Automated tracking
Statistics
Batch-generated aggregates from independent performance and pricing models. Updated without an admin console via scheduled rebuilds.
Dataset synthetic-v1 · generated 2026-07-26T02:09:30.013Z
Synthetic events
18,000
Sports covered
9
Average overround
6.48%
Model versions
1.0.0, 1.1.0
Calibration significance
Does the synthetic model calibrate?
Across every odds band in this batch, a Pearson chi-square goodness-of-fit compares observed win rates to price-implied rates. Individual bands are tested with a 95% Wilson interval and a proportion z-test — a deviation only counts when it clears both significance and a minimum sample. Open a band or a sport to see the full breakdown.
Bands tracked
12
χ² (df)
33.5 (11)
Approx p-value
0.000
Significant bands
7
Real-world use case
- Reload a sport's statistics page with a different batch seed and check whether a flagged band stays significant — one-off flags are usually noise.
- Compare a favourite band's tight Wilson interval against a longshot band's wide one to see why long prices need far more events before a deviation means anything.
1.0.0
Horse racing
2,000 events · avg margin 18.18%
1.1.0
Soccer
2,000 events · avg margin 5.01%
1.1.0
American football
2,000 events · avg margin 5.07%
1.1.0
Ice hockey
2,000 events · avg margin 4.84%
1.1.0
Basketball
2,000 events · avg margin 5.16%
1.1.0
Baseball
2,000 events · avg margin 4.26%
1.1.0
Tennis
2,000 events · avg margin 4.34%
1.1.0
Golf
2,000 events · avg margin 6.45%
1.1.0
Handball
2,000 events · avg margin 5.02%