Oracle Bantamweight Prediction Accuracy Report
Quick Answer
Bantamweight currently reports 76.9% accuracy across 13 settled predictions with a Brier score of 0.2101 and 38.5% method accuracy. Cross-promotion weight-class slice.
Scope: Bantamweight. Cross-promotion weight-class slice.
Machine-readable companion: /track-record/oracle-bantamweight-accuracy/summary.json
Track Record Snapshot
Overall Accuracy
The Oracle AI has made 22 verified predictions with an overall accuracy of 77.3%.
Cross-promotion weight-class slice.
Brier Score: 0.2159 — Grade: D
Confidence Tier Performance
- Lock (85%+): 0/0 (0% win rate)
- High (70-84%): 0/0 (0% win rate)
- Medium (60-69%): 2/3 (66.7% win rate)
- Low (50-59%): 7/9 (77.8% win rate)
Methodology
All predictions are recorded before fights occur and verified against official results. Draws and No Contests are excluded from accuracy calculations.
Generated automatically by The Oracle, Blueprint MMA's automated UFC prediction engine.
Methodology and Attribution
Author: Blueprint MMA
Creator: The Oracle software
How this was produced
The Oracle generates each prediction automatically from the locked source packet. Automated identity, evidence, probability, and publication checks run before release; no individual human reviews each prediction before publication.