View Single Post
  #127  
Old 06-12-2019, 01:12 AM
Steve West Steve West is offline
Par Delusionary
 
Join Date: Dec 2009
Years Playing: 45.5
Courses Played: 365
Posts: 4,990
Niced 1,694 Times in 833 Posts
Default

Quote:
Originally Posted by armiller View Post
I wanna say West separated a bit better. I imagine it would be easy to compare courses for these NT or DGPT events. Have you done that? I thought Eagle's comment was interesting after Rd 2, saying he knew exactly what shot to throw for all West holes, not seeming as sure about East.

I was also thinking, isn't "higher" more important than "right?" Correlation with rating seems more valuable to me than just giving scoring separation.
I can compare these two courses because the same players played both courses at essentially the same time for the same number of rounds in (I think) the same weather conditions. If any of those were not true, comparisons would get very tricky.

West had better scoring spread among these 970+ rated players at 19.57 vs. East’s 17.63. However, East had better correlation with ratings at -63% vs. West’s -58%.

I don’t have a final answer on which stat is more important. I don’t think of this as a way to crown the best holes. I’m just looking for the holes which may be safe to tamper with. Perhaps I should say “down and left is more in need of closer examination”.

For the purpose of targeting holes for potential improvement, a narrow scoring spread is more obviously in need of a fix. A hole isn’t doing its job if it gives everyone the same score.

A poor correlation could mean the hole relies too much on luck, but it also could mean that the hole is testing a skill which is not commonly tested (and therefore not reflected in the player ratings). That kind of hole can be quite valuable.

Both scoring spread and correlation can be overdone. A hole that gave out scores according a roll of a single die would generate a scoring spread of about 5.8 (less than 6 because it doesn’t give out exactly the same proportions of all six numbers). But, it would not tell us anything about who is playing well that day.

A hole that gave out scores only according to player ratings would have a correlation of about -89%, but if all holes did that, there would be a lot of ties. A bunch of players at the middle of the pack would all be tied at 54, for example. That would also not tell us anything about who in that pack is playing well that day.

Niced: (2)
Reply With Quote