Tech @ Pittsburgh
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.917
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Sacramento State +9.5 Spread Kent State @ Sacramento State
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.952
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
South Florida -7.5 Spread UAB @ South Florida
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-27
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.870
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Tulane -10.5 Spread Tulane @ Charlotte
Bet365 · straight · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.952
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Tulane -3.5 Spread Tulane @ Charlotte | UCF +5.5 Spread Baylor @ UCF | Kent State -2.5 Spread Kent State @ Sacramento State
Bet365 · round robin 3 · Stake $15.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
Not supplied
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
Tulane -3.5 Spread Tulane @ Charlotte
UCF +5.5 Spread Baylor @ UCF
Kent State -2.5 Spread Kent State @ Sacramento State
Under 46.5 Total Tulane @ Charlotte
Bet365 · straight · Stake $10.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.909
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Under 51.5 Total Kennesaw State @ Middle Tennessee
Bet365 · straight · Stake $10.00
Kickoff time not verified
Pending
Ticket details
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
Pending; no settlement supplied
Source decimal price
1.885
Stake + source net profit
Unavailable until settled
Arithmetic return; not a verified payment.
Settled tickets · 25 shown of 212
Sam Houston State +5.5 Spread Sam Houston State @ Missouri State
Bet365 · Stake $15.00
Settled 2026-10-27
Lost
Net -$15.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
1.909
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Over 50.5 Total Sam Houston State @ Missouri State
Bet365 · Stake $10.00
Settled 2026-10-27
Lost
Net -$10.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
1.885
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Under 62.5 Total Delaware @ Western Kentucky | Over 45.5 Total Sam Houston State @ Missouri State
Bet365 · Stake $10.00
Settled 2026-10-27
Lost
Net -$10.00
Ticket details
round robin 2 · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
Not supplied
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
Under 62.5 Total Delaware @ Western Kentucky
Over 45.5 Total Sam Houston State @ Missouri State
Western Kentucky +4.0 Spread Delaware @ Western Kentucky | Sam Houston State +11.0 Spread Sam Houston State @ Missouri State
Bet365 · Stake $15.00
Settled 2026-10-27
Lost
Net -$15.00
Ticket details
round robin 2 · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
Not supplied
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
Western Kentucky +4.0 Spread Delaware @ Western Kentucky
Sam Houston State +11.0 Spread Sam Houston State @ Missouri State
Under 57.5 Total Delaware @ Western Kentucky
Bet365 · Stake $10.00
Settled 2026-10-27
Won
Net $9.26
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
1.926
Stake + source net profit
$19.26
Arithmetic return; not a verified payment.
Western Kentucky -1.5 Spread Delaware @ Western Kentucky
Bet365 · Stake $15.00
Settled 2026-10-27
Lost
Net -$15.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-26
Source settlement · Chicago date
2026-10-27
Source decimal price
1.971
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Kentucky -10.5 Spread Vanderbilt @ Kentucky
Bet365 · Stake $55.00
Settled 2026-10-24
Won
Net $49.55
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-24
Source decimal price
1.901
Stake + source net profit
$104.55
Arithmetic return; not a verified payment.
UCLA -14.5 Spread Michigan State @ UCLA
Bet365 · Stake $25.00
Settled 2026-10-24
Won
Net $24.28
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-24
Source decimal price
1.971
Stake + source net profit
$49.28
Arithmetic return; not a verified payment.
Ohio -10.5 Spread Eastern Michigan @ Ohio
Bet365 · Stake $25.00
Settled 2026-10-24
Lost
Net -$25.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-24
Source decimal price
1.901
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Navy +2.5 Spread North Texas @ Navy
Bet365 · Stake $25.00
Settled 2026-10-24
Lost
Net -$25.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-19
Source settlement · Chicago date
2026-10-24
Source decimal price
1.769
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Kent State -2.5 Spread Akron @ Kent State
Bet365 · Stake $25.00
Settled 2026-10-24
Won
Net $20.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-19
Source settlement · Chicago date
2026-10-24
Source decimal price
1.800
Stake + source net profit
$45.00
Arithmetic return; not a verified payment.
TCU -8.5 Spread West Virginia @ TCU
Bet365 · Stake $55.00
Settled 2026-10-24
Lost
Net -$55.00
Ticket details
straight · Net profit from the reviewed source record.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-19
Source settlement · Chicago date
2026-10-24
Source decimal price
1.633
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
James Madison -4.5 Spread James Madison @ Appalachian State | Liberty -2.5 Spread Liberty @ Kennesaw State | Memphis -3.5 Spread East Carolina @ Memphis
Bet365 · Stake $25.00
Settled 2026-10-22
Lost
Net -$25.00
Ticket details
round robin 3 · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-22
Source settlement · Chicago date
2026-10-22
Source decimal price
Not supplied
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
One ticket; stakes and profit are not allocated to individual legs.
James Madison -4.5 Spread James Madison @ Appalachian State
Liberty -2.5 Spread Liberty @ Kennesaw State
Memphis -3.5 Spread East Carolina @ Memphis
James Madison -10.5 Spread James Madison @ Appalachian State
Bet365 · Stake $25.00
Settled 2026-10-22
Lost
Net -$25.00
Ticket details
straight · Net profit from the reviewed source record.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-20
Source decimal price
1.926
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Delaware -2.5 Spread Missouri State @ Delaware
Bet365 · Stake $25.00
Settled 2026-10-20
Won
Net $20.33
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-20
Source decimal price
1.813
Stake + source net profit
$45.33
Arithmetic return; not a verified payment.
UCLA -14.5 Spread Michigan State @ UCLA
Bet365 · Stake $55.00
Settled 2026-10-19
Cashed out
Net $0.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-19
Source decimal price
1.971
Stake + source net profit
$55.00
Arithmetic return; not a verified payment.
Alabama -14.5 Spread Texas A&M @ Alabama
Bet365 · Stake $55.00
Settled 2026-10-19
Cashed out
Net $0.00
Ticket details
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-18
Source settlement · Chicago date
2026-10-19
Source decimal price
1.909
Stake + source net profit
$55.00
Arithmetic return; not a verified payment.
UCLA -2.5 Spread Wisconsin @ UCLA
Bet365 · Stake $15.00
Settled 2026-10-17
Won
Net $12.94
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-15
Source settlement · Chicago date
2026-10-17
Source decimal price
1.862
Stake + source net profit
$27.94
Arithmetic return; not a verified payment.
Syracuse +15.5 Spread Louisville @ Syracuse
Bet365 · Stake $15.00
Settled 2026-10-17
Won
Net $13.77
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-15
Source settlement · Chicago date
2026-10-17
Source decimal price
1.917
Stake + source net profit
$28.77
Arithmetic return; not a verified payment.
Bowling Green -6.5 Spread Ball State @ Bowling Green
Bet365 · Stake $15.00
Settled 2026-10-17
Lost
Net -$15.00
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-15
Source settlement · Chicago date
2026-10-17
Source decimal price
1.870
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Louisiana Monroe +15.5 Spread Louisiana Tech @ UL Monroe
Bet365 · Stake $15.00
Settled 2026-10-17
Lost
Net -$15.00
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-13
Source settlement · Chicago date
2026-10-17
Source decimal price
1.917
Stake + source net profit
$0.00
Arithmetic return; not a verified payment.
Florida +10.5 Spread Florida @ Texas
Bet365 · Stake $15.00
Settled 2026-10-17
Won
Net $13.77
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-13
Source settlement · Chicago date
2026-10-17
Source decimal price
1.917
Stake + source net profit
$28.77
Arithmetic return; not a verified payment.
