The same fights, but with alternate primes / dates / weight-class shifts. Each variant is a separate model run on the V14 ensemble — conversation pieces, not predictions.
The model picks Jack Della Maddalena based on a unanimous consensus across four sub-models. The model favours Jack Della Maddalena due to a positive Elo Rating Gap and a higher Output in significant strikes per fight. Additionally, the model identifies an advantage for Jack Della Maddalena in the Striking Index Gap.
The model picks Jack Della Maddalena with a unanimous consensus across all four sub-models. This preference is driven by a positive Elo Rating Gap of 56 points and a higher Output in significant strikes per fight. Additionally, the model favours Jack Della Maddalena due to a notable advantage in the Striker vs Grappler Matchup.
The model picks Quillan Salkilld with a 74% probability based on a unanimous four-way ensemble. The model favours Salkilld due to advantages in the Striking Index Gap, the Striker vs Grappler Matchup, and the Output gap. Additionally, the model identifies a negative Win/Loss Streak gap for Salkilld compared to Dariush.
The model picks Quillan Salkilld with a 67% probability based on a unanimous four-way ensemble. The model favours Salkilld due to a specific Striking Index Gap and a Reach Advantage of 3.0 inches. Additionally, the model identifies a Striker vs Grappler Matchup edge for Salkilld.
The model picks Steve Erceg based on a unanimous consensus across all four sub-models. The model favours Erceg due to his advantages in the Striking Index Gap, Output, and Average Fight Duration Gap. While Tim Elliott holds an advantage in the Experience Gap, the model's output remains focused on the statistical edges identified for Erceg.