7 Years. 2,475 Fights.
Zero Data Leakage.

Walk-forward backtest results for the FightIQ V12 dual-ensemble model. Every prediction made using only data available before fight night — no hindsight, no cheating, no exceptions.

Walk-forward period: Jan 2020 — Apr 2026 • 1,990 fights with odds data

Past performance does not guarantee future results. Gamble responsibly.
Headline Performance
Seven years of out-of-sample predictions. The model sees only historical data at the time of each prediction.
2,475
Total Fights Tested
walk-forward 2020-2026
66.2%
Overall Accuracy
1,990 fights with odds
81.3%
LOCK + HIGH Accuracy
460 fights
87.4%
LOCK Tier Accuracy
87 fights, 76 wins
+$1,235
LOCK + HIGH Profit
flat $10 per fight
+2.7%
LOCK + HIGH ROI
460 bets
Not All Picks Are Equal
The model assigns a confidence tier to every prediction. The edge lives in LOCK and HIGH — that's where you should focus.
Moderate
MED
67.3%
Fights 701
Wins 472
P&L -$1,259.18
ROI -1.8%
Low Signal
LOW
64.6%
Fights 379
Wins 245
P&L +$2,988.93
ROI +7.9%
Avoid
TOSS-UP
50.4%
Fights 450
Wins 227
P&L -$3,165.46
ROI -7.0%
The Model Is Getting Better
Accuracy has trended upward from 63% to 74% as the training dataset grows and the ensemble learns. Each year uses only prior data.
2020
63.1%
339 fights
-$370.03
P&L
2021
62.0%
413 fights
-$1,075.84
P&L
2022
64.0%
419 fights
-$821.06
P&L
2023
67.4%
389 fights
+$1,948.95
P&L
2024
66.9%
429 fights
-$942.50
P&L
2025
69.9%
425 fights
+$651.00
P&L
2026
73.8%
61 fights (YTD)
+$408.86
P&L
63.1% in 2020 → 73.8% in 2026 — accuracy trending up
Methodology
Built for transparency and statistical rigour. No black boxes.

V12 Dual Ensemble

Two independent model stacks (4 algorithms each) make separate predictions. The final output blends both ensembles, reducing variance and overfitting risk.

4 Algorithms

Gradient-boosted trees, logistic regression, neural network, and random forest. Each brings different strengths — the ensemble captures what no single model can.

Walk-Forward Testing

Every prediction in this backtest was made using only data available before that fight. The model is retrained on a rolling window — no future information ever leaks in.

5-Year Half-Life

Time-weighted training data with a 5-year half-life. Recent fights count more, but deep history still matters. The sport evolves — the model adapts.

The most common mistake in backtesting is data leakage — accidentally using future information to make predictions about the past. This makes results look incredible on paper but useless in reality.

Walk-forward testing eliminates this entirely. Imagine you're standing in January 2022 and want to predict a fight. The model can only use data from before that date — fighter records, historical stats, training camp info, all from the past. It makes a prediction, we record it, then move forward in time.

We repeat this for every fight from 2020 to 2026. The model is periodically retrained as new data becomes available, but it never sees the outcome of a fight before predicting it. This is exactly how the model runs in production today.

The 66.2% overall accuracy and 81.3% LOCK+HIGH accuracy you see on this page are real out-of-sample results — the same kind of performance you can expect going forward.

See it in action

Check out the latest UFC predictions — every fight graded by confidence tier, with full probability breakdowns.

View UFC 327 Predictions Back to FightIQ