Penn State +6.5 Spread Penn State @ Michigan
Bet365 · Stake $15.00
Settled 2026-10-17
Won
Net $13.89
Ticket details
Event date 2026-10-17; separate from placement and settlement.
straight · Net profit from the reviewed source record.
Placed · Chicago date
2026-10-13
Source settlement · Chicago date
2026-10-17
Source decimal price
1.926
Stake + source net profit
$28.89
Arithmetic return; not a verified payment.
Accounting and source evidence
Pikkit aggregate record has not been verified for this filtered population or source update.
All-stakes ROI: source net profit ÷ all original stakes, including open and void tickets. Settled-stake yield: completed-ticket net profit ÷ completed-ticket stakes.
212 completed tickets · 36 cash-outs · 1 pushes · 0 voids. Cash-outs, pushes and voids count in dollars, separately from wins–losses.
25 pending · $360.00 staked · Open stakes are excluded from settled-stake yield. Open tickets and confirmed outcomes with unknown settlement dates use placement year.
0 confirmed outcomes have unknown settlement dates; their profit is included once, using placement year. These tickets sort by placement date.
Confirmed in Pikkit ticket details and already included once in tracker profit, yield and wins–losses above. Settlement dates remain unknown. The two previously unmatched Indiana entries are now identified as separate Bet365 tickets. Observations follow the year and sportsbook filters.
Review observed outcomes · 34
James Madison -8.5 Spread James Madison @ Georgia Southern
Bet365 · Historical stake $15.00 · Observed loss · -$15.00 reported net
Exact college-football tickets (NCAAFB), across synced sportsbooks. Mixed college/NFL tickets and other sports are excluded.
Known settlement dates use America/Chicago. Confirmed Pikkit outcomes with unknown settlement dates retain null dates and years; year filters explicitly use placement year for those tickets. Game-final and capture dates are not settlement dates.
Settled-stake yield uses net profit on completed tickets divided by their original stakes, including cash-outs, pushes and voids. Open stakes and source-recorded open adjustments are excluded from yield. Source net profit includes those open adjustments, shown separately.
One record is one ticket, including each parlay and round-robin. Selections are descriptive legs, not separately staked records.
Prior history retained; October 28 user-supplied transaction export reconciled 160 exact NCAAFB tickets, including 25 open tickets. NFL and mixed tickets remain outside this college-football ledger. Source settlement amounts and Chicago dates retained; this exported range is not a full live account sync. Book-direct payment and separate fee treatment remain unverified.
USD assumed from supplied account screenshots; currency was not specified in the transaction exports.
No single unit size was supplied. Display source stakes in dollars; do not invent unit counts.
These are personal wagers. They are separate from frozen model-pick performance and do not establish that a wager followed the model.
Statuses and net profit come from reviewed source exports and Pikkit ticket details. Scores and rounded odds do not determine settlement or payment.
Every settled pick across all weeks, graded at the line the model saw when it made the pick. Full-game spreads.
Every pick the model has made since Week 1, nothing hidden. Win rate leaves out pushes; you need 52.4% to break even at −110. Units assume 1 unit on every pick. Quote gaps compare the saved pick spread with a recorded CFBD market quote, often a postgame median across books. Positive means a larger picked-side spread. Verified CLV is unavailable: book-specific closing time, spread price and a close definition are not recorded per row. These proxies do not establish a tradeable closing quote, bettor skill or model edge.
Record
111–110–2
Win rate
50.2%
Units
-9.09
Average quote gap (proxy)
-0.04 pts
By model
Group
Picks
Record
Win rate
Units
Quote gap (proxy)
In-season EPA model
114
57–56–1
50.4%
-4.18
+0.00 pts (42 of 114 positive)
Preseason SP+ ratings
109
54–54–1
50.0%
-4.91
-0.08 pts (34 of 109 positive)
Weeks 1–5 picks come from preseason SP+ ratings plus home field; the in-season EPA model takes over once it has enough games (Week 6). The preseason lane did not hold up in backtests, so treat its record as a live test.
By lane
Group
Picks
Record
Win rate
Units
Quote gap (proxy)
⭐ C1 plays
6
6–0
100.0%
+5.45
+0.08 pts (4 of 6 positive)
Standard picks (3+ point gap)
217
105–110–2
48.8%
-14.55
-0.04 pts (72 of 217 positive)
⭐ C1 plays are in-season picks in the rule card’s best spot: a gap of 14+ points in weeks 6–9, or 12–14 points in weeks 10–16. C1 is a tag, not a filter; every other pick is still listed.
By week
Group
Picks
Record
Win rate
Units
Quote gap (proxy)
Week 1
10
4–6
40.0%
-2.36
+0.53 pts (6 of 10 positive)
Week 2
27
11–16
40.7%
-6.00
-0.32 pts (8 of 27 positive)
Week 3
22
12–10
54.5%
+0.91
-0.02 pts (8 of 22 positive)
Week 4
27
13–13–1
50.0%
-1.18
+0.09 pts (9 of 27 positive)
Week 5
23
14–9
60.9%
+3.73
-0.30 pts (3 of 23 positive)
Week 6
34
21–13
61.8%
+6.09
-0.26 pts (10 of 34 positive)
Week 7
42
17–24–1
41.5%
-8.55
-0.06 pts (13 of 42 positive)
Week 8
38
19–19
50.0%
-1.73
+0.30 pts (19 of 38 positive)
Showing 223 of 223 picks
Details
Louisiana -10
vs Southern Miss · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401869967; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–35 final
loss
Missed by 14
Recorded market spread (unverified close)
-11.5
Source: CFBD median across books; not a verified close
Model edge
-7.52 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+1.50 points
San José State +6.5
vs Nevada · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401864528; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–28 final
loss
Missed by 8.5
Recorded market spread (unverified close)
+5.75
Source: CFBD median across books; not a verified close
Model edge
-3.52 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.75 points
Boise State -14.5
vs Washington State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401860907; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–10 final
win
Covered by 6.5
Recorded market spread (unverified close)
-14.75
Source: CFBD median across books; not a verified close
Model edge
-4.58 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.25 points
Utah State +13.75
vs Texas State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401860906; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–20 final
win
Covered by 10.75
Recorded market spread (unverified close)
+14.25
Source: CFBD median across books; not a verified close
Model edge
-3.88 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.50 points
Colorado State +1.25
vs San Diego State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401860905; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
56–49 final
win
Covered by 8.25
Recorded market spread (unverified close)
+0.25
Source: CFBD median across books; not a verified close
Model edge
+6.77 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+1.00 points
Wisconsin +2.75
vs USC · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858500; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–7 final
win
Covered by 19.75
Recorded market spread (unverified close)
+2
Source: CFBD median across books; not a verified close
Model edge
+9.62 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.75 points
North Carolina -8.75
vs Syracuse · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858275; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
30–31 final
loss
Missed by 9.75
Recorded market spread (unverified close)
-8.25
Source: CFBD median across books; not a verified close
Model edge
+5.41 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.50 points
Arizona -11
vs Iowa State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856838; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
39–3 final
win
Covered by 25
Recorded market spread (unverified close)
-12
Source: CFBD median across books; not a verified close
Model edge
+5.31 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+1.00 points
Colorado +13.5
vs Oklahoma State · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856836; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–57 final
loss
Missed by 19.5
Recorded market spread (unverified close)
+13
Source: CFBD median across books; not a verified close
Model edge
-7.02 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.50 points
Ole Miss +3
vs Texas · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856732; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–34 final
loss
Missed by 4
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
-3.12 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Kentucky -10.5
vs Vanderbilt · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856729; settled 2026-10-25T13:47:54.671690+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
34–17 final
win
Covered by 6.5
Recorded market spread (unverified close)
-10
Source: CFBD median across books; not a verified close
Model edge
+11.47 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.50 points
Troy -11.5
vs UL Monroe · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401871094; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–21 final
loss
Missed by 18.5
Recorded market spread (unverified close)
-13.5
Source: CFBD median across books; not a verified close
Model edge
+4.35 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+2.00 points
Central Michigan +3.5
vs Miami (OH) · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401866476; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–28 final
win
Covered by 2.5
Recorded market spread (unverified close)
+3.5
Source: CFBD median across books; not a verified close
Model edge
+4.90 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Toledo -2.75
vs Western Michigan · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401866450; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
45–38 final
win
Covered by 4.25
Recorded market spread (unverified close)
-2.5
Source: CFBD median across books; not a verified close
Model edge
+5.06 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.25 points
Ohio -10
vs Eastern Michigan · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401866449; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–20 final
loss
Missed by 6
Recorded market spread (unverified close)
-11.5
Source: CFBD median across books; not a verified close
Model edge
+6.24 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+1.50 points
Buffalo -3
vs Bowling Green · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401866448; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–24 final
win
Covered by 4
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
+7.10 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Northern Illinois +3.75
