GCA Mythbusters: Does Randomness Negate Skill?

[Jason Thurman]

Please Note: This is not another thread about the greens at Chambers Bay. It is instead a thread about randomness in architecture and maintenance that uses the greens at Chambers Bay to initiate its examination. Please do not make this a thread about the greens at Chambers Bay. I want instead to use this thread to examine how features that encourage randomness - undulation and contour and firm and fast conditions, etc. - ACTUALLY affect outcomes of golf tournaments.


Previous installment in this series: http://www.golfclubatlas.com/forum/index.php/topic,59158.0.html


Following the US Open at Chambers Bay, much was made of the bumpy greens with numerous golfers and media members suggesting that the conditions emphasized luck over skill.


This is the same argument that we've heard for years about links golf from advocates for lush and green - that the random bounces on links courses reward luck instead of skill. We've been told that if a course intends to "identify the best player," "good shots" must be rewarded with predictable results.


Still, I've always been compelled by looking at the long list of legendary ballstrikers who have also been winners in the Open Championship. I've always personally suspected that the ability to control ballflight in turn gives a player more control than his opponent over the outcome of shots on a fast and firm and unpredictable course. To simplify, I've always figured that great skill makes the unpredictable more predictable, in the same way that slippery roads may lead to more random skids, but ultimately also reveal the drivers who can best handle their vehicle while exposing the ones who are incompetent.


The controversy surrounding the greens at Chambers Bay seems to present an opportunity to explore the effects of randomness in design/maintenance on outcomes in golf. By consensus, the greens were bumpy and inconsistent. Also by consensus, the greens at Augusta National during the Masters roll as true as any greens in the world. With that in mind, I decided to do a quick look at the outcome of the season's first two majors, and examine whether good putters were significantly favored on one course versus the other.


"Methodology"


I made a spreadsheet and compared a player's finish against their rank in strokes gained putting. Then I plotted the points and calculated the correlation coefficient.


Result


So, Chambers Bay clearly didn't uniquely favor good putters. In fact, there was a slightly negative correlation coefficient between a player's finish and their rank in strokes gained putting. Meanwhile, good putters at the Masters were slightly more likely to finish higher on the leaderboard.





However, the correlation between strokes gained putting and final position for BOTH tournaments is so tiny and over such a small data sample that we can effectively say it's completely insignificant. Whether the greens are perfect or bad enough to make Billy Horschel take a dump in the hole, putting skill still wasn't of much importance to where a player finished in each tournament. At Chambers Bay, the top five featured the 158th, 121st, and 189th ranked putters on Tour. At Augusta, 8 players who finished T12th or better were ranked 120th or worse in putting.


All this raises the question for me of how other random features affect the importance of skill in a golf tournament's outcome. Do windy or firm conditions raise or lower the importance of good ballstriking? Does lush and unpredictable rough around greens increase or decrease the importance of short game skills? What about green contours? If smooth greens don't necessarily reward great putters, do flat greens reward them? Or does contour better emphasize putting skill?


Also, I put "methodology" in quotes for a reason. What data has my crappy spreadsheet study failed to take into account? I'm not smart enough to control for a player's world ranking or things like that. A real statistician may look at my conclusion and see that I've completely overlooked several factors and reached a totally incorrect conclusion as a result.


I continue to believe that a modicum of randomness doesn't negate skill, and may in fact emphasize it. What other data should be examined to determine whether this is true?

[Matthew Petersen]

Good data, but it seems like there are two totally different types of randomness being discussed here, in theory (and this goes to complainers, not saying jason is wrong to bring them up).


The "randomness" of a bumpy green is much closer to true randomness. Three puts from the same spot on a green may react in entirely different ways based on how they are rolling at a certain moment, what they hit, etc. That would seem to work against a good putter. Player A who hits a putt with good speed on the correct line could still miss the putt if it hits an odd clump of grass. Player B could hit a putt that would miss on a true green, but either find that his putt actually goes in after being redirected by a bump, or simply gains the advantage of the better putter's putt missing.


That to me is a totally different idea than the concept of randomness on links, humpy golf courses. When most players talk about that being random, they generally are just upset that what they thought was a good shot (i.e., hit directly at the hole) didn't finish near the hole due to firmness, contour, or some other reason.


As such, it makes sense that we see not only good ball strikers but good course strategists succeeding in such tournaments. Not only do these players see the best way to get their ball near the hole, they have the skills to hit shots that end up in the right places. This is of course only exacerbated by conditions--that's why pros love wet courses. They know the ball will just stop where it lands. Even on very green courses, you can see shots land in similar places and one will take a big hop and another will stop quick. Whether that's random or the result of ball striking is pretty much impossible to tell from TV or the gallery.

[Tom_Doak]

Jason:
Not sure your methodology here makes sense, because you are assuming that these golfers are equal in all other skills, which they're not.

I think what you seek to prove would be better illustrated by analyzing whether the year's leaders in strokes saved putting gained more strikes against the field at Chambers than at other venues.

What you've shown is that Chambers more than most courses favors long hitters.

[Jason Thurman]

Tom, that was what I wanted to explore initially. However, strokes gained putting statistics are not available for the US Open. Not sure if they will be made available eventually or not. I agree that it would be preferential to isolate strokes gained at each course if the calculation is eventually made available.

[Patrick_Mucci]

Is there a golf course in the world that doesn't favor longer hitters ?

[Tom_Doak]

Is there a golf course in the world that doesn't favor longer hitters ?

Some more than others, surely.  It would be quite interesting to examine all the Tour's stats from week to week to see which courses really do reward which strengths.  Too bad they don't play more courses that are interesting ...

[Carl Johnson]

Doesn't quantum mechanics say that everything is random?  And therefore that skill doesn't really exist?     (I'm inspired by the recent book, The Quantum Moment, by Robert P. Crease and Alfred Scharff Goldhaber.)

[Dave McCollum]

It takes something like 300 volunteers to run all that shotlink stuff.  Not sure the PGA Tour would want to do it for the USGA unless they were in for a cut of the action. 

[Paul Gray]

I'm not sure this is anything that anyone doesn't already know but, since Jason raises the question, I will briefly chirp in by categorically stating that windy conditions DEFINITELY require better ball striking. Spend a little time playing links golf and see how your ball DOESN'T deviate so much when struck well as opposed to when the strike is less than perfect.