UFC 328 tonight — Chimaev vs Strickland · 12 fights · live Polymarket odds + Trifecta v2 picks

Predict
every
fight.

A dual-ensemble ML model trained on 8,500+ fights. Live-trial accuracy 55.3% on n=47 across 4 cards. LOCK tier 11-0 in 2026 (mostly mechanical agreement with extreme market favorites — see methodology). No guesswork, no gut feelings — just data.

Presented by mouthguards.pro Custom Combat Mouthguards

The numbers speak for themselves

Live 2026 results + 7-year walk-forward backtest (2020–2026). Not cherry-picked. Not overfit. Zero data leakage.

11-0
LOCK Picks in 2026
10 mechanical agreement · 1 real LOCK (Malkoun, Perth)
84.1%
High-Confidence Accuracy
37-7 · LOCK + HIGH tier · 2026
55.3%
Live Trial Accuracy
26 of 47 fights · 4 cards live
$1,404
2026 Profit (cum.)
115 bets · flat $100 · backtest + live
+48.8%
Method Prop ROI (2026)
2026 window · lifetime +22.3%
+74.1%
Round Prop ROI (2026)
2026 window · lifetime +24.9%

Past performance does not guarantee future results. Live-trial cumulative is the honest 2026 number; the 75.6% headline blended walk-forward backtest with live trial — split out above.

Built different

Most prediction sites give you gut feelings dressed up as analysis. We give you a machine that learns from every fight ever thrown.

01

Every Fight. Ever.

8,500+ UFC fights spanning 12+ years. Striking, grappling, physical attributes, Elo ratings, historical odds — all cleaned and structured.

02

Engineered Intelligence

We don't just look at records. Proprietary features capture style matchups, momentum, aging curves, and how fighters perform under pressure.

03

Multi-Model Ensemble

Four algorithms compete and collaborate. A dual architecture blends market-aware and market-blind predictions to avoid overfitting.

04

Find the Edge

We scan betting markets in real time. When our model disagrees with the market, we flag it. That's where the money is.

Four models. One prediction.

Each algorithm sees the fight differently. When they agree, we're confident. When they don't, we tell you.

Model A

Statistical foundation — finds the signal in the noise

Model B

Pattern recognition — learns complex matchup dynamics

Model C

Speed optimised — processes every angle at scale

Model D

Robustness engine — guards against overfitting

This Saturday's predictions

UFC 328 tonight · Chimaev vs Strickland headline · 12 fights · Trifecta v2 selects 5. Below: the headline picks model + market disagree on most.

UFC 328 · Tonight
9 May 2026 · main card 10 PM ET / 3 AM BST · live Polymarket odds
LIVE
Khamzat Chimaev
84.2%
HIGH
SUB 34% · LEAN +1.7pp
Sean Strickland
15.8%
Grant Dawson
71.5%
MED
Dec 53% · STRONG +11.0pp
Mateusz Rebecki
28.5%
Sean Brady TRI v2
56.1%
TRIFECTA
Brady by DEC · 2.96× fair odds
Joaquin Buckley
38.8%
Pat Sabatini
45.8%
UNDERDOG
Sub threat · +14.3pp value
William Gomis
54.2%

Not all picks are equal

Every prediction comes with a confidence tier. The edge lives in LOCK and HIGH — that's where you should focus.

LOCK
11-0
2026 · mostly mechanical agreement w/ market
HIGH
79.4%
2026 · 27-7
MED
67%
solid · worth tracking
LOW / TOSS-UP
~55%
coin flip · proceed with caution

Spot mispriced fights

We scan Polymarket and sportsbook lines in real time and compare them against our model probabilities. When the market is wrong, we flag it.

If our model says a fighter wins 76% of the time but the market prices them at 62%, that's a +14% edge. We surface these opportunities so you can act on them.

Edge Detection — UFC 328 · Tonight
Sabatini vs Gomis
Model 45.8%
Market 31.5%
+14.3% UDV
Susurkaev vs Santos
Model 30.2%
Market 18.0%
+12.2% UDV
Dawson vs Rebecki
Model 71.5%
Market 60.5%
+11.0% STRONG
◆ Six UNDERDOG VALUE flags this card. Full card + Trifecta v2 picks →

Every card. Every prediction. Every result.

Full pre-fight analysis, FightIQ model predictions, and post-fight results for every UFC event in 2026.

Get predictions before fight night

Join the waitlist and be the first to receive model output, edge alerts, and confidence-rated picks for every UFC event.