Skip to content

NCAAB model accuracy

NCAA Division I Men's Basketball

Every final NCAAB game, forecast with only what was known the day before and then scored. Lower Brier and log loss are better; a coin flip scores 0.250 Brier. All leagues

Scorecard

Accuracy of each model on final games
ModelGamesBrierLog lossAccuracySkillCalib. error
Game simulator (backtest)5,8120.2010.58568.4%+13.3%6.1%
Elo5,8120.1970.57970.4%+14.8%3.6%
Base rate5,8120.2320.65664.3%—6.2%
Published forecasts18small sample0.1260.41283.3%+17.1%9.3%

Calibration

When a model gives the home side a 60% chance, does it win about 60% of the time? Dots on the dashed line mean yes; whiskers show the 90% range the observed rate could land in by chance.

Brier score by month

Early-season forecasts lean on last season's regressed ratings, so expect them to improve as games accumulate.

Head to head

Game simulator (backtest) vs Elo

5,812 games both forecast

Brier
0.201 vs 0.197
Log loss
0.585 vs 0.579
Accuracy
68.4% vs 70.4%

Elo has the lower Brier score here.

Published forecasts vs Game simulator (backtest)

18 games both forecast

Brier
0.126 vs 0.148
Log loss
0.412 vs 0.477
Accuracy
83.3% vs 83.3%

Published forecasts has the lower Brier score here, but on too few games to call.

Market consensus isn't connected: it needs a licensed odds source, which an admin can import. Until then the base rate and Elo are the benchmarks.

How this is measured

The backtest replays the game simulator for every final game from the efficiency ratings and Elo as they stood before that date, the league's scoring and home advantage so far, rest and travel, the venue and the game's weather. Injuries and lineups of past games aren't on record, so the replay is team level; the published forecasts, which include them, are scored separately as they accumulate.

Brier = ½ Σ (probability − outcome)² over home, away. Log loss = −ln(probability given to the result), with probabilities clipped at 0.5%. Tied finals are left out. Full details are in the model methodology document.

Pick record

Every forecast the site published before a game, scored once it finished. A pick is the likeliest outcome. A well-calibrated model's 70% picks come in about 70% of the time, so compare each row's hit rate with its confidence.

Record

15–3 83%

Average confidence 80%

Hit rate by confidence
Confidence PicksHit
70%–80%667%
80%+1292%
Latest picks
GamePickResult
Nov 6 EWU @ OREORE 80%71–98
Nov 6 VAL @ PURPUR 71%83–80
Nov 6 GASO @ GTGT 76%83–95
Nov 6 BU @ BRWNBRWN 74%72–67
Nov 6 APSU @ KENNKENN 80%69–71
Nov 6 INST @ BALLBALL 82%69–66
Nov 6 ARST @ ALAALA 79%86–101
Nov 6 DEL @ WISWIS 83%64–73
Nov 6 FAU @ WRSTWRST 82%71–97
Nov 6 DETM @ MARQMARQ 85%32–93
Nov 6 STBK @ NUNU 85%65–78
Nov 6 IDST @ WYOWYO 80%62–93
Nov 6 MTSU @ KUKU 84%45–89
Nov 6 WAG @ MIAMIA 76%57–83
Nov 6 UNCO @ ASUASU 81%63–90
Nov 6 IDHO @ CSUCSU 80%67–78
Nov 6 CSUN @ KSUKSU 83%78–95
Nov 6 CBU @ ORSTORST 84%68–100

Data as of Source: SportsAPI Pro

Cookies

Cookies keep you signed in. Optional ones remember your settings. No ads or tracking. Cookie policy