NCAA BASKETBALL · MODEL PERFORMANCE

The whole record, one page

Spreads and totals, each shown twice: the premium cut the model actually publishes, and every tracked pick behind it. Both are graded against the closing line, and all four price at the same -110, so unlike the baseball products these win rates can be read against each other directly.

✓ LOGGED BEFORE TIPEvery pick published before the game and graded against the closing line.

3,361

Graded games

65.1%

Spreads win rate

Apr 2, 2026

Graded through

SEASON

SPREADSPREMIUM · EDGE ≥ 6 PTS

65.1%

257W – 138L · 395 graded · 95% CI 60.2–69.6%

published · break-even 52.4%

+24.3%

ROI, 1u at −110

+95.9u

Units won

ALL ATSEDGE ≥ 2 PTS

57.1%

1148W – 864L · 2,012 graded · 95% CI 54.9–59.2%

published · break-even 52.4%

+9.0%

ROI, 1u at −110

+180.7u

Units won

LINE QUALITYSpreads

52.2%

Share of games BBMI landed closer than the market

computed from 395 paired games in this cut

9.18

BBMI margin MAE

9.51

Market margin MAE

Calibration by edge

Spreads · 395 decided

Every pick in this cut, split by the tier its edge places it in. This feed stores no tier on the pick, so the tier is APPLIED from the published thresholds rather than read off the row, and the group is headed as posted because it is what the published thresholds posted, not what today’s would. The total is the product card’s own cohort, which is what makes the two agree.

EDGE (PTS)	TIER AS APPLIED	N	WIN %	ROI
AS POSTED · THE TIER EACH PICK ACTUALLY FIRED UNDER
Edge 6.0+	Premium	395	65.1%	+24.3%
All picks by tier	TOTAL	395	65.1%	+24.3%
BY EDGE INTERVAL · DESCRIPTIVE, SAME PICKS
6.0 to 7.0 pts	Premium	143	61.5%	+17.5%
7.0+ pts	Premium	252	67.1%	+28.1%

The interval rows below do NOT add to the headline the way the tier rows do: a band counts every decided pick at that edge, which is what the thresholds are selecting from rather than what they delivered. This cut is a single tier by construction, so its tier split and its total are the same picks. The all-cut pill beside it is where the two bands separate.

Season to date

Spreads

Cumulative win rate after each week, against the 52.4% break-even. The curve is the running rate, not each week's own, so a late swing moves it less than an early one.

13 weeksbreak-even 52.4%now 65.1%

BY WEEK ▼	PICKS	WIN %	ROI	CUMULATIVE
3/27 ’26	6	50.0%	-4.5%	65.1%
3/20 ’26	21	47.6%	-9.0%	65.3%
3/13 ’26	18	55.6%	+6.1%	66.3%
3/6 ’26	30	60.0%	+14.6%	66.9%
2/27 ’26	33	57.6%	+10.0%	67.5%
2/20 ’26	28	67.9%	+29.6%	68.6%
2/13 ’26	15	86.7%	+65.5%	68.7%
2/6 ’26	29	69.0%	+31.7%	67.6%
1/30 ’26	34	44.1%	-15.7%	67.4%
1/23 ’26	32	71.9%	+37.3%	71.8%
1/16 ’26	59	81.4%	+55.4%	71.8%
1/9 ’26	54	70.4%	+34.4%	65.6%
1/2 ’26	36	58.3%	+11.4%	58.3%

BBMI vs Vegas

395 games

Computed from this product’s own 395 paired games, on the margin. The Record card publishes no closer split for this sport, so it is computed rather than read.

WHICH LANDED CLOSER206 / 179 / 10 tie

BBMI52.2%

VEGAS45.3%

MARGIN MAEBBMI by 0.33 pts

BBMI9.18

VEGAS9.51

Published, across every priced game. The rows above are computed from this cut’s own 395 logged games. These are the site’s own aggregates over 3,361 games, full 2026 season (model launch late Dec 2025), read rather than recomputed, each with the verdict the Record card reached.

PROJECTED WINNER ACCURACYBBMI by 2.6 pts

Share of games each projection’s implied winner was right

BBMI70.6%

VEGAS68.0%

BRIER SCOREBBMI by 0.007

Probability accuracy: rewards honest odds, not bold ones

BBMI0.192

VEGAS0.199

MARGIN MAE, PTSBBMI by 0.03 pts

Projected home margin vs actual margin, in pts

BBMI8.80

VEGAS8.82

GAME-TOTAL MAE, PTSBBMI by 0.24 pts

Projected total vs actual combined score, in pts

BBMI13.13

VEGAS13.36

Where the lines diverge

395 games

6 to 7 pts36.2%

7+ pts63.8%

How far apart the two numbers were on this cut’s games. Every pick here already clears the 6-point floor, so the bands below it are not drawn at all.

Calibration

Brier 0.192 · Published · full 3,361-game sample

A calibrated 60% should win about 60% of the time. Predicted win probability against the actual win rate, across every priced game rather than only the ones a pick fired on. This panel does not follow the product pill: the scorecard publishes one cohort for the sport.

