Menu
Login Get Started
All Sports Performance
AD TEST MODE
PERFORMANCE HEADER
Mock Placement 728x90
UF

UFC / MMA

combat • Performance Deep Dive

Trained Never Predicted Never
All-Time Average (151 folds)
Flat ROI
+84.7%
Hit Rate
67.8%
+15.4pp vs b/e
Recent Trend (last 0 folds · Nov 22–Jan 31, 2026 )
Flat ROI
Hit Rate
Out-of-sample Accuracy
63.3%

Avg across 151 folds

AUC-ROC
0.684

0.5 = coin flip · 1.0 = perfect

Log Loss
0.6071

Lower is better

Brier Score
0.2140

Probability calibration

151 validation folds
Mar 4, 2017 – Jan 31, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal.

No holdout data available. Re-run prediction-system run --sport ufc to generate holdout metrics.

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 (151 folds)
Flat ROI
+48.6%
Hit Rate
66.1%
+13.7pp vs b/e
Recent Trend (last 2 folds · Nov 22–Jan 31, 2026 ) ↑ Improving
Flat ROI
+82.9%
+34.3pp vs avg
Hit Rate
80.0%
Out-of-sample Accuracy
66.0%

Avg across 151 folds

AUC-ROC
0.683

0.5 = coin flip · 1.0 = perfect

Log Loss
0.5793

Lower is better

Brier Score
0.2038

Probability calibration

151 validation folds
Mar 4, 2017 – Jan 31, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal.

No holdout data available. Re-run prediction-system run --sport ufc to generate holdout metrics.

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: LightGBMModel
Calibration: platt
All-Time Average (151 folds)
Flat ROI
+100.0%
Hit Rate
80.2%
+27.8pp vs b/e
Recent Trend (last 3 folds · Nov 22–Jan 31, 2026 ) ↑ Improving
Flat ROI
+170.0%
+70pp vs avg
Hit Rate
57.1%
Out-of-sample Accuracy
72.5%

Avg across 151 folds

AUC-ROC
0.675

0.5 = coin flip · 1.0 = perfect

Log Loss
0.5348

Lower is better

Brier Score
0.1828

Probability calibration

151 validation folds
Mar 4, 2017 – Jan 31, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal.

No holdout data available. Re-run prediction-system run --sport ufc to generate holdout metrics.

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: LightGBMModel
Calibration: platt

No backtest data yet

The model hasn't been run for this market yet. Data will appear here once the pipeline has been executed.

All-Time Average (152 folds)
Flat ROI
+100.0%
Hit Rate
73.2%
+20.8pp vs b/e
Recent Trend (last 0 folds · Dec 13–Feb 21, 2026 )
Flat ROI
Hit Rate
Out-of-sample Accuracy
64.3%

Avg across 152 folds

AUC-ROC
0.657

0.5 = coin flip · 1.0 = perfect

Log Loss
0.6320

Lower is better

Brier Score
0.2211

Probability calibration

152 validation folds
Mar 4, 2017 – Feb 21, 2026

Final Holdout Test

Held back from all training and fold evaluation — the cleanest out-of-sample signal.

No holdout data available. Re-run prediction-system run --sport ufc to generate holdout metrics.

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: LightGBMModel
Calibration: platt