vs Hawai'i · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401864526; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–56 final
loss
Missed by 31.25
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
+9.76 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.75 points
Navy +3.25
vs North Texas · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401862810; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
39–49 final
loss
Missed by 6.75
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
+5.11 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.25 points
UCLA -15.25
vs Michigan State · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858499; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
28–10 final
win
Covered by 2.75
Recorded market spread (unverified close)
-16.5
Source: CFBD median across books; not a verified close
Model edge
+12.29 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+1.25 points
Northwestern -15.5
vs Rutgers · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858498; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
42–14 final
win
Covered by 12.5
Recorded market spread (unverified close)
-15.5
Source: CFBD median across books; not a verified close
Model edge
+7.86 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Boston College +6
vs Georgia Tech · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858273; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
49–27 final
win
Covered by 28
Recorded market spread (unverified close)
+6.5
Source: CFBD median across books; not a verified close
Model edge
-4.12 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.50 points
TCU -10.5
vs West Virginia · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856841; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–24 final
loss
Missed by 14.5
Recorded market spread (unverified close)
-10
Source: CFBD median across books; not a verified close
Model edge
+12.07 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.50 points
Cincinnati +11.5
vs Texas Tech · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856840; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
37–24 final
win
Covered by 24.5
Recorded market spread (unverified close)
+10.75
Source: CFBD median across books; not a verified close
Model edge
+6.41 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.75 points
Kansas State -1.25
vs Arizona State · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856839; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–28 final
loss
Missed by 12.25
Recorded market spread (unverified close)
-1.5
Source: CFBD median across books; not a verified close
Model edge
-7.87 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.25 points
UCF +7.5
vs BYU · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856835; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
39–31 final
win
Covered by 15.5
Recorded market spread (unverified close)
+7.5
Source: CFBD median across books; not a verified close
Model edge
+4.81 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Kansas +1.25
vs Baylor · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856834; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
25–24 final
win
Covered by 2.25
Recorded market spread (unverified close)
-1.5
Source: CFBD median across books; not a verified close
Model edge
+4.61 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+2.75 points
Tennessee +2.5
vs South Carolina · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856731; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–27 final
loss
Missed by 0.5
Recorded market spread (unverified close)
+2.5
Source: CFBD median across books; not a verified close
Model edge
-4.55 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Mississippi State +2.25
vs Oklahoma · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856730; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
15–23 final
loss
Missed by 5.75
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
+7.55 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.75 points
LSU -7
vs Auburn · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856728; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–17 final
win
Covered by 14
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
-4.52 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
+0.00 points
Alabama -13.5
vs Texas A&M · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401856727; settled 2026-10-25T02:35:33.061285+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–20 final
loss
Missed by 2.5
Recorded market spread (unverified close)
-13.25
Source: CFBD median across books; not a verified close
Model edge
+8.83 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-25
Quote gap (proxy)
-0.25 points
Air Force -5.25
vs Wyoming · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401864525; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–28 final
loss
Missed by 19.25
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
-4.06 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
+1.75 points
Tulsa -1.75
vs Army · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401862808; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–38 final
loss
Missed by 4.75
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
+4.31 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
-3.25 points
Memphis -10.5
vs East Carolina · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401862807; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–34 final
loss
Missed by 3.5
Recorded market spread (unverified close)
-10.5
Source: CFBD median across books; not a verified close
Model edge
+5.01 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
+0.00 points
Massachusetts +14.25
vs UConn · Week 8
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401861965; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
16–35 final
loss
Missed by 4.75
Recorded market spread (unverified close)
+12.75
Source: CFBD median across books; not a verified close
Model edge
-12.10 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
+1.50 points
Stanford +6.5
vs NC State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858272; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–28 final
win
Covered by 9.5
Recorded market spread (unverified close)
+7.5
Source: CFBD median across books; not a verified close
Model edge
+3.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
-1.00 points
Duke +1.75
vs Virginia · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401858271; settled 2026-10-24T11:12:20.352745+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
19–16 final
win
Covered by 4.75
Recorded market spread (unverified close)
+2.5
Source: CFBD median across books; not a verified close
Model edge
-4.02 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-24
Quote gap (proxy)
-0.75 points
Delaware -3
vs Missouri State · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401871069; settled 2026-10-21T11:11:26.676321+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
61–38 final
win
Covered by 20
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
+5.99 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-21
Quote gap (proxy)
+0.00 points
Florida International -10
vs Middle Tennessee · Week 8
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 8, game 401871055; settled 2026-10-21T11:11:26.676321+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
23–26 final
loss
Missed by 13
Recorded market spread (unverified close)
-10.25
Source: CFBD median across books; not a verified close
Model edge
+4.15 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 8
Settled
2026-10-21
Quote gap (proxy)
+0.25 points
UL Monroe +15.5
vs Louisiana Tech · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401871091; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
0–24 final
loss
Missed by 8.5
Recorded market spread (unverified close)
+14.5
Source: CFBD median across books; not a verified close
Model edge
+3.57 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+1.00 points
Louisiana -3
vs Troy · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401869966; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
33–6 final
win
Covered by 24
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
+9.73 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Georgia State +9.5
vs James Madison · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401869957; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
22–49 final
loss
Missed by 17.5
Recorded market spread (unverified close)
+7.75
Source: CFBD median across books; not a verified close
Model edge
-5.17 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+1.75 points
Coastal Carolina +1.5
vs App State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401869844; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
32–35 final
loss
Missed by 1.5
Recorded market spread (unverified close)
+1.25
Source: CFBD median across books; not a verified close
Model edge
+8.48 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Miami (OH) -9.75
vs Akron · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401866478; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
34–14 final
win
Covered by 10.25
Recorded market spread (unverified close)
-9.5
Source: CFBD median across books; not a verified close
Model edge
+5.71 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Sacramento State +15.5
vs Ohio · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401866445; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
7–48 final
loss
Missed by 25.5
Recorded market spread (unverified close)
+16.5
Source: CFBD median across books; not a verified close
Model edge
+5.97 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-1.00 points
Central Michigan +7
vs Western Michigan · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401866443; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–21 final
push
Push · exactly on the line
Recorded market spread (unverified close)
+7.5
Source: CFBD median across books; not a verified close
Model edge
+5.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.50 points
Massachusetts +7.25
vs Buffalo · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401866442; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
18–25 final
win
Covered by 0.25
Recorded market spread (unverified close)
+6.75
Source: CFBD median across books; not a verified close
Model edge
-9.48 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.50 points
Bowling Green -6.5
vs Ball State · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401866441; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–32 final
loss
Missed by 7.5
Recorded market spread (unverified close)
-6.25
Source: CFBD median across books; not a verified close
Model edge
+5.35 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Air Force -0.25
vs UNLV · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401864524; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–31 final
win
Covered by 9.75
