Arnold Palmer Invitational 2026: Model Performance Recap
What Happened
Akshay Bhatia won his third PGA Tour title at Bay Hill, defeating Daniel Berger in a one-hole playoff at 15-under. Berger had led wire-to-wire entering the final round and looked destined for his fifth Tour title before Bhatia birdied 15 and 16 late on Sunday to force extra holes. Berger, making his return from multiple injuries, showed his game is back despite the heartbreaking finish.
The week's most talked-about moment was Scottie Scheffler losing his tee shot in a palm tree in the bunker at the par-3 14th on Sunday, forcing a re-hit and a double bogey that ended his title chances. Rory McIlroy withdrew before his third round. Ludvig Aberg and Cameron Young finished T3 at -12.
Why We Did Not Bet Bhatia to Win
Bhatia won at +6500, implying a market probability of 1.52%. Our model had him at 0.9% raw win probability, which calibrates to 0.63% after applying our standard 0.7x win market calibration factor. To justify a bet, we needed odds of approximately +15700 or greater. The market was actually overrating Bhatia relative to our model—no edge, no bet. His winning does not make that the wrong decision. That is how positive expected value betting works.
Outright Model Performance
The Arnold Palmer was a challenging week on the outright card, finishing 7-27-0 for -$21.14 (essentially breakeven at -0.11u). This result masked significant variation between markets, with Top 10 delivering exceptional value while other markets struggled.
Top 10 was the standout performer, going 2-5 for +$393.59 (+1.97u). Sahith Theegala at +550 returned $376 and Rickie Fowler at +220 returned $169—both representing genuine model edge at a signature PGA Tour event. This strong week brought the Top 10 market's season total to slightly positive (+0.11u) after struggling earlier in the year.
Top 5 went 0-7 for -$131.63 and Top 20 went 5-13 for -$94.00. Bay Hill is one of the tougher courses to model for finishing positions given the difficulty variance between rounds—a calibration challenge the model will address as the historical dataset grows.
Make Cut had a rough outing at 0-1 for -$188.35, while FRL went 0-1 for -$0.75.
Best Bet of the Week
Top 10: Sahith Theegala +550 - Returned +$376 on a $68.36 stake. The model identified Theegala at strong value entering the week, and he delivered with a T6 finish. This single bet nearly covered the entire week's losses across all other markets.
P&L Summary
Top 5: 0-7 / -$131.63 / -0.66u
Top 10: 2-5 / +$393.59 / +1.97u
Top 20: 5-13 / -$94.00 / -0.47u
FRL: 0-1 / -$0.75 / -0.00u
Make Cut: 0-1 / -$188.35 / -0.94u
Overall: 7-27 / -$21.14 / -0.11u
Takeaway
A near-breakeven week at one of the toughest courses on the PGA Tour schedule. The Top 10 market delivered elite value identification this week, turning what could have been a -$415 loss into an essentially neutral result. The Theegala and Fowler hits demonstrate the model's ability to find finishing position edges at high-profile events where market pricing is most scrutinized.
The week reinforced the model's core principle: no edge, no bet, regardless of who ultimately wins. Bay Hill's difficulty and scoring variance present ongoing calibration challenges for finishing position markets, particularly Top 5 and Top 20, which will continue to be refined as more tournament data becomes available.
---
Want to see this week's +EV bets? View current opportunities →