Model Performance

2024 Season · 3,078 graded predictions

2026 2025 2024 2023 2022 All

In-sample data: All predictions for this season were generated after games completed (historical backfill), so metrics reflect in-sample fit rather than true out-of-sample accuracy. Metrics will become trustworthy as the current season progresses and daily ETL generates pre-game predictions.

66%

Overall Accuracy

68%

Last 50 Games

0.2060

Brier Score (↓ better)

0.5947

Log Loss (↓ better)

0.7284

AUC (↑ better)

3,078

Predictions Graded

Calibration

When the model says X%, does the favored team win X% of the time? The vertical bar shows the predicted probability; the filled bar shows actual.

50–55%381 games51% actual(-2pp)

55–60%418 games51% actual(-7pp)

60–65%445 games55% actual(-8pp)

65–70%450 games64% actual(-4pp)

70–75%355 games70% actual(-2pp)

75–80%339 games73% actual(-4pp)

80–100%690 games87% actual

Actual win rate Predicted

Score Projection Accuracy

How close were the Monte Carlo score projections to actual results?

4.7

Avg margin error (runs)

+1.0

Margin bias (under-projects)

56%

Own-line cover rate

Within 2 runs of projected margin28%

Within 5 runs of projected margin62%

Based on 3078 games with score projections.

D1 Diamond Top 25 — 2026 Season

Composite ranking of all D1 teams with ≥5 games played.

#	Team	Record	Elo	wOBA	FIP	Run Diff	Score
1	UCLAUCLA Bruins	38–3	1847	0.426	3.69	+5.3	91
2	GTGeorgia Tech Yellow Jackets	34–7	1820	0.480	4.32	+6.1	91
3	TEXTexas Longhorns	31–8	1794	0.418	2.89	+4.5	87
4	UNCNorth Carolina Tar Heels	36–8	1832	0.405	3.92	+4.4	85
5	MSSTMississippi State Bulldogs	32–10	1764	0.421	3.24	+4.7	83
6	UGAGeorgia Bulldogs	32–10	1769	0.463	4.19	+4.5	83
7	TA&MTexas A&M Aggies	31–8	1778	0.449	4.72	+4.9	83
8	AUBAuburn Tigers	29–12	1761	0.391	3.16	+3.1	79
9	ORSTOregon State Beavers	31–9	1747	0.383	3.03	+3.1	79
10	MISSOle Miss Rebels	30–12	1766	0.392	3.63	+3.0	78
11	FSUFlorida State Seminoles	29–12	1744	0.391	3.52	+3.0	76
12	OREOregon Ducks	30–11	1706	0.424	4.17	+3.2	75
13	KUKansas Jayhawks	31–11	1761	0.403	4.41	+2.0	75
14	FLAFlorida Gators	29–14	1722	0.369	3.08	+2.2	73
15	MIAMiami Hurricanes	30–12	1693	0.418	4.14	+3.9	73
16	JVSTJacksonville State Gamecocks	34–9	1673	0.397	3.67	+3.4	73
17	NEBNebraska Cornhuskers	31–11	1719	0.411	4.34	+2.6	73
18	TENNTennessee Volunteers	29–15	1732	0.383	3.96	+2.5	72
19	USMSouthern Miss Golden Eagles	30–12	1699	0.381	3.30	+1.7	72
20	ARKArkansas Razorbacks	29–15	1742	0.379	4.29	+2.5	71
21	WAKEWake Forest Demon Deacons	28–15	1687	0.420	4.36	+3.5	71
22	OUOklahoma Sooners	28–13	1723	0.390	4.15	+2.3	71
23	WVUWest Virginia Mountaineers	27–10	1695	0.401	4.30	+3.1	71
24	ASUArizona State Sun Devils	29–14	1652	0.436	4.17	+3.0	71
25	UCSBUC Santa Barbara Gauchos	27–12	1685	0.372	3.41	+2.4	70

Minimum 5 games played. wOBA/FIP from FanGraphs.

By Confidence Tier

Accuracy when the model is highly confident in its pick.

Confidence	Games	Accuracy
≥ 60%	2279	72%
≥ 65%	1834	76%
≥ 70%	1384	80%
≥ 75%	1029	83%

Top Teams by Elo

Rolling Elo rating (1500 = average). Updated after every game result.

Rank	Team	Games	Elo
1	UCLA Bruins	276	1847
2	North Carolina Tar Heels	293	1832
3	Georgia Tech Yellow Jackets	276	1820
4	Texas Longhorns	290	1794
5	Texas A&M Aggies	292	1778
6	Georgia Bulldogs	277	1769
7	Ole Miss Rebels	281	1766
8	Mississippi State Bulldogs	273	1764
9	Auburn Tigers	278	1761
10	Kansas Jayhawks	267	1761

How to Read This

Overall Accuracy — % of games where the model's favored team won.

Brier Score — measures probability calibration quality. Lower is better.

Log Loss — penalises confident wrong predictions more harshly than Brier score. Lower is better.

AUC — probability that the model ranks a random home win above a random home loss. 0.5 = no skill; 1.0 = perfect.

Calibration chart — a perfectly calibrated model's bar would exactly touch the line in every bucket.

Elo ratings — self-correcting power ratings that update after every game. 1500 is average.