I would like to hear as to why the people who run DataGolf had Russell Henley rated higher than Rory going into the tournament when Henley is nowhere close as good of a player as Rory is especially in his current form. I get that Rory isn’t the absolute best course fit for Sawgrass and Henley was a better course fit but Rory being around +2900 to win The Players pre-tournament vs the books all being around +800 or so should rise some suspicion in regard to model error/bias imo. Do you think you guys overweighed the driving accuracy for Rory here? Rory is currently playing like the best player in the world right now especially with Scottie and Xander both not in form plus he is in the best form he has been in years. Was there not as much of an adjustment to his skill rating to highlight the current form he is in? Because the major difference in DG’s numbers on Rory vs the market skewed all the fair value simulations for all nationalities/groups that Rory was in and as well as live post round outright winners. Id just like to hear as to why you guys were so low on Rory going into this tournament. Thanks.
I think the player skill decomposition page gives a good breakdown.
On baseline skill, Rory was about 0.4 strokes better than Henley. But Henley’s course fit adjustment was 0.6 strokes better than Rory.
Rory has been a very reliable +1 SG:OTT guy (or a bit better) for a long time. At Sawgrass, it looks like he’s historically averaged +0.7. This week he’s gaining 0.4/round at the moment (which will be adjusted up slightly with field strength). So that alone kind of explains his course fit bump (-0.35).
This week is basically treated as a major from the books, and it seems like they were pricing Rory and Scottie the same way they would at any of the majors (which are typically courses that favour their skillset more). We were also very low on Scottie all week, which has ended up being an okay prediction.
But I agree we were low on Rory all week, wasn’t something I was too comfortable with. But hopefully users weren’t blindly simulating things and unknowingly making a lot of correlated bets against Rory.
edit: should add that I don’t think Rory was particularly close to Scheffler’s level coming into the tournament. He does have the win this year, but from an SG standpoint they’ve been pretty similar so far in 2025. And then obviously Scottie is coming off an historic season, so you had to give him a sizeable edge in terms of baseline form still before this week.
He can still lose to Spaun tomorrow DM don’t capitulate now!
On a (very slightly) more serious note even if you disagree with the model pricing, +800 is at least as wrong and sharper places were quite a bit longer than that.
I was 1000% agree that Rory being priced at +800 is a major rip off and its the books egregiously shading players like Scottie/Rory etc because they know all the casual bettors will betting the favorites. Very similar to how Fanduel would price touchdowns for players like Jamar Chase, Justin Jefferson etc.
My main gripe was just pricing Henley more likely to win over Rory. Despite Henley benefiting from good course fit and Rory not benefiting from course fit - I don’t personally think that he should ever be more likely to win a tournament over Rory especially when Rory is playing the best golf he has played in years.
Rory 2022 and 2023 was at least as good as him now, no? Wasn’t exactly imperious in Genesis or API.
Henley is a top 7 player in the world right now. It’s fine to wonder if he can do it under the brightest lights but he’s been unbelievable for a while…
Ya, I’m definitely sympathetic to that. I remember way back, 5-6 years ago, before our model accounted for much course-specific stuff, laughing at someone who made Harman a favourite over DJ at Harbour Town (back when DJ was good). And now we are making Henley a favourite over Rory at Sawgrass, which feels similar.
I think one thing to keep in mind is that our model is primarily meant to be a betting model, and not a bookmaking model. On weeks like this, as long as we are directionally correct with Henley and Rory, there can be good bets to make. But if you were to use our model’s prices (+ juice) for people to bet into, you could definitely run into trouble pricing Rory where we did.
It certainly seems that when there are heavy positive adjustments for shorter accurate hitters (and naturally negative adjustments for the longer more erratic players) that there is more green on the board (this week and players week)
I’m not too sure how sharp or not sharp books are making odds for outrights, but it would seem natural to be skeptical on how real the edges are if 20+ edges are popping up.
While I think there was certainly reason to downgrade and upgrade based on accuracy, I wondered myself if the distance adjustment was too much. I was more comfortable with what the sim result gave me when I put the distance adjustment to ~.7 (about average) in the custom sim, but I do most of my outright betting before the release anyway.
I would also just note that guys who gain more strokes with accuracy v. distance performed quite well last week. 11/19 were guys who are clearly more accurate than long. Only 4 are clearly more long than accurate type guys (knapp, cantlay, walker, rory).
Just the guy who won didn’t fit the profile. SSS, 3% things happen fairly often, Rory was 6th most likely to win… pick your explanation, bottom line the model was very predictive last week and if folks are expecting it to do better than that, good luck!
Sure seems like accuracy skill was probably a better predictor of success this week than distance skill, which is what we would expect here