By Austin Tymins
The picturesque Whistling Straits, a course not unfamiliar with hosting major tournaments, will host this week’s PGA Championship. And like nearly all professional events, some skill sets will be more suitable than others to the Whistling Straits layout. For example, as I described in my previous post, long hitters typically play very well in major championships (though this effect was weakest for the PGA Championship). In this post, I set out to determine which tournaments and courses present the biggest advantage for good ball-strikers, putters, and players over the field. To do this, I looked at the last five years of PGA and major tournament results and the current season’s strokes gained statistics at year-end.
The strokes gained statistics were invented by Professor Mark Broadie of Columbia University utilizing ShotLink data to determine the individual’s relative ability in different facets of the game. The strokes gained method is most easily described using a simple example. Suppose a golfer has an eight-foot putt and the tour average number of strokes to hole out from that distance is 1.5. Therefore, a one-putt gains 0.5 strokes, but a two-putt loses 0.5 strokes on the field. These fractions are aggregated from every green, and the per round averages are published by the PGA tour. Strokes gained tee to green, equivalently, is the measure of how efficiently a player is getting the ball to a spot on the green versus the field.
My analysis is done for 43 different tournaments and produced nearly 30,000 observations. I then regressed the individual player’s average round score at each tournament on their strokes gained averages from the applicable season. This was done so as to include the variation presented by cut players and to avoid potential selection bias.
Before I get to the results, a couple quick notes. I presented values for the Quicken Loans/AT&T National, BMW Championship, Sanderson Farms Championship, The Barclays, the Frys.com Open, and the three rotating majors even though these tournaments took place on more than a single course from 2010-2014. Additionally, the Farmers Insurance Open, AT&T Pebble Beach Pro-Am, and Humana Challenge are traditionally played on more than one course. For these reasons, it may be best to view the results for these tournaments with the proverbial grain of salt since one would expect the assumed advantage to be seen most clearly on repeated playing on a singular course layout.
Below are the results with the majors highlighted in yellow. The regression results and t-statistics for SG:T2G are shown in the right columns.
As we would obviously expect, all the coefficients are negative and statistically significant below 1%. The tournaments are listed in order of the largest coefficients on the SG:T2G term. This should be interpreted as the ranking of tournaments that present the greatest advantage to better than average ball-strikers. These are the courses that we would expect Hideki Matsuyama, Bubba Watson, Jordan Spieth, Sergio Garcia, and Adam Scott to play best at as they are the top 5 on tour this season in SG:T2G.
Next are the results from the same style regression using strokes gained putting as the primary regressor.
Once again all the terms are negative and the vast majority are statistically significant at the 1% level. The single asterisk denotes 5% significance, and the double asterisk means no significance in this context. Again, the tournaments are listed in order of the largest coefficients on the SG:P term. The top tournaments on this list are where the top putters have scored well over the data range. Watch out for Jimmy Walker, Aaron Baddeley, Freddie Jacobson, Brendon Todd, Brandt Snedeker, and Jordan Spieth (again) come Sunday at these events.
The last table shows the results using Total Strokes Gained as the independent variable. This table simply shows which tournaments are most advantageous to the best players over the course of the season. Strokes Gained Total is currently one of the best, if not the best, quantitative measure of player skill.
All the coefficients here are statistically significant below the 1% level as expected. The Puerto Rico Open at Trump International GC was the worst course over this period at separating the best from the worst players. Conversely, the Barracuda/Reno-Tahoe Open was the best at creating field separation (although t-stat reflects smaller sample size). There is an argument to be made that the best tournaments and courses effectively separate the best from the worst players; however, a subjective look at these rankings doesn’t necessarily confirm that notion.
Having seen the cross-sectional evidence, what can we say about the upcoming PGA Championship? Well firstly, it has been the best major at separating the good from bad players as measured by Strokes Gained Total over the last five years. Additionally, it is the fourth most advantageous tournament of all studied for solid ball-striking. Come Sunday on the shores of Lake Michigan this weekend, don’t be surprised to see top ball-strikers competing for the Wanamaker trophy, as occurred when Dustin Johnson, Bubba Watson, Rory McIlroy, and Martin Kaymer went toe-to-toe in 2010.