odds, translated: +205 ≈ 33% implied chance · -230 ≈ 70%
Our model picks Costa by decision. The model gives Costa 56%.
The current price implies roughly 70%. The market likes Costa even more than the model does.
Durden still wins this 44% of the time in the model, and with a 53% chance the fight ends inside the distance, one clean sequence can beat the forecast.
Model pick, no official bet. A prediction is not automatically a bet; the price didn't clear the rules on this one.
Bigger polygon = more of that trait. Career averages (not opponent-adjusted); the date is each fighter's last bout.
Each cell is the model's chance the fight ends that way, in that round. Brighter = more likely; hover for the exact number.
The line is the model's chance the fight is still going after each round — where it drops, finishes cluster.
Moneyline edges are the model's calibrated edge vs the no-vig sharp fair. Prop edges marked ~ have no sharp line (thin / one-book) — treat as unverified. Only the official plays above are bets; everything else here is model reference. Prices as of Jul 10, 11:00 UTC.