Loading...
Prediction: Melissa Dixon by Decision
Reality: Darya Zheleznyakova by Decision
Where the Model Failed:
Underestimated Darya’s forward pressure and physicality. She disrupted Dixon’s rhythm early and forced her to fight backward.
Overvalued Dixon’s composure and volume against someone as aggressive and relentless as Darya.
Missed the judges’ favor toward forward momentum and control, even in closer striking exchanges.
Correction:
Upgrade Darya’s ability to overwhelm technical strikers with pace and aggression.
In closely matched women’s fights, slightly favor the fighter with more physicality and forward movement unless there’s a clear skill gap.
Prediction: Oban Elliott by Decision
Reality: Ko Seok-hyeon by Decision
Where the Model Failed:
Underestimated Ko’s cardio and top control—he maintained relentless pressure for three rounds.
Overestimated Elliott’s ability to stuff takedowns and create separation; he got stuck defending and couldn’t establish range or rhythm.
Ko’s grappling pace was more sustainable than expected and made Elliott look flat.
Correction:
Elevate Ko’s rating as a relentless control grappler with sustainable cardio.
Downgrade fighters like Elliott who rely on being the hammer but haven’t proven they can be the nail under pressure.
Prediction: Jamahal Hill by Decision
Reality: Khalil Rountree by Decision
Where the Model Failed:
Didn’t adequately penalize Hill’s recent knockout loss (vs Pereira) and rushed return.
Underappreciated Rountree’s confidence surge and power-shot timing—he landed clean with composure.
Hill looked slower and less reactive—possibly due to mental or physical effects of the KO layoff.
Correction:
Heavily penalize fighters returning from brutal KOs in under 3–6 months, especially when facing explosive strikers.
Increase volatility score for knockout returnees facing early power threats.
Recognize Rountree’s development as a measured finisher, not just a reckless brawler.