Recorded market spread (unverified close)
-0.25
Source: CFBD median across books; not a verified close
Model edge
+9.43 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
San José State -7
vs UTEP · Week 7
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401864523; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–11 final
win
Covered by 6
Recorded market spread (unverified close)
-6.75
Source: CFBD median across books; not a verified close
Model edge
-14.69 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
New Mexico -14.5
vs Hawai'i · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401864521; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–10 final
loss
Missed by 3.5
Recorded market spread (unverified close)
-14.5
Source: CFBD median across books; not a verified close
Model edge
-11.11 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
South Florida -20.5
vs Kent State · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401862806; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
29–2 final
win
Covered by 6.5
Recorded market spread (unverified close)
-20.5
Source: CFBD median across books; not a verified close
Model edge
+4.77 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Memphis -7
vs Tulane · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401862801; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–31 final
loss
Missed by 21
Recorded market spread (unverified close)
-6.5
Source: CFBD median across books; not a verified close
Model edge
-9.53 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.50 points
Colorado State +14
vs Texas State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401860923; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–66 final
loss
Missed by 25
Recorded market spread (unverified close)
+15.25
Source: CFBD median across books; not a verified close
Model edge
-7.64 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-1.25 points
Fresno State -7.25
vs San Diego State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401860904; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
23–26 final
loss
Missed by 10.25
Recorded market spread (unverified close)
-6.75
Source: CFBD median across books; not a verified close
Model edge
-11.08 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.50 points
Oregon State -18.5
vs Washington State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401860903; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
48–30 final
loss
Missed by 0.5
Recorded market spread (unverified close)
-18.5
Source: CFBD median across books; not a verified close
Model edge
+12.50 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
UCLA -2.5
vs Wisconsin · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858494; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–7 final
win
Covered by 7.5
Recorded market spread (unverified close)
-2
Source: CFBD median across books; not a verified close
Model edge
+7.16 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.50 points
Purdue -3.5
vs Washington · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858493; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
26–38 final
loss
Missed by 15.5
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
+5.46 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.50 points
Nebraska +7
vs Oregon · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858492; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–17 final
win
Covered by 4
Recorded market spread (unverified close)
+6.75
Source: CFBD median across books; not a verified close
Model edge
-12.19 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Northwestern -7
vs Michigan State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858491; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–21 final
loss
Missed by 4
Recorded market spread (unverified close)
-7.25
Source: CFBD median across books; not a verified close
Model edge
-9.33 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Penn State +6.5
vs Michigan · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858490; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
18–21 final
win
Covered by 3.5
Recorded market spread (unverified close)
+6
Source: CFBD median across books; not a verified close
Model edge
-3.28 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.50 points
Maryland -5.75
vs Rutgers · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858489; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
30–37 final
loss
Missed by 12.75
Recorded market spread (unverified close)
-5.75
Source: CFBD median across books; not a verified close
Model edge
+5.69 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Indiana +3
vs Ohio State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858488; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
15–19 final
loss
Missed by 1
Recorded market spread (unverified close)
+2.75
Source: CFBD median across books; not a verified close
Model edge
+8.69 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Wake Forest -10
vs California · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858269; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–7 final
win
Covered by 24
Recorded market spread (unverified close)
-10
Source: CFBD median across books; not a verified close
Model edge
-9.13 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Virginia +9.75
vs SMU · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858268; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–23 final
loss
Missed by 0.25
Recorded market spread (unverified close)
+10
Source: CFBD median across books; not a verified close
Model edge
-4.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Boston College +6.5
vs Pittsburgh · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858267; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–21 final
win
Covered by 12.5
Recorded market spread (unverified close)
+6.5
Source: CFBD median across books; not a verified close
Model edge
+3.01 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Syracuse +14.75
vs Louisville · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858265; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
26–34 final
win
Covered by 6.75
Recorded market spread (unverified close)
+14.5
Source: CFBD median across books; not a verified close
Model edge
+7.46 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Virginia Tech -8.5
vs Georgia Tech · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858264; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
34–24 final
win
Covered by 1.5
Recorded market spread (unverified close)
-8.5
Source: CFBD median across books; not a verified close
Model edge
+5.43 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Miami -20.5
vs Florida State · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401858263; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
43–35 final
loss
Missed by 12.5
Recorded market spread (unverified close)
-20.25
Source: CFBD median across books; not a verified close
Model edge
+3.31 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Houston -4.5
vs Oklahoma State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856831; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–32 final
loss
Missed by 1.5
Recorded market spread (unverified close)
-4.25
Source: CFBD median across books; not a verified close
Model edge
+11.98 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Kansas +3.5
vs Kansas State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856830; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–35 final
loss
Missed by 7.5
Recorded market spread (unverified close)
+3.25
Source: CFBD median across books; not a verified close
Model edge
-10.61 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Cincinnati +1.5
vs West Virginia · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856829; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
42–45 final
loss
Missed by 1.5
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
-11.30 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Texas Tech -15.75
vs Arizona State · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856828; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
53–25 final
win
Covered by 12.25
Recorded market spread (unverified close)
-16
Source: CFBD median across books; not a verified close
Model edge
+12.87 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.25 points
Florida +11.25
vs Texas · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856724; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–24 final
win
Covered by 14.25
Recorded market spread (unverified close)
+13
Source: CFBD median across books; not a verified close
Model edge
-4.16 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-1.75 points
Alabama -7.75
vs Tennessee · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856723; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–26 final
loss
Missed by 23.75
Recorded market spread (unverified close)
-7.75
Source: CFBD median across books; not a verified close
Model edge
-8.63 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.00 points
Kentucky +6.5
vs Oklahoma · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856722; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–13 final
win
Covered by 14.5
Recorded market spread (unverified close)
+6.75
Source: CFBD median across books; not a verified close
Model edge
-5.73 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Missouri +5.25
vs Ole Miss · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856721; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
3–31 final
loss
Missed by 22.75
Recorded market spread (unverified close)
+4.5
Source: CFBD median across books; not a verified close
Model edge
-8.70 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
+0.75 points
Mississippi State +17.5
vs LSU · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401856720; settled 2026-10-18T18:47:45.212082+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
0–31 final
loss
Missed by 13.5
Recorded market spread (unverified close)
+17.75
Source: CFBD median across books; not a verified close
Model edge
-7.48 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-18
Quote gap (proxy)
-0.25 points
Sam Houston +1.5
vs Western Kentucky · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401871068; settled 2026-10-16T01:48:29.984112+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
3–16 final
loss
Missed by 11.5
Recorded market spread (unverified close)
+1.25
Source: CFBD median across books; not a verified close
Model edge
+7.55 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-16
Quote gap (proxy)
+0.25 points
Middle Tennessee +1.25
vs Delaware · Week 7
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401871067; settled 2026-10-16T01:48:29.984112+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
33–30 final
win
Covered by 4.25
Recorded market spread (unverified close)
+1.25
