In my previous two articles, I went over an exploration of correlated player stats and a way to isolate independent descriptive statistics for golf playing styles. I wanted to wrap up this series talking about the applications of doing so, and what I think the next-level analysis of golf looks like.
Dealing With Luck
Last we left it, we had these independent variables we created out of traditional stats, but we don’t really have a good bearing on what they are or what they describe. Based on concepts like explained variance ratio, we have some intuition that some of these numbers are descriptive of skill, and others aren’t. We can figure out which ones are by seeing if they have predictive value of things like DraftKings points, birdies, odds to make the cut, etc. And if a metric doesn’t have predictive power, it’s most likely a rough-cut estimate of luck.
If we can isolate and quantify luck, we can use that and adjust raw numbers we get–like scoring average–before it hits adjustment methods like Long-Term Adjust Round Score. That leads us to some future metrics with really long names, like “Luck-Independent Long-Term Adjusted Round Score.” The FantasyLabs’ Acronym division is hard at work for figuring out ways to condense that. (Editor’s Note: LILTARS is a dope acronym.)
The Future of Golf Analysis
This, in a nutshell, is I think the biggest breakthrough waiting to happen in golf analysis: Figuring out what is and isn’t luck, and using it to adjust our understanding of results. Baseball’s first stats breakthrough was introducing the concept of BABIP and subsequently Fielding Independent Pitching (FIP) to remove luck from our understanding of pitcher performance.
Football had a similar breakthrough with ideas like fumble recovery rate and understanding how much luck there is in turnovers. Golf has had no such breakthrough yet, even though complex interactions are fewer, similar to baseball. The methodology in the last couple articles is but one method to try and get there; I’d love to see more people take a crack at their own concepts of luck in golf and how to quantify it. In the meantime, we’ll be busy here trying to incorporate these findings into our PGA product and at least capitalize on giving our customers an edge with quantifying luck.