In the last article we went over some of the systematic biases in Vegas odds-based pricing and showed that upside is a major component of each player’s DraftKings price. It follows that a good cash game strategy would be to identify golfers who may not have upside but have good average outcomes. Such golfers could serve as the bedrock of value in PGA lineups.
We’ll go through some of the key PGA stats I’ve identified to find such golfers. Unfortunately for regular readers and users of the PGA Player Models, these stats aren’t much different from what we usually emphasize week to week. Nevertheless, it’s a useful illustration of why we’ve chosen the values we did for the default models.
Long-Term Adjusted Round Scores
If you have to pick one metric to identify golfers who perform well on average independent of pricing, long-term adjusted round is your best bet. Here’s how two golfers can have the same average expected scoring but different upside and therefore different pricing (and note the similarities that this graph shares with the distribution graph from last week):
Upside is defined as the right-tailed region of that graph, so the orange golfer may have a better chance to win (and therefore a higher price) but will perform the same on average as the blue golfer. The nice thing about average score is that it’s an easy calculation and is completely independent of upside, so once scores are properly adjusted and normalized, simple averages of strokes per round are a great way to identify average expectations (and, by extension, value).
Missed Cut Percentages
One of the more interesting questions in golf is whether or not cut-making is a distinct skill from being generally good at golf. In other words, if two golfers have the same average adjusted strokes per round, is it possible that one will be predictably better than the other at making the cut from week to week? Based on the research I’ve done to date, I’d say yes: There is a distinct skill in cut making, which is why it deserves its own weight.
In terms of adjustments, making the cut is worth roughly one stroke per round above not making the cut, all things being equal. But be careful when weighting this statistic: There is a lot of overlap with average strokes per round as outlined above, so if you use both of them, don’t overweight missed cut percentages.
Course Fit Statistics
Accounting for course fit is not straightforward, since it is frequently priced into golfers’ average expectations as well as upside, but it’s still a helpful way to adjust golfers who might otherwise rate well using the statistics above. Course fit will have its own set of articles going forward, but it’s important to apply course fit stats the right way. The most common way people figure out course fit is to look at the top-10 leaderboards each year, figure out the most common stats, and call those the “key statistics.”
That approach has a host of problems: The overlap with course history (distinct from course fit), failure to consider if there were also underperforming players with those stats, and the refusal to consider otherwise low- to mid-tier golfers who didn’t crack the top but still overperformed. All the same, using the key course statistics we provide in Player Models can help you adjust the players found using the first two stats closer to their “true” averages.