Source: CFBD median across books; not a verified close
Model edge
+10.05 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-16
Quote gap (proxy)
+0.00 points
Kennesaw State -3
vs Missouri State · Week 7
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 7, game 401871054; settled 2026-10-16T01:48:29.984112+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–24 final
loss
Missed by 6
Recorded market spread (unverified close)
-2.5
Source: CFBD median across books; not a verified close
Model edge
-5.89 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 7
Settled
2026-10-16
Quote gap (proxy)
-0.50 points
Florida International -6.5
vs New Mexico State · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401871066; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
22–3 final
win
Covered by 12.5
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
+4.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.50 points
Liberty -13.5
vs Sam Houston · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401870766; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–3 final
win
Covered by 18.5
Recorded market spread (unverified close)
-13.5
Source: CFBD median across books; not a verified close
Model edge
+5.41 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Louisiana +3
vs Louisiana Tech · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401869965; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–32 final
win
Covered by 2
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
-12.68 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Coastal Carolina +4.25
vs Marshall · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401869943; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–21 final
win
Covered by 10.25
Recorded market spread (unverified close)
+4.5
Source: CFBD median across books; not a verified close
Model edge
-6.28 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.25 points
Massachusetts +4.75
vs Miami (OH) · Week 6
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401866440; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
19–20 final
win
Covered by 3.75
Recorded market spread (unverified close)
+6
Source: CFBD median across books; not a verified close
Model edge
+17.42 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-1.25 points
Western Michigan -14
vs Kent State · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401866439; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–7 final
win
Covered by 17
Recorded market spread (unverified close)
-13.75
Source: CFBD median across books; not a verified close
Model edge
+3.26 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.25 points
Eastern Michigan -5.5
vs Akron · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401866435; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
25–23 final
loss
Missed by 3.5
Recorded market spread (unverified close)
-5
Source: CFBD median across books; not a verified close
Model edge
-3.02 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.50 points
San José State -4.5
vs Wyoming · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401864519; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–16 final
loss
Missed by 7.5
Recorded market spread (unverified close)
-4.25
Source: CFBD median across books; not a verified close
Model edge
+8.48 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.25 points
North Dakota State -2.5
vs UNLV · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401864518; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
7–13 final
loss
Missed by 8.5
Recorded market spread (unverified close)
-2.75
Source: CFBD median across books; not a verified close
Model edge
-7.96 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.25 points
Nevada -9.75
vs UTEP · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401864517; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
30–13 final
win
Covered by 7.25
Recorded market spread (unverified close)
-10
Source: CFBD median across books; not a verified close
Model edge
-10.80 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.25 points
Air Force -7
vs Northern Illinois · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401864516; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
28–7 final
win
Covered by 14
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
-3.00 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Charlotte +28.5
vs North Texas · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401862796; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–50 final
loss
Missed by 11.5
Recorded market spread (unverified close)
+27.75
Source: CFBD median across books; not a verified close
Model edge
-6.29 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.75 points
Temple -3.5
vs UConn · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401861964; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–34 final
loss
Missed by 17.5
Recorded market spread (unverified close)
-3.75
Source: CFBD median across books; not a verified close
Model edge
+7.30 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.25 points
Utah State -5.5
vs Washington State · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401860922; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–16 final
loss
Missed by 4.5
Recorded market spread (unverified close)
-4
Source: CFBD median across books; not a verified close
Model edge
+10.77 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-1.50 points
Oregon State -16.25
vs San Diego State · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401860902; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
66–14 final
win
Covered by 35.75
Recorded market spread (unverified close)
-14.5
Source: CFBD median across books; not a verified close
Model edge
+6.85 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-1.75 points
Purdue +2.75
vs Minnesota · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858486; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–13 final
win
Covered by 30.75
Recorded market spread (unverified close)
+2.75
Source: CFBD median across books; not a verified close
Model edge
+12.32 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Penn State -1.75
vs USC · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858485; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
23–31 final
loss
Missed by 9.75
Recorded market spread (unverified close)
+0.25
Source: CFBD median across books; not a verified close
Model edge
+7.16 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-2.00 points
UCLA +10.5
vs Oregon · Week 6
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858484; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–15 final
win
Covered by 12.5
Recorded market spread (unverified close)
+10.25
Source: CFBD median across books; not a verified close
Model edge
-15.27 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.25 points
Nebraska +8
vs Indiana · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858481; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–20 final
win
Covered by 5
Recorded market spread (unverified close)
+9
Source: CFBD median across books; not a verified close
Model edge
+8.41 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-1.00 points
Michigan State +2.5
vs Illinois · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858480; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
37–38 final
win
Covered by 1.5
Recorded market spread (unverified close)
+2.5
Source: CFBD median across books; not a verified close
Model edge
+3.57 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Wake Forest -3.25
vs NC State · Week 6
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858260; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–14 final
win
Covered by 20.75
Recorded market spread (unverified close)
-3.75
Source: CFBD median across books; not a verified close
Model edge
-14.87 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.50 points
Syracuse +11.5
vs Virginia · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858258; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
25–28 final
win
Covered by 8.5
Recorded market spread (unverified close)
+14
Source: CFBD median across books; not a verified close
Model edge
-6.35 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-2.50 points
Pittsburgh -3.75
vs North Carolina · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858256; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–21 final
loss
Missed by 7.75
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
+8.58 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.25 points
Duke -6.5
vs Georgia Tech · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401858255; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
25–28 final
loss
Missed by 9.5
Recorded market spread (unverified close)
-4.5
Source: CFBD median across books; not a verified close
Model edge
-6.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-2.00 points
Kansas +15.75
vs Utah · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856827; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–41 final
loss
Missed by 1.25
Recorded market spread (unverified close)
+15.75
Source: CFBD median across books; not a verified close
Model edge
-10.49 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
BYU -10.5
vs Iowa State · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856826; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–10 final
win
Covered by 3.5
Recorded market spread (unverified close)
-11.25
Source: CFBD median across books; not a verified close
Model edge
+7.89 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.75 points
Houston +2.5
vs Kansas State · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856825; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–16 final
win
Covered by 21.5
Recorded market spread (unverified close)
+2.5
Source: CFBD median across books; not a verified close
Model edge
-4.43 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
UCF +10.25
vs Oklahoma State · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856824; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
28–42 final
loss
Missed by 3.75
Recorded market spread (unverified close)
+10.25
Source: CFBD median across books; not a verified close
Model edge
-9.17 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Arizona -3.5
vs West Virginia · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856823; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
52–31 final
win
Covered by 17.5
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
-7.77 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Hawai'i +21
vs Arizona State · Week 6
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856808; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
12–30 final
win
Covered by 3
Recorded market spread (unverified close)
+20.75
Source: CFBD median across books; not a verified close
Model edge
-18.65 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.25 points
