The benefits of evaluating driving in golf with strokes gained: off the tee are obvious — it’s the most stable metric in all of golf, making it one of the most predictive, as well. However, it’s intuitive that some courses benefit bombers while others benefit precision and accuracy. The problem is, until now, we haven’t had a way to reliably measure accuracy off the tee.

 

The PGA Tour has long used the percentage of fairways hit as their “driving accuracy” measure. It doesn’t take much math to know why this is bad — shorter hitters have an easier time hitting fairways since their margin for error, measured by degrees off-line, is larger. Like many statistics in sports, the effect the statistic is meant to capture is not the one primarily being captured. In other words, the real effect being captured by the percentage of fairways hit is an inverse correlation between the percentage and driving distance. In fact, this inverse correlation is so strong that the percentage of fairways hit is even inversely correlated with SGOTT.

However, the issue doesn’t stop there. Even if you account for driving distance, the percentage of fairways hit is an unreliable measure of accuracy because it treats all missed fairways equally. Take the sixth hole at PGA National as an example. There’s water down the entire left side and deep bunkers down the right. Some players take less than driver to maximize their chance of hitting the fairway, even though it leaves them a long iron into a difficult green and puts the first bunker in play. Here’s what Gary Woodland did in his third round this year:

A screenshot of a computer
Description automatically generated with medium confidence

Woodland took driver, effectively eliminating the first bunker, and aimed well right. This shot was surely right of his intended line, but his intended line was surely well right of the center of the fairway. A perfectly accurate shot here would have flirted with the right edge of the fairway, potentially being recorded as a missed fairway, just like his actual shot, which resulted in birdie. Numerous players sniped it left off this sixth tee, losing more than a stroke to the field, yet you wouldn’t be able to distinguish between their miss and Gary’s. This type of nuance prevents us from directly measuring driving accuracy in a reliable way, so we must do it indirectly.

 

Here’s how we do it:

  • Step 1: Isolate the distance component of SGOTT, creating strokes gained: driving distance
  • Step 2: Subtract SGDD from SGOTT, resulting in strokes gained: driving accuracy 

The isolation of the distance component is actually quite simple. Using data from the last three seasons, I regressed SGOTT on driving distance to create a model that reliably estimates a player’s SGDD based off their observed driving distance. Let’s compare Collin Morikawa and Matthew Wolff in 2020:

Given their similar SGOTT and vastly different distances, it’s obvious that Morikawa was more accurate. But what about when the SGOTT isn’t so similar?

Suppose we know a course takes away distance advantages, solely rewarding accuracy off the tee. Which player would have been the better option in 2021 (disregarding the rest of their games)? Try to guess before scrolling lower for the answer.

Hard to tell, right? Not anymore:

Did you guess correctly? If not, don’t feel bad — I bet the majority of the industry would have said Berger was more accurate off the tee than DeChambeau. In my opinion, the most incredible part of DeChambeau’s driving distance explosion has been his ability to remain hyper-accurate. Now we have the data to back that up.

This is groundbreaking stuff, and the beauty of it is, we at FTN are the only ones who have it. Not only are we now the most equipped to identify who the most accurate players are off the tee, but we also have the benefit of the course fit model telling us whether distance or accuracy is most rewarded at each course. Gone are the days of speculation — we’ll deal exclusively in fact and data.