All Sports Performance

AD TEST MODE

PERFORMANCE HEADER

Mock Placement 728x90

College Basketball (M)

basketball • Performance Deep Dive

Trained 3d ago Predicted Never

All-Time Average (24 folds)

Flat ROI

-4.8%

Hit Rate

28.3%

-24.1pp vs b/e

Recent Trend (last 3 folds · Jan 9–Mar 5, 2026 ) ↑ Improving

Flat ROI

+6.1%

+10.9pp vs avg

Hit Rate

32.5%

Out-of-sample Accuracy

64.1%

Avg across 24 folds

AUC-ROC

0.550

0.5 = coin flip · 1.0 = perfect

Log Loss

0.6383

Lower is better

Brier Score

0.2238

Probability calibration

24 validation folds

Jan 26, 2024 – Mar 5, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal. Period: Mar 16, 2026 – Apr 6, 2026

Holdout AUC

0.591

0.5 = coin flip

Fold-by-Fold Metrics

Supporter+ required

Supporter access required

Unlock fold-level charts to identify training stability.

Unlock Supporter

Betting Simulation

Supporter+

Recommendation ROI

�

Recommendation Win Rate

�

Universe ROI

�

Universe Win Rate

�

Supporter access required

Unlock Supporter

Model: XGBoostModel

Calibration: platt

All-Time Average (27 folds)

Flat ROI

+25.6%

Hit Rate

65%

+12.6pp vs b/e

Recent Trend (last 2 folds · Jan 10–Mar 6, 2026 ) ↑ Improving

Flat ROI

+51.6%

+26pp vs avg

Hit Rate

78.6%

Out-of-sample Accuracy

52.7%

Avg across 27 folds

AUC-ROC

0.544

0.5 = coin flip · 1.0 = perfect

Log Loss

0.6889

Lower is better

Brier Score

0.2480

Probability calibration

27 validation folds

Dec 16, 2023 – Mar 6, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal. Period: Mar 16, 2026 – Apr 6, 2026

Holdout AUC

0.631

0.5 = coin flip

Fold-by-Fold Metrics

Supporter+ required

Supporter access required

Unlock fold-level charts to identify training stability.

Unlock Supporter

Betting Simulation

Supporter+

Recommendation ROI

�

Recommendation Win Rate

�

Universe ROI

�

Universe Win Rate

�

Supporter access required

Unlock Supporter

Model: XGBoostModel

Calibration: platt

All-Time Average (27 folds)

Flat ROI

+15.2%

Hit Rate

60.5%

+8.1pp vs b/e

Recent Trend (last 2 folds · Jan 10–Mar 6, 2026 ) ↑ Improving

Flat ROI

+19.5%

+4.3pp vs avg

Hit Rate

62.3%

Out-of-sample Accuracy

53.4%

Avg across 27 folds

AUC-ROC

0.537

0.5 = coin flip · 1.0 = perfect

Log Loss

0.6890

Lower is better

Brier Score

0.2480

Probability calibration

27 validation folds

Dec 16, 2023 – Mar 6, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal. Period: Mar 16, 2026 – Apr 6, 2026

Holdout AUC

0.510

0.5 = coin flip

Fold-by-Fold Metrics

Supporter+ required

Supporter access required

Unlock fold-level charts to identify training stability.

Unlock Supporter

Betting Simulation

Supporter+

Recommendation ROI

�

Recommendation Win Rate

�

Universe ROI

�

Universe Win Rate

�

Supporter access required

Unlock Supporter

Model: XGBoostModel

Calibration: platt