Vanderbilt +9.25
vs Ole Miss · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856718; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–52 final
loss
Missed by 25.75
Recorded market spread (unverified close)
+9.25
Source: CFBD median across books; not a verified close
Model edge
+6.22 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+0.00 points
Oklahoma +7.5
vs Texas · Week 6
51% historical accuracy3-7 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856717; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–24 final
win
Covered by 10.5
Recorded market spread (unverified close)
+7.75
Source: CFBD median across books; not a verified close
Model edge
+5.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
-0.25 points
Missouri -3.5
vs Texas A&M · Week 6
⭐ C1 · 61% C1 in-sample cover rate (106 games) · rule chosen on these gamesIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856716; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–6 final
win
Covered by 17.5
Recorded market spread (unverified close)
-4.5
Source: CFBD median across books; not a verified close
Model edge
+16.47 points
Edge tier / historical accuracy
14+ points / 75%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 6
Settled
2026-10-12
Quote gap (proxy)
+1.00 points
Alabama -2.5
vs Georgia · Week 6
46% historical accuracy7-14 pt edge tierIn-season EPA model
Output:
ATS result at the saved pick line; cover margin (points)
Model:
EPA-OLS
Training:
Unavailable — not recorded in this artifact
Features:
EPA net difference; source snapshot unavailable
Evaluation:
2026 week 6, game 401856712; settled 2026-10-12T02:59:24.727937+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–20 final
loss
Missed by 5.5
Recorded market spread (unverified close)
-2.5
Source: CFBD median across books; not a verified close
Model edge
+13.38 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401871050; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
30–14 final
win
Covered by 9.25
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
-3.95 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401869962; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
45–17 final
win
Covered by 9.75
Recorded market spread (unverified close)
-16.25
Source: CFBD median across books; not a verified close
Model edge
+6.65 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401869956; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–42 final
loss
Missed by 29.5
Recorded market spread (unverified close)
+2
Source: CFBD median across books; not a verified close
Model edge
-3.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401869842; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
29–31 final
loss
Missed by 3.25
Recorded market spread (unverified close)
+0.25
Source: CFBD median across books; not a verified close
Model edge
-5.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401866477; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–24 final
loss
Missed by 16.25
Recorded market spread (unverified close)
-11.25
Source: CFBD median across books; not a verified close
Model edge
+12.35 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401866434; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–10 final
loss
Missed by 0.5
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
-5.90 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401866433; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–38 final
loss
Missed by 29.75
Recorded market spread (unverified close)
-5.75
Source: CFBD median across books; not a verified close
Model edge
+8.55 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401866432; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–20 final
win
Covered by 9.75
Recorded market spread (unverified close)
+13.25
Source: CFBD median across books; not a verified close
Model edge
+5.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401864515; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
28–0 final
win
Covered by 11.25
Recorded market spread (unverified close)
-17.25
Source: CFBD median across books; not a verified close
Model edge
+6.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401864514; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
61–7 final
win
Covered by 30.75
Recorded market spread (unverified close)
-23.25
Source: CFBD median across books; not a verified close
Model edge
+10.95 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401862792; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–17 final
loss
Missed by 9.75
Recorded market spread (unverified close)
-5.5
Source: CFBD median across books; not a verified close
Model edge
+6.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401862787; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
59–8 final
win
Covered by 30.5
Recorded market spread (unverified close)
-19.5
Source: CFBD median across books; not a verified close
Model edge
-6.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401862786; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
44–45 final
win
Covered by 0.25
Recorded market spread (unverified close)
+2
Source: CFBD median across books; not a verified close
Model edge
+6.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401860921; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
26–6 final
win
Covered by 21.75
Recorded market spread (unverified close)
+2.25
Source: CFBD median across books; not a verified close
Model edge
-8.25 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401860900; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–29 final
win
Covered by 11.5
Recorded market spread (unverified close)
+9.75
Source: CFBD median across books; not a verified close
Model edge
+6.00 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401858476; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–34 final
loss
Missed by 23.75
Recorded market spread (unverified close)
-2.75
Source: CFBD median across books; not a verified close
Model edge
-4.35 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401858472; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–24 final
loss
Missed by 17.25
Recorded market spread (unverified close)
-9.75
Source: CFBD median across books; not a verified close
Model edge
+7.15 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401858253; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
7–38 final
loss
Missed by 32.25
Recorded market spread (unverified close)
-1.25
Source: CFBD median across books; not a verified close
Model edge
-4.95 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401858252; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–42 final
win
Covered by 5.5
Recorded market spread (unverified close)
+6.5
Source: CFBD median across books; not a verified close
Model edge
+6.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401856821; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
29–7 final
win
Covered by 8.5
Recorded market spread (unverified close)
-13.5
Source: CFBD median across books; not a verified close
Model edge
-7.50 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401856819; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–27 final
win
Covered by 0.5
Recorded market spread (unverified close)
+10.5
Source: CFBD median across books; not a verified close
Model edge
-4.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401856817; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
55–19 final
win
Covered by 39.5
Recorded market spread (unverified close)
+3.5
Source: CFBD median across books; not a verified close
Model edge
-6.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 5, game 401856708; settled 2026-10-04T15:32:17.567874+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
45–17 final
win
Covered by 33.5
Recorded market spread (unverified close)
+5.5
Source: CFBD median across books; not a verified close
Model edge
+3.20 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401871048; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
23–13 final
win
Covered by 3
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
+11.20 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401869961; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
46–20 final
win
Covered by 20
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
-4.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401866429; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–35 final
loss
Missed by 24.5
Recorded market spread (unverified close)
+6
Source: CFBD median across books; not a verified close
Model edge
+6.00 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401864579; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
42–18 final
win
Covered by 12.75
Recorded market spread (unverified close)
-11.5
Source: CFBD median across books; not a verified close
Model edge
-6.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401864511; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–35 final
loss
Missed by 10.75
Recorded market spread (unverified close)
+8.75
Source: CFBD median across books; not a verified close
Model edge
-7.65 points
Edge tier / historical accuracy
7-14 points / 46%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401862785; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–6 final
loss
Missed by 9.75
Recorded market spread (unverified close)
-17.75
Source: CFBD median across books; not a verified close
Model edge
-3.65 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401862782; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
45–17 final
win
Covered by 16.5
Recorded market spread (unverified close)
-11.5
Source: CFBD median across books; not a verified close
Model edge
-6.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401862780; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
34–7 final
win
Covered by 17
Recorded market spread (unverified close)
-9.25
Source: CFBD median across books; not a verified close
Model edge
-4.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401862778; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–24 final
loss
Missed by 10.5
Recorded market spread (unverified close)
-6.5
Source: CFBD median across books; not a verified close
Model edge
-5.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401860896; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
45–59 final
push
Push · exactly on the line
Recorded market spread (unverified close)
+14.25
Source: CFBD median across books; not a verified close
Model edge
-6.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401860895; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
33–7 final
win
Covered by 15
Recorded market spread (unverified close)
-11.5
Source: CFBD median across books; not a verified close
Model edge
-3.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401860894; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–21 final
loss
Missed by 8
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
-3.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401860893; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
