BOSTON – The recently concluded MIT Sloan Sports Analytics Conference brought together some of the most cutting-edge researchers and brightest minds in the world of sports statistics. Former Microsoft CEO and current Los Angeles Clippers owner Steve Ballmer spoke, as did former NBA stars Shane Battier and Steve Nash, former baseball star Alex Rodriguez, former president and general manager of the Philadelphia 76ers Sam Hinkie and former President of the United States Barack Obama.
Panel discussions and lectures included, “Leveraging Machine Learning and Pitch F/X Data to Illuminate the Impact of Pitch Framing,” and “Real-time Interactive Play Sketching with Synthesized NBA Defenses.”
Golf was there, too, in the form of Peter Sanders, the founder of ShotByShot.com and the stats guru to numerous PGA Tour players, including two-time major winner Zach Johnson. His lecture, “From Stats to Strokes Gained: The Evolution of Golf Analytics,” was given in front of about 150 people and focused on why some of the most popular stats in the history of golf are, basically, worthless.
But thanks to the PGA Tour’s ShotLink system, which began in 2003, strokes gained statistics have given us a better understanding of the game.
Here are a few insights Sanders shared:
- Every week, the player who wins makes an inordinate number of putts from 11 to 20 feet.
- About 33 percent of all first putts are 20 feet and longer, and when faced with a putt from more than 20 feet, the best putters in the world leave themselves within 7 percent of the first putt’s length, or 3 feet, more often than other players. Because players on the PGA Tour make putts from inside 3 feet about 99 percent of the time, the best putters rarely three-putt.
- Everyone on the PGA Tour plays under par when they play from the fairway. “If you look at the scores after an average Tour player, top-five players and even the guys who get cut, when they hit the fairway, they all average under par on those holes,” Sanders said, citing the data in the table below. “From the rough, they don’t, except for the players in the top five. The average difference between playing from the fairway and hitting the rough is about a quarter of a shot.”
| Players | Avg. Fairway Score Relative to Par | Avg. Fairway Score Relative from Rough | Difference |
| Tour | -0.16 | 0.12 | 0.28 |
| Top 5 | -0.25 | -0.01 | 0.25 |
| Cut | -0.05 | 0.23 | 0.28 |
Sanders said the most frequent approach-shot distance range is 150-175 yards. From that distance in the fairway, the average Tour player hits the green about 71 percent of the time. But when playing from the rough, to achieve that level of consistency he would have to be 50 to 75 yards closer to the hole (in the 76- to 100-yard range).
“The message I give to the players I work with is that if you have to lay back 25, 30 or 35 yards to avoid the rough or trouble, you are not giving up any shots,” Sanders said. “You are gaining shots.”
However, when golfers do not have water or a hazard guarding a green on a par 5, they are almost always better off going for the green instead of laying up. The reason, according to Sanders, is in the chart below.
| 75-100 yards | Percent Who Hit the Green | Average Distance to the Hole |
| Tour | 85 percent | 15 feet |
| <50 yards | Percent Who Hit the Green | Average Distance to the Hole |
| Tour | 93 percent | 7.1 feet |
Players who go for the green in two shots on a par 5 and hit their third shot from within 50 yards of the hole hit the green with their third more often and their typical shot finishes closer to the hole too.
“They cut their error percentage in half, and they cut their average putting distance in half,” Sanders said. “So my message is that if there is no obvious downside, go for the green every time.”
Surprisingly, Sanders said that most coaches who work with PGA Tour players see the hundreds of stats created in the ShotLink system as a giant, confusing mess. But even more surprising, four minutes into Sanders talk, he said, “If you know what strokes gained is, then you are in the vast minority. I start our all my initial meetings with Tour players and their coaches by giving examples, and it invariably leads to very spirited conversations about what these numbers really mean and how they relate to traditional statistics.”
To ensure that you are in that wise minority, it is important to understand that stokes gained statistics are all based on the average number of shots it will take to get the ball in the hole from a given distance. Those average numbers have been created by studying tens of thousands of shots measured using ShotLink. Here is a simple example of how strokes gained stats work:
Imagine that a PGA Tour golfer is playing a 460-yard par 4. According to professor Mark Broadie’s book, “Every Shot Counts,” historical data shows that the average number of strokes needed to get the ball in the hole from that distance is about 4.17. After the player hits a tee shot 300 yards, and his ball rests in the fairway 160 yards from the hole, the data says that the average number of strokes needed to get the ball in the hole is now 2.98. If you take 4.17 and subtract 2.98 (which equals 1.19), and subtract one to account for the shot that was taken, we know the value of that 300-yard drive was 0.19 strokes.
If the player hits his next shot on the green, but 60 feet from the flag, the data says the average number of strokes needed to get the ball in the hole will now 2.21. So, 2.98 minus 2.21, minus one for the shot the player just took means the approach shot was -0.23. In other words, it was not a good shot. However, had that second shot stopped 10 feet from the hole, where the average number of shots needed to get the ball in the hole is 1.61, the approach shot would have been worth 0.37 strokes (2.98-1.61-1 equals 0.37).
If the player makes the 10-foot putt, the value of the putt would be 1.61 minus one, or 0.61 shots. If you add up the strokes gained value of the drive, the approach shot to 10 feet and the putt, you will see that this player gained 1.17 shots.
Having attended the MIT Sloan Sports Analytics Conference five times before this year’s event, I was not surprised that golf had a relatively small voice. Last year I moderated the only panel discussion on golf and the year before there was just one presentation on golf delivered. Basketball, baseball, football and soccer have adopted analytics more aggressively. Still, golf is not going away. Sanders is one of many stat gurus who work with Tour pros, elite amateurs and motivated recreational golfers. The ShotLink database is getting richer ever week and shot-tracking systems for recreational golfers are coming down in price and increasing in utility. Networks are working analytics and data into the storylines of television broadcasts every weekend too.
Golf is an ancient game, but when it comes to stats, there is no denying that it is moving into the modern world. Gwk
Be the first to comment