What do Daniel Berger, Tony Finau and Branden Grace have in common? I see them all in my crystal ball. When drawing up a list of the winless PGA Tour players who are most likely to break through with a victory in 2016, I could have formed an opinion based on swing aesthetics or my intuition, but I prefer a quantitative "predictive analytics" approach -- to use big data in the same way that, say, Netflix predicts your taste in movies.
There's still plenty of room for judgment—okay, hunches—in quantitative approaches. Numbers don't yield the truth any more than a dictionary teaches you how to write an essay. So here's my hunch: A simple but solid predictive model for identifying probable first-time winners involves two factors. The first is a Tour player's Strokes Gained in the previous season (that is, their scoring average in each round, adjusted for the difficulty of the course). Players who gain more strokes against the field during one season are likelier to win in the next. The second factor is the number of top 10s a player had in the previous season. This helps identify guys who can string together four good rounds. The more times a player knocks on the door, the more likely it is to open.
When I ran this model in October, at the end of the 2014-15 season, three players—Justin Thomas, Russell Knox and Kevin Kisner—were at or near the top of the list of Tour pros most likely to win their first event in the 2015-16 season. Mission already accomplished! In November, those three players all notched their maiden Tour victories. Who are the players still crowding my quantitative crystal ball? In order: Grace, Finau, Berger and Will Wilcox. The model also identifies Patton Kizzire, Hiroshi Iwata and Harold Varner III as Tour rookies to look out for.
Running the same model for one-time winners most likely to capture their second Tour win gives the edge to (again, in order) Hideki Matsuyama, Kevin Na, Justin Thomas, Paul Casey and Robert Streb.
Pro tip: If anyone starts a sentence with "The numbers say…," don't believe them. Making good predictions from raw data takes skill and some imagination—a bit like the ingredients that winners use to conquer a course.