16–41 final
loss
Missed by 22.25
Recorded market spread (unverified close)
+2.75
Source: CFBD median across books; not a verified close
Model edge
-3.15 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401860891; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–24 final
win
Covered by 0.75
Recorded market spread (unverified close)
-12.25
Source: CFBD median across books; not a verified close
Model edge
+3.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858469; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–41 final
loss
Missed by 10.5
Recorded market spread (unverified close)
+3.5
Source: CFBD median across books; not a verified close
Model edge
+3.20 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858466; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–24 final
loss
Missed by 14
Recorded market spread (unverified close)
-9.75
Source: CFBD median across books; not a verified close
Model edge
+6.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858465; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
19–42 final
win
Covered by 3.5
Recorded market spread (unverified close)
+25.5
Source: CFBD median across books; not a verified close
Model edge
-5.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858464; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–13 final
win
Covered by 12.25
Recorded market spread (unverified close)
-6
Source: CFBD median across books; not a verified close
Model edge
-3.95 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858463; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–19 final
win
Covered by 6.5
Recorded market spread (unverified close)
+4.75
Source: CFBD median across books; not a verified close
Model edge
-5.20 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401858240; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–34 final
loss
Missed by 10.5
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
-5.80 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856881; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–24 final
win
Covered by 18.75
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
-4.15 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856704; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–20 final
win
Covered by 2.25
Recorded market spread (unverified close)
+5.5
Source: CFBD median across books; not a verified close
Model edge
+3.85 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856703; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–31 final
loss
Missed by 1
Recorded market spread (unverified close)
+3.5
Source: CFBD median across books; not a verified close
Model edge
-6.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856702; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–35 final
loss
Missed by 21
Recorded market spread (unverified close)
+8.5
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856698; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
15–21 final
win
Covered by 3.5
Recorded market spread (unverified close)
+10
Source: CFBD median across books; not a verified close
Model edge
-5.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856697; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–34 final
loss
Missed by 21
Recorded market spread (unverified close)
+7
Source: CFBD median across books; not a verified close
Model edge
-3.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 4, game 401856696; settled 2026-09-28T01:06:58.332666+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
18–49 final
loss
Missed by 18.5
Recorded market spread (unverified close)
+12.5
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401867804; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
19–36 final
win
Covered by 2.75
Recorded market spread (unverified close)
+20
Source: CFBD median across books; not a verified close
Model edge
-4.55 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401864574; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
26–21 final
loss
Missed by 12.25
Recorded market spread (unverified close)
-16
Source: CFBD median across books; not a verified close
Model edge
+3.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401864507; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–31 final
win
Covered by 6
Recorded market spread (unverified close)
+26.5
Source: CFBD median across books; not a verified close
Model edge
+3.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401864443; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–27 final
loss
Missed by 10.5
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
-7.70 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401862776; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–14 final
loss
Missed by 0.5
Recorded market spread (unverified close)
-7
Source: CFBD median across books; not a verified close
Model edge
+5.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401861963; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
48–20 final
win
Covered by 25
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401860887; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
26–13 final
win
Covered by 14.5
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401858460; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–54 final
loss
Missed by 20.5
Recorded market spread (unverified close)
+23.5
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401858456; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–55 final
loss
Missed by 1.25
Recorded market spread (unverified close)
+39.5
Source: CFBD median across books; not a verified close
Model edge
-6.65 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401858449; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
0–38 final
win
Covered by 6.5
Recorded market spread (unverified close)
+44.5
Source: CFBD median across books; not a verified close
Model edge
-3.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401858231; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–7 final
win
Covered by 18
Recorded market spread (unverified close)
-9.5
Source: CFBD median across books; not a verified close
Model edge
+3.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856801; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
0–33 final
loss
Missed by 5.5
Recorded market spread (unverified close)
+26.5
Source: CFBD median across books; not a verified close
Model edge
-3.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856800; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
44–30 final
loss
Missed by 3.5
Recorded market spread (unverified close)
-17.5
Source: CFBD median across books; not a verified close
Model edge
+12.20 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856796; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–7 final
win
Covered by 30.5
Recorded market spread (unverified close)
-3.5
Source: CFBD median across books; not a verified close
Model edge
+3.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856795; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–23 final
win
Covered by 0.75
Recorded market spread (unverified close)
-17.5
Source: CFBD median across books; not a verified close
Model edge
-3.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856793; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–42 final
win
Covered by 8.75
Recorded market spread (unverified close)
+34.5
Source: CFBD median across books; not a verified close
Model edge
-4.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856694; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–31 final
loss
Missed by 27
Recorded market spread (unverified close)
-16.5
Source: CFBD median across books; not a verified close
Model edge
+3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856693; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–30 final
win
Covered by 6.25
Recorded market spread (unverified close)
+29.5
Source: CFBD median across books; not a verified close
Model edge
-4.85 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856691; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
34–41 final
loss
Missed by 11
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
+5.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856690; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–14 final
win
Covered by 13.75
Recorded market spread (unverified close)
+21.5
Source: CFBD median across books; not a verified close
Model edge
-5.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856688; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–32 final
loss
Missed by 11
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
-3.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 3, game 401856685; settled 2026-09-20T14:18:32.247296+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
36–50 final
win
Covered by 5.75
Recorded market spread (unverified close)
+18.5
Source: CFBD median across books; not a verified close
Model edge
-3.75 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856677; settled 2026-09-20T18:16:01.963175+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
13–38 final
loss
Missed by 24.75
Recorded market spread (unverified close)
-1
Source: CFBD median across books; not a verified close
Model edge
+4.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401864578; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
29–19 final
win
Covered by 2.5
Recorded market spread (unverified close)
-7.5
Source: CFBD median across books; not a verified close
Model edge
+7.50 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401860883; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
49–3 final
win
Covered by 28
Recorded market spread (unverified close)
-17.5
Source: CFBD median across books; not a verified close
Model edge
+4.90 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856670; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–43 final
loss
Missed by 20.5
Recorded market spread (unverified close)
+12.5
Source: CFBD median across books; not a verified close
Model edge
-3.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401864502; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
32–38 final
loss
Missed by 3.5
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
+4.00 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856671; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
43–8 final
win
Covered by 1.5
Recorded market spread (unverified close)
-32.5
Source: CFBD median across books; not a verified close
Model edge
+3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856676; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–9 final
loss
Missed by 15.5
Recorded market spread (unverified close)
-44.5
Source: CFBD median across books; not a verified close
Model edge
+3.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856682; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
23–24 final
win
Covered by 0.75
Recorded market spread (unverified close)
+2.5
Source: CFBD median across books; not a verified close
Model edge
