Loading...
Prediction: Naimov by Decision
Reality: Santos by KO/TKO
1. Underestimated Santos’ physicality & explosiveness
Pre-fight model treated Santos as a “regional chaos striker” but he showed UFC-level power + timing early.
We did not apply enough PPA because we assumed Naimov was the physically stronger fighter.
2. Overvalued Naimov’s durability
DIC should have been harsher: Naimov had been rocked multiple times historically.
Santos’ early chaos (CTM) should have had a higher weight.
3. Incorrect grappling expectation
Model assumed Naimov could neutralize Santos’ aggression with clinch/wrestling. He couldn’t get respect early.
Increase volatility and power-risk flags when a pressure wrestler faces an explosive young striker with underrated power indicators.
Prediction: Aslan by KO/TKO
Reality: Baraniewski by KO/TKO
1. Over-reliance on Aslan’s early KO pattern
The model overweighted Aslan’s “round-1 death touch” despite inconsistent defensive structure.
2. Underestimated Baraniewski’s toughness and power
DIC wasn’t applied enough to Aslan (he has been hurt before).
Baraniewski had improving striking, but model treated him as a less dangerous underdog.
3. Incorrect physicality assumption
Baraniewski matched (and surpassed) Aslan’s explosiveness.
PPA bias favored Aslan too strongly.
Avoid overvaluing one-dimensional early KO artists unless their defensive durability also passes thresholds.
Prediction: McKinney by KO/TKO R1
Reality: Duncan by Submission R1
1. Underweighted McKinney’s grappling liability
Duncan’s submission game is rarely used, but McKinney’s reckless scrambles made it easy.
2. CTM was correct but incomplete
Chaos = volatility, but the model incorrectly assumed chaos favors McKinney.
Duncan’s composure under fire is much better than earlier in his career.
3. Overvalued McKinney’s early threat
June logic update says not to overvalue early finishes vs low-tier opponents—this should have applied harder.
In high-volatility fights, always consider self-destructive grappling mistakes as finish paths.
Prediction: Dawson by Decision/Sub
Reality: Torres by KO/TKO
1. Massive underestimation of Dawson’s post-KO fragility
Dawson had just been brutally KO’d — DIC penalty should have been much heavier.
Fighters returning from KO < 12 months are extremely fragile.
2. Overassuming Dawson would get takedowns
No strong setup striking → Torres landed early before grappling began.
3. Underweighting Torres’ explosiveness
Again, the model mistakenly downplayed Torres’ chaos threat because of opponent quality earlier in his UFC run.
→ When a wrestler returns from a KO and faces an explosive striker, volatility should be upgraded into “danger zone” territory.
Prediction: Pantoja by Decision/Sub
Reality: Van by KO/TKO (arm dislocation injury)
This was not predictive failure - it was a freak accident.
The model would NOT adjust based on this outcome, because the fight did not reflect competitive reality.
No logic adjustment. Injury-based outcomes should not influence predictive calibration.
Prediction: Merab by Decision
Reality: Yan by Decision
1. Overvalued Merab’s pressure against elite defensive technicians
Memory note #49 already warned about downgrading pressure fighters vs elite defensive strikers.
Merab’s pressure was neutralized by Yan’s improved footwork and defensive shifts.
2. Underestimated Yan’s anti-wrestling improvements
Yan stuffed takedowns with more urgency and maintained balance better than expected.
3. Ignored Merab’s pace decline indicators
PDT should have been applied. His output has been dropping over last two fights.