Kevin Borjas vs Andre Lima, UFC Fight Night: Kape vs. Horiguchi prediction and deep analysis
Kevin Borjas/Andre Lima
FlyweightWinner Committee
Thirty-three specialists answer distinct winner-only questions. Eligible evidence is renormalized only inside its original lane.
Fight-State Projection
These are distributions of eligible committee signal mass—not invented percentages of fight time.
Conditional Method Model
Each row is conditional on that fighter already winning; every fighter row sums to 100%.
Conditional Round Model
Round-event mass is derived from survival-conditioned hazards. The final scheduled round is absorbing.
Probability Reconciliation
The exact-outcome grid preserves both winner marginals and sums to 100%. Decision branches have no round.
Output Selection
The winner and every outcome label come from one immutable reconciled distribution. Oversight may widen uncertainty, never move the point forecast.
Key Takeaways
- The War Room Committee picks Andre Lima (58% win probability).
- Andre Lima path. "Defensive volume to decision": Use height to manage range and force Borjas to overextend
- Kevin Borjas path. "Early power exchanges": Close distance behind feints to bypass Lima's reach
- Risk: Layoff timing collapse
- Risk: Chin gives out under first clean shot
Frequently Asked Questions
Who won Kevin Borjas vs Andre Lima?
Kevin Borjas won. Blueprint MMA's pre-fight pick was Andre Lima.
Who wins Kevin Borjas vs Andre Lima?
Blueprint MMA's War Room Committee picks Andre Lima with a 58% win probability.
What could change the outcome of Kevin Borjas vs Andre Lima?
Layoff timing collapse: Lima's extended absence could manifest in mistimed entries, letting Borjas land the power shots his volume deficit demands. Chin gives out under first clean shot: Borjas's flagged damage history means one well-placed Lima counter could end the firefight earlier than the decision lean suggests. Cardio inversion: If Borjas paces himself unexpectedly well, his heavier hands could steal late rounds Lima's volume path needs.
How does Blueprint MMA make this prediction?
Blueprint MMA validates and freezes the fight data, then runs 68 Kimi K2.6 agent executions: 33 Winner, 10 Method, 20 Timing, and 5 non-voting Oversight. Kimi K2.6 is the only inference model; LangGraph only orchestrates stage order and state. Winner agents fix the point probability first, Method and Timing agents fill conditional branches without reopening it, and Oversight audits without voting. The public probability is held to a sane 5-95% range and logged before release. Published winner probabilities are AI-only. Market prices may be shown for benchmarking, CLV, and workflow context, but they do not select, rank, or alter the published forecast. If the direct pipeline cannot return a valid read, the forecast is held. Methodology is documented at https://blueprintmma.com/manifesto and accuracy is tracked publicly at https://blueprintmma.com/track-record.