-7.35 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401858219; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
22–7 final
loss
Missed by 5.5
Recorded market spread (unverified close)
-19.5
Source: CFBD median across books; not a verified close
Model edge
+3.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401867796; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–45 final
win
Covered by 4.25
Recorded market spread (unverified close)
+34.5
Source: CFBD median across books; not a verified close
Model edge
-4.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401858443; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–28 final
loss
Missed by 5.5
Recorded market spread (unverified close)
+13.5
Source: CFBD median across books; not a verified close
Model edge
-3.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401858441; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
7–56 final
loss
Missed by 19.25
Recorded market spread (unverified close)
+30.5
Source: CFBD median across books; not a verified close
Model edge
-6.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856681; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–45 final
loss
Missed by 8.25
Recorded market spread (unverified close)
+12.5
Source: CFBD median across books; not a verified close
Model edge
+5.25 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401871045; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
28–26 final
loss
Missed by 11.5
Recorded market spread (unverified close)
-12.5
Source: CFBD median across books; not a verified close
Model edge
+8.60 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401869955; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–31 final
loss
Missed by 23.5
Recorded market spread (unverified close)
-7.5
Source: CFBD median across books; not a verified close
Model edge
+8.80 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401866414; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–33 final
loss
Missed by 2.75
Recorded market spread (unverified close)
+8.5
Source: CFBD median across books; not a verified close
Model edge
-9.55 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401866418; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
27–29 final
win
Covered by 0.5
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
-5.90 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856684; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
35–26 final
loss
Missed by 12
Recorded market spread (unverified close)
-20.5
Source: CFBD median across books; not a verified close
Model edge
+4.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401862705; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
6–44 final
loss
Missed by 41.25
Recorded market spread (unverified close)
-3
Source: CFBD median across books; not a verified close
Model edge
-8.85 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401858216; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–18 final
win
Covered by 6.5
Recorded market spread (unverified close)
+4
Source: CFBD median across books; not a verified close
Model edge
-5.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401860884; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
31–26 final
win
Covered by 6.5
Recorded market spread (unverified close)
+1.5
Source: CFBD median across books; not a verified close
Model edge
-3.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856673; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
20–70 final
loss
Missed by 9.25
Recorded market spread (unverified close)
+40.5
Source: CFBD median across books; not a verified close
Model edge
-7.45 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856782; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
39–31 final
win
Covered by 31.25
Recorded market spread (unverified close)
+24
Source: CFBD median across books; not a verified close
Model edge
+4.55 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856679; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–10 final
win
Covered by 12.5
Recorded market spread (unverified close)
+5.5
Source: CFBD median across books; not a verified close
Model edge
+6.90 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401864571; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–27 final
loss
Missed by 10.25
Recorded market spread (unverified close)
-6.5
Source: CFBD median across books; not a verified close
Model edge
+5.35 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401856678; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–21 final
win
Covered by 11.5
Recorded market spread (unverified close)
-4
Source: CFBD median across books; not a verified close
Model edge
-3.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 2, game 401858214; settled 2026-09-14T20:22:03.110051+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–28 final
loss
Missed by 3.75
Recorded market spread (unverified close)
+3
Source: CFBD median across books; not a verified close
Model edge
-4.05 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858429; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
30–20 final
loss
Missed by 1
Recorded market spread (unverified close)
-10
Source: CFBD median across books; not a verified close
Model edge
+3.40 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858432; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
56–3 final
win
Covered by 3
Recorded market spread (unverified close)
-50.5
Source: CFBD median across books; not a verified close
Model edge
+9.50 points
Edge tier / historical accuracy
7-14 points / 44%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858204; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
38–16 final
loss
Missed by 0.5
Recorded market spread (unverified close)
-27.5
Source: CFBD median across books; not a verified close
Model edge
+3.10 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858207; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
14–59 final
loss
Missed by 28.5
Recorded market spread (unverified close)
+16.5
Source: CFBD median across books; not a verified close
Model edge
-4.60 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401870763; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
24–21 final
loss
Missed by 13
Recorded market spread (unverified close)
-16.5
Source: CFBD median across books; not a verified close
Model edge
+6.80 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858423; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
21–37 final
loss
Missed by 46.5
Recorded market spread (unverified close)
-29.5
Source: CFBD median across books; not a verified close
Model edge
+4.70 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858209; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
17–3 final
win
Covered by 4.75
Recorded market spread (unverified close)
-9.5
Source: CFBD median across books; not a verified close
Model edge
+4.45 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401856634; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
10–48 final
loss
Missed by 11
Recorded market spread (unverified close)
+28
Source: CFBD median across books; not a verified close
Model edge
-4.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858426; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
40–0 final
win
Covered by 10
Recorded market spread (unverified close)
-31.5
Source: CFBD median across books; not a verified close
Model edge
+4.30 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
ATS result at the saved pick line; cover margin (points)
Model:
SP+ rating-difference fallback
Training:
Unavailable — not recorded in this artifact
Features:
SP+ ratings; source snapshot unavailable
Evaluation:
2026 week 1, game 401858438; settled 2026-09-08T17:09:16.459276+00:00. Exact forecast identity unavailable; historical EPA buckets are a separate reference, not ensemble results.
41–13 final
win
Covered by 7.5
Recorded market spread (unverified close)
-21
Source: CFBD median across books; not a verified close
Model edge
-3.50 points
Edge tier / historical accuracy
3-7 points / 51%
Tier groups the absolute gap between the model and market in points. Accuracy is the saved historical spread-pick win rate for that tier, excluding pushes; it is not this game’s win probability. Sample size and training dates are not included in this saved ledger.
Season / week
2026 / 1
Settled
2026-09-08
Quote gap (proxy)
+0.50 points
Experiment lanes
Graded after the fact at CFBD closing numbers. These are research records of every game that met a lane's rule, not picks made before kickoff and not bets placed. Weeks graded: 1, 2, 3, 4, 5, 6, 7, 8.
Wind unders0–1 (0.0%) · -1.00 units
Outdoor game, wind 10 to 15 mph, total 58 or higher: take the under.
Pre-registered 2026-08-15. Backtest 2017-2023 56.1% (n=453), holdout 2024-2025 55.7% (n=79). Half unit. 2027: 30+ games; under 52.4% ends it, 54% or better keeps it.
Week 5 · Oregon State at Colorado State · Under 61.5 · loss (-1.00)
Humid unders (collection only)3–2 (60.0%) · +0.73 units
Outdoor game, humidity 80% or higher, total 58 or higher: take the under.
Collection cohort X4: 55.8% pooled (n=353); the holdout is driven by 2025 alone, so this is unproven. Not bet. 2027: 30+ games; under 52.4% ends it, 54% or better keeps it.
Week 1 · Florida Atlantic at Florida · Under 59.5 · loss (-1.00)
Week 2 · Louisiana at USC · Under 58.5 · loss (-1.00)
Week 3 · Kennesaw State at Tennessee · Under 60.75 · win (+0.91)
Week 3 · Florida International at Florida Atlantic · Under 62.25 · win (+0.91)
Week 6 · South Carolina at Florida · Under 59.5 · win (+0.91)
Middles0 of 1 middled (0.0%; break-even 4.8%)
Two different books close 3+ points apart: take both sides; it pays when the final margin lands between the numbers.
Backtest: 6.27% of such games middled (n=287) against a 4.76% break-even. Depends on both numbers being available. Collection lane; half unit max.
Week 5 · Marshall at James Madison · James Madison -14.5 (DraftKings) and Marshall +18 (Bovada) · win/loss (-0.09)
Holdover fade3–5 (37.5%) · -2.27 units
Weeks 1-2: bet against a team that finished last season in the Elo top 10 when it plays a non-power-five opponent.
Backtest 58.7% (n=75, 6 of 7 seasons), holdout 63.6% (n=22). Thin sample. Collection only, not bet. 2027: 30+ games; under 52.4% ends it, 54% or better keeps it.
Week 1 · East Carolina at Alabama · East Carolina +28.25 (against Alabama) · loss (-1.00)
Week 1 · North Texas at Indiana · North Texas +40 (against Indiana) · win (+0.91)
Week 1 · Ball State at Ohio State · Ball State +50.75 (against Ohio State) · loss (-1.00)
Week 1 · Marshall at Penn State · Marshall +23.25 (against Penn State) · loss (-1.00)
Week 2 · Western Kentucky at Georgia · Western Kentucky +40.25 (against Georgia) · loss (-1.00)
Week 2 · Charlotte at Ole Miss · Charlotte +44.5 (against Ole Miss) · win (+0.91)
Week 2 · Penn State at Temple · Temple +24.25 (against Penn State) · win (+0.91)
Week 2 · Rice at Notre Dame · Rice +44.75 (against Notre Dame) · loss (-1.00)