Skip to content
Browse › Calibration Over Time
Analytics module · as of

MLB calibration is improving month over month -- but still trails the market

The MLB model's calibration error fell from 0.1165 in June to 0.0834 in July, yet the market stayed sharper in both months (0.0944, then 0.0635).
View full size ↗
confirmednull (a finding)not testabledescriptivepending
Calibration Over Time
Chart: Calibration Over Time -- The MLB model's calibration error fell from 0.1165 in June to 0.0834 in July, yet the market stayed sharper in both months (0.0944, then 0.0635).
scripts/platformkit/analytics_showcase/out/calibration_over_time.json

What it means

Splitting the record by calendar month checks whether the model is stable or drifting. MLB is encouraging -- the model got better-calibrated as July games came in (ECE 0.1165 to 0.0834) -- but the market improved in lockstep and kept its lead. Soccer went the other way: the model's monthly Brier climbed from 0.1729 to 0.331.

Caveats & confounds

July soccer rests on only n=3129 rows, so its worse numbers may reflect a small, harder sample rather than genuine month-over-month degradation.

Ask Scout about this
Is calibration improving month over month?Ask anything →