UFC Fight Intelligence

Model-Driven
Fight Prediction

Cage Analytics builds a statistical profile for every UFC fighter — tags, archetypes, momentum scores, and style matchups — then applies a multi-factor model trained on historical fight outcomes.

Get started free Already have an account →
74%
Prediction Accuracy
Cross-validated
19,052
Bouts Analyzed
Training set
500+
Fighters Tracked
Active roster
41
Predictive Tags
Backtested signals
How the model works
01
Ingest
Full fight history — round-by-round strikes, takedowns, submission attempts, and finish data for every fighter on the UFC roster.
02
Tag & Classify
41 classification signals applied per fighter — stance, physical traits, finishing ability, grappling style, and 8 momentum signals with backtested win rates.
03
Archetype
15 fighting style archetypes assigned based on stat signatures. Each archetype carries a tier score (−1 to 3) validated against historical win rates.
04
Momentum Score
A 0–100 form score built from recent record, finish rate, opponent quality, activity, and stat trend — time-weighted so recent fights count most.
05
Edge Prediction
Multi-factor matchup scoring across momentum, striking, grappling, and tag signals — producing a confidence-tiered edge call per fight.
Top predictive signals
on_a_run
83%
rising_star
82%
power_striker
79%
chin
79%
layoff_risk
25%
on_a_skid
38%
Win rate when fighter with tag faces fighter without. Tested across all historical bouts in dataset.
Ready to dig in?
Card analysis, matchup tool, momentum rankings, roster browser, and more — free.
Create an account Sign in →