Did the model beat the market?
Charts home
Rankings
Scorecard
Matchup
Model
Season
Tournament
Dashboard
Three predictors scored on the same season, and the calibration curve that
decides which of them means what it says.
Each predictor is scored only on the games it actually priced. Read
Forecast?
in the table first: a model allowed to see the season it is predicting is not
forecasting, it is remembering. Only Elo and the market consensus are honest
on that column, and the game counts differ because each model prices a
different slice of the schedule.
Every model, every scoring rule
Regular season. A coin flip scores 0.250 on Brier and 0.693 on log loss.
Model
Accuracy
Brier
Log Loss
Skill vs Coin Flip
Margin Error
Margin Bias
Games
Forecast?
Blended season ratings
76
%
0.161
0.483
36
%
8.6
−0.22
5,435
No
Market consensus
72
0.178
0.524
29
9.0
+0.17
5,431
Yes
Elo, pregame
73
0.186
0.553
26
12.4
−4.29
5,968
Yes
0.000
0.050
0.100
0.150
0.200
Blended season ratings
Market consensus
Elo, pregame
0.161
0.178
0.186
Brier score, lower is better
Mean squared error on the probability itself, not on the pick
2022
2023
2024
2025
2026
70%
75%
Regular-season accuracy
Elo, season by season
The only model here that never sees the season it is predicting
Reliability
Accuracy rewards a model for being right. Calibration asks the harder
question: whether its confidence means anything. A line above the diagonal is
a model underselling itself, and one below it is a model charging for certainty
it does not have.
0%
20%
40%
60%
80%
100%
Probability the model stated
0%
50%
100%
Share that actually won
When a model says 70%, how often does it happen?
One point per probability bucket. A model on the diagonal means what it says.
Perfect calibration
Elo, pregame
Blended season ratings
Market consensus
Regular season
Other postseason
NCAA tournament
0%
20%
40%
60%
80%
Accuracy
Blended season ratings
Elo, pregame
Market consensus
The tournament is where confidence goes to die
Same models, same season, harder games
Data as of 01:14 UTC on 16 Sep 2026
made with
dbt Charts