HOME WIN P	N	PRED	ACTUAL	GAP
< 35%	671	21.3%	24.7%	+3.4
35–45%	291	40.2%	45.0%	+4.8
45–55%	354	50.0%	50.0%	-0.0
55–65%	384	59.5%	56.0%	-3.5
≥ 65%	1,661	82.3%	78.7%	-3.6

GAP is actual minus predicted, so a positive number means the model was too cautious in that bucket. Brier 0.192 over 3,361 games; lower is better, and 0.25 is a coin flip.

Products & rules

thresholds from config

These are the definitions in force today. They change through the season as the model is refined, and a change applies from its date forward only: a pick is never refitted to a rule that did not exist when it fired, and the log is never edited to match a newer one. Every pick is counted in the record exactly as it posted.

PRODUCT	THE BET	TRACKED	PREMIUM
Spreads	Against the spread, the sign of the edge picks the side.	edge ≥ 2	edge ≥ 6
Totals	Full-game over / under, the sign of the edge picks the call.	edge ≥ 2	edge ≥ 4

All four cuts settle at -110, so their win rates are directly comparable. Thresholds are read from the published config, never restated here.

All sports →

The log

395 rows

Every pick in this cut, with the market’s number beside ours and which one landed closer. BBMI ERR and VEGAS ERR are each side’s projected margin against the actual, so CLOSER is decided by those two columns and nothing else.

								FINAL
2026-04-02	Stanford @ West Virginia	+1.5	-4.5	6.0	66%	Premium	WIN	77–82	0.5	6.5	BBMI
2026-03-29	Tennessee @ Michigan	-7.5	-13.5	6.0	90%	Premium	WIN	62–95	19.5	25.5	BBMI
2026-03-29	UConn @ Duke	-5.5	-12.5	7.0	88%	Premium	LOSS	73–72	13.5	6.5	VEGAS
2026-03-27	St. John's @ Duke	-6.5	-13.5	7.0	90%	Premium	LOSS	75–80	8.5	1.5	VEGAS
2026-03-27	Alabama @ Michigan	-9.5	-15.5	6.0	93%	Premium	WIN	77–90	2.5	3.5	BBMI
2026-03-27	Michigan State @ UConn	-1.5	+7.5	9.0	24%	Premium	LOSS	63–67	11.5	2.5	VEGAS
2026-03-26	Texas @ Purdue	-7.5	-16.5	9.0	94%	Premium	LOSS	77–79	14.5	5.5	VEGAS
2026-03-26	Iowa @ Nebraska	-1.5	-8.5	7.0	79%	Premium	LOSS	77–71	14.5	7.5	VEGAS
2026-03-26	Illinois @ Houston	-3.5	+6.5	10.0	27%	Premium	WIN	65–55	3.5	13.5	BBMI
2026-03-26	Arkansas @ Arizona	-7.5	-13.5	6.0	90%	Premium	WIN	88–109	7.5	13.5	BBMI
2026-03-25	Illinois State @ Dayton	-7.5	-14.5	7.0	91%	Premium	LOSS	61–55	20.5	13.5	VEGAS
2026-03-22	Kentucky @ Iowa State	-4.5	-10.5	6.0	84%	Premium	WIN	63–82	8.5	14.5	BBMI
2026-03-22	UNLV @ Tulsa	-4.5	-11.5	7.0	86%	Premium	WIN	66–77	0.5	6.5	BBMI
2026-03-22	UCLA @ UConn	-4.5	+1.5	6.0	45%	Premium	LOSS	57–73	17.5	11.5	VEGAS
2026-03-22	Texas Tech @ Alabama	+1.5	-6.5	8.0	73%	Premium	WIN	65–90	18.5	26.5	BBMI
2026-03-21	TCU @ Duke	-11.5	-17.5	6.0	95%	Premium	WIN	58–81	5.5	11.5	BBMI
2026-03-21	Texas @ Gonzaga	-6.5	-14.5	8.0	91%	Premium	LOSS	74–68	20.5	12.5	VEGAS
2026-03-21	Vanderbilt @ Nebraska	+1.5	-4.5	6.0	66%	Premium	WIN	72–74	2.5	3.5	BBMI
2026-03-20	Akron @ Texas Tech	-8.5	+0.5	9.0	48%	Premium	LOSS	71–91	20.5	11.5	VEGAS
2026-03-20	Wright State @ Virginia	-18.5	-6.5	12.0	73%	Premium	WIN	73–82	2.5	9.5	BBMI
2026-03-20	Tennessee State @ Iowa State	-24.5	-11.5	13.0	86%	Premium	LOSS	74–108	22.5	9.5	VEGAS
2026-03-20	Hofstra @ Alabama	-11.5	-2.5	9.0	59%	Premium	LOSS	70–90	17.5	8.5	VEGAS
2026-03-20	Miami (OH) @ Tennessee	-11.5	-2.5	9.0	59%	Premium	LOSS	56–78	19.5	10.5	VEGAS
2026-03-20	Queens University @ Purdue	-25.5	-12.5	13.0	88%	Premium	LOSS	71–104	20.5	7.5	VEGAS
2026-03-20	Prairie View A&M @ Florida	-35.5	-14.5	21.0	91%	Premium	LOSS	55–114	44.5	23.5	VEGAS

All picks locked before tip and graded against the closing line. Past performance is not a guarantee.