← Back to Track Record

Fight Night: Lebosnoyani vs Ko

Sports Β· Kalshi Β· Called NO, verified LOSS Β· Jul 18, 2026
View the original Kalshi market →
Verified Outcome
LOSS
PillarLab called Jean-Paul Lebosnoyani will NOT win (NO @ 38%).
Result: Jean-Paul Lebosnoyani won. The call was wrong.
Analyzed: Jul 18, 2026, 7:41 AM UTC
Market resolved: Jul 18, 2026, 11:46 PM UTC
Called 16h before resolution. The call was made before the outcome was known.

https://kalshi.com/markets/kxufcfight/ufc-fight/KXUFCFIGHT-26JUL18LEBSEO?utm_source=kalshiapp_eventpage&referral=143c1f38-859e-4cca-880d-fe56e65e5945

PillarLab
πŸ“‘ Fetching Kalshi data...
PillarLab
βœ… Got data from Kalshi API!0.8s
PillarLab
πŸ” Step 1: Detecting market type and category...
PillarLab
βœ… Category: ⚽ SPORTS βœ… Type: BINARY πŸ“Š Market: Fight Night: Lebosnoyani vs Ko πŸ“… Date: 2026-07-19T03:20:00.000Z πŸ’° Volume: $105,478 Current Odds: β€’ Seok Hyun Ko: 64% β€’ Jean-Paul Lebosnoyani: 38%2.7s
PillarLab
⏳ Step 2: Gathering intelligence (parallel)...
PillarLab
βœ… Intelligence gathered: Vegas Lines Β· Injury Report Β· Team Stats Β· Reddit Sentiment Β· Situational Factors Β· Line Movement Β· Handle Split (Public vs Sharp) Β· Candidate Dossiers (case for + case against, every candidate β‰₯2%)20.9s
PillarLab
🧠 Step 3: Deep Analysing across all 9 pillars β€” hang tight, ~20s...
PillarLab

πŸ“Š Market Summary

⚽ Fight Night: Lebosnoyani vs Ko

Team/PlayerCurrent Odds
Seok Hyun Ko64%
Jean-Paul Lebosnoyani38%

2 options total

PillarLab

🎯 Best Trade

  • Action: Jean-Paul Lebosnoyani β€” NO @ 38%

  • Confidence: 🟑 MEDIUM Β· Uncertainty: MEDIUM

  • Why:

  • Lebosnoyani's UFC sample is limited to a single split decision over Phil Rowe, the same opponent Ko beat more decisively.
  • Lebosnoyani faces a quick turnaround of ~5 months from his Feb. 2026 debut, which may affect preparation.
  • Ko's wrestling threat could neutralize Lebosnoyani's striking advantage, as seen in Ko's takedown-heavy game plan.
  • The line has already moved toward Lebosnoyani from +210 to +180, but he remains the underdog.
PillarLab

πŸ“ˆ Where the Market Stands

OptionMarketRead
Seok Hyun Ko64%🟒 FAVORED
Jean-Paul Lebosnoyani38%πŸ”΄ TRAILING

I price with the market β€” 🟒 favored Β· πŸ”΄ trailing. The analysis below is the why.

PillarLab

πŸ”¬ 9-Pillar Analysis (Sharp Trading Method)

πŸ“Š 1. Vegas Line

My Vegas Line has Ko at approximately 69% implied, Lebosnoyani at 31%.

πŸ“ˆ 2. Line Movement

Line movement slightly toward Lebosnoyani, with Ko's odds shortening by 25 cents and Lebosnoyani's improving by 30 cents from open.

πŸ₯ 3. Injury Report

No reported injuries or weight issues for either fighter.

πŸ“± 4. Reddit Sentiment

Reddit sentiment leans toward Ko, but a minority sees Lebosnoyani as underrated.

πŸ”’ 5. Team Stats & Form

Ko is on a 6-fight win streak (13-2 overall, 2-0 UFC) with decision wins over Elliott and Rowe; Lebosnoyani is 10-2 overall, 1-0 UFC with a split decision over Rowe.

🌀️ 6. Situational Factors

Fight is in Oklahoma City with no home-venue edge; Lebosnoyani faces a quick turnaround from his Feb. 2026 fight.

πŸ’° 7. Handle Split (Public vs Sharp)

Public handle split data is not published for this fight, so no read available.

⏰ 8. Time Decay

MEDIUM - The market has already absorbed the line move toward Lebosnoyani, but no major injury news is pending; price appears stable.

⚑ 9. Contrarian Check

The market is slightly fading Lebosnoyani despite his sharp striking in his UFC debut; if he can keep the fight standing, he has live dog potential.

PillarLab

⚠️ Disclaimer: AI can make mistakes. This is not financial advice. Do your own research before placing any trades.

Stop guessing. See the edge.

Paste any Kalshi or Polymarket market. PillarLab runs a full 9-pillar analysis and hands you a Best Trade call in about 30 seconds.

Free to start Β· 10 credits Β· no card

PillarLab AI is an AI prediction-market analyzer that runs a 9-pillar analysis on Kalshi and Polymarket markets and publishes a public, settlement-verified track record of its decisive calls.