Re: Understanding variation in variance

Interesting post-- enjoyed reading.

My intuition has always been that the courses with higher variation in their scores tend to be the most challenging-- those that really separate not just good shots from bad shots, but also great from good and good from ok, ok from iffy, iffy from bad, etc.

Not only do I think such rounds should not be standardized, but I would push for them to receive even more weight than they would in the standard methodology. When players compete on a course that really tests every skill, rewards every good shot, and punishes every bad one, it is an opportunity to get real insight into who is actually really good!

It perhaps offers an explanation for how Morikawa floats to the top of the leaderboard when the setup demands real precision tee to green, but I am probably reaching here.

I think you bring up an interesting point in the case that the source of the variance is primarily penalty strokes-- this definitely introduces a challenge. My thought would probably be to still not normalize, given that part of the sport is knowing that it is bad to hit the ball in the water and trying hard not to do it, but I could easily be convinced that something like a half standardization is best in the case of lots of OB or something.

Again, thanks for writing that post it was a really good read.

1 Like

I moved this to the General Discussion topic class if that’s ok.

I kind of agree with you, I go back and forth on this. The problem is you don’t really see it in the data. For example, if a course is separating good shots from bad more effectively, then that should show up in our course fit tool as a course that emphasizes all skills more than average. Now there are limitations to that methodology (it’s hard to pick up anything to do with putting / arg play due to noise it seems), so that’s part of it, but I still think it’s an important counterpoint to the claim that some of these high-variance courses are really good at separating good from marginal shots.