Golf Gets into the Swing of Analytics

Golf coach Pat Goss helped Luke Donald become the No. 1 golf player in the world from May 2011 to March 2012 (and on and off for a few months thereafter). And though Donald has slipped in the standings in recent years, Goss is working with his longtime protégé to get him back on top. Through it all, he’s taken advantage of data analytics to refine Donald’s game.

“As Donald turned professional, we really took advantage of analytics, understanding where he had room to improve and designing practices to improve specific statistics,” says Goss, who is director of golf and player development at Northwestern University, where he first coached Donald. “It was a big part of his development as he ascended the world rankings, and when he struggled, it’s helped us understand why he’s struggled and understand where he could improve.”

The PGA Tour’s ShotLink system, powered by CDW, has tracked and recorded every golf shot in real time since 2001, allowing golfers to embrace the data revolution that has transformed baseball, basketball, football and other sports.

Steve Evans, the PGA Tour’s senior vice president of information systems, says golf analytics is still in its infancy, but the golf industry has made great strides in recent years as ShotLink has paved the way for more accurate and meaningful statistics to measure player performance.

For example, strokes gained-putting, developed by Columbia Business School professor Mark Broadie and introduced in 2011, measures overall putting performance. It calculates the number of putts a player takes to reach a hole and compares it with that of his opponents while taking putting distance into account.

In recent years an increasing number of golfers and their coaches have hired statisticians to analyze the deluge of data to discover their strengths and weaknesses in order to improve their performance and strategy. For fans, the data provide better insight into player performance as they watch tournaments. And for the PGA Tour, this information helps architects redesign golf courses, Evans says.

Photo Credit: Stuart Franklin/Sam Greenwood/Getty Images Sport

ShotLink Data Provides Useful Direction

Broadie, author of the golf analytics book Every Shot Counts, studied the ShotLink data of PGA Tour professionals over an 11-year period and discovered that the scoring advantage of the best PGA Tour players comes more from the long game than the short one.

In his analysis, approach shots accounted for 40 percent of the players’ scoring advantage, and driving the golf ball was responsible for 28 percent. The short game accounted for 17 percent of their scoring advantage, and putting covered the remaining 15 percent. The analysis debunks the long-held golf myth of “drive for show and putt for dough.”

Broadie, who provides statistical analysis for numerous golf coaches and PGA Tour players, uses raw ShotLink data to crunch the numbers.

“It’s a diagnostic tool to pinpoint players’ strengths and weaknesses,” he says. “If you figure out that you have a particular weakness in medium-range putts or nine-iron shots from the rough, the data will give you that insight.”

After reviewing the analysis reports, coaches step in and help their players change their swing, strategy or practice routine.

Goss is one of the first coaches to embrace advanced golf analytics. Using ShotLink data, for example, he can evaluate bunker shots and see how well his players get out of the bunker. And if they are subpar, he can have them focus on these shots during practice.

Goss turned to Broadie to analyze Donald’s game using his strokes-gained formula. The statistics confirmed much of what Goss already knew: that while Donald didn’t drive the ball super far, he was very good at his short game and putting. But the analysis also provided an extra level of detail that Goss was able to use to pinpoint areas where he could help Donald improve his play.

“Through Broadie’s stats, we were able to understand Luke’s formula for success,” he says. “The year he was number one, his strokes gained was around a positive 2.0. Luke had to get almost a full shot in putting and the rest from the short game inside 100 yards and from iron play, and just break even with driving. That really is his formula, and that’s how he can be the best player in the world. It really allowed us to identify Luke’s strengths and prepare for those strengths to try to maximize them.”

Goss created drills to help Donald further improve his putting numbers. Whereas the average Tour player makes about 69 percent of their 4- to 8-foot putts, Donald in his best years made 75 percent, he says.

“We constantly created drills where we measure that and try to perform above his set level,” he says.

Analytics, through the use of historical data, also help golfers make strategic risk-reward decisions, Broadie says. So if the hole is on the right side of the green right next to water, should you aim for the hole, which is more risky and provides more of a reward? Or should you play it conservatively and aim to the left side of the green?

The right answer, according to Broadie, depends on the individual golfer’s skill level and shot patterns: If the player knows his last 20 to 50 shots from ShotLink data, then he knows what his tendencies are – and that can help him decide his strategy.

“If you know this information, you can do more careful analysis on how much risk you should take,” he says.

Data Helps Redesign Golf Courses

PGA Tour pros regularly give feedback on what they like or don’t like on any given golf course. If players complain that a fairway bunker on a particular hole is unfair, PGA Tour officials use ShotLink data to help analyze shots and determine whether it is unfair, Evans says.

“Whatever the complaints are, you can actually look at all the shots hit into the bunker and around the bunker, as opposed to listening to a limited amount of feedback that’s not super objective,” he says. “The ShotLink data provides an objective view. Maybe the bunker needs to be moved or maybe the lip is too high, but you can use objective data to draw that conclusion.”

Architects also work with the PGA Tour’s in-house design experts to redesign golf courses, such as changing the layouts of greens or placing bunkers in new areas. They look at the ShotLink data to determine if they should make a hole harder or provide more risk-reward, Evans says.

More Analytics to Come

Broadie has worked with the PGA Tour to extend his strokes-gained formula to include strokes gained-tee to green, which measures player performance before they reach the putting surface.

In the near future, they will break up the tee to green statistic into its three smaller components: drives off the tee, approach shots and around-the-green shots.

“Strokes gained has become the most visible byproduct of analytics,” Evans says. “It provides a benchmark of how you’ve performed against your peers. It’s a very effective way of ranking players.”

Overall, while golf professionals and statisticians have made inroads with analytics, they are still just scratching the surface of what’s possible, Evans adds.

“We have information on the type of grass on courses, the speeds of the greens, and the weather conditions they are playing in,” he says. “We have not done a lot of work to date correlating those data components with scoring and performance, so there are still a lot of opportunities to get more value from the data we capture.”