“Angles Don’t Matter”

[Don Mahaffey]

Erik, on tour, more putts are made for par than for birdie.  It’s a fact. They are not robots.   They don’t all play the same no matter the circumstance. 

[Charlie Goerges]

First, still, is that I don't think most courses have much strategic value to them, so saying that angles don't matter because of stats is just a tautology.

So the flag on the right, you don't think it's generally advantageous to be coming into that from the left side of the fairway over the right side? Generally speaking?

If someone told me 90+% of courses had no discernable strategy whatsoever, nothing in my experience would contradict that.

I'd strongly disagree.

I'm probably exaggerating the numbers a bit, but the types of courses I've played most of my golf on really haven't had much to them. The small number of courses where the angles have mattered were always the most engaging to me. My preference for the latter is large, even if the scoring difference may not be.

[Rob Marshall]

Erik, on tour, more putts are made for par than for birdie.  It’s a fact. They are not robots.   They don’t all play the same no matter the circumstance.

https://www.golfwrx.com/247705/study-why-do-tour-players-make-more-par-putts-than-birdie-putts/




The article examines the misses of Tour players through ShotLink data. They tested to see if there was a bias by Tour players in the direction of their misses (left or right) and found none. What they did find, however, were three major points:  Tour players miss birdie putts short of the cup a higher percentage of the time than on par-or-worse putts.

• Tour players miss more birdie putts early in events.

• Tour players miss a higher percentage of birdie putts short of the cup earlier in events.

• That’s how they came to their conclusion of loss aversion — Tour players are missing birdie putts short of the cup because they are afraid of hitting them too far past the cup and three-putting. And with par-or-worse putts, Tour players are missing them past the cup because they are averse to losing strokes and therefore willing to hit putts with a more aggressive speed to make sure the ball gets to the hole.

[Charlie Goerges]

This might be a little off the wall, but here goes. In the NBA some have bemoaned the effect analytics have had by prejudicing the 3-point shot. An article I read suggested that the older style of play could be encouraged by changing the rules to move the 3-point line. I wonder what effect a rule change like a shorter, spinnier ball would have on this whole thing?

[Jason Topp]

If accurate, this data undermines the fundamental concept underlying strategic design - tempt a hazard to gain an advantage.

That may be true, but I don't think it necessarily is true. Someone has already brought up the idea that being on the wrong side may result in more conservative play, which is often the correct play anyway. Additionally there is the possibility that most courses (even those played by the pros) are only really loosely strategic at best. And finally, there is the conditioning. All added up, they blunt the effects as far as I'm concerned, and don't take away the efficacy of truly strategic design.


It might mean that we need to alter the types of hazards and features we use in order to influence play. Rough and sand don't really force the player to hit lower-trajectory shots, maybe more use of severe contours or trees would help force lower shots, especially on better players.

Charlie - the other aspect of the analysis is that even the best players in the world will scatter tee shots in a 65 yard area left to right.  The idea is that the penalty of not being in play is severe enough you need to aim far enough away from trouble to make sure the penalty will not come into play.  The cost of a full shot eliminates any advantage that might accrue from having a better angle at a particular green.   I find the analysis compelling even though I do not like it.   

[Erik J. Barzeski]

This stuff isn't even on topic.

Erik, on tour, more putts are made for par than for birdie.  It’s a fact. They are not robots.   They don’t all play the same no matter the circumstance.

I haven't denied that. I have shown data that very little of the already small difference is due to "human nature" or loss aversion or whatever. That very little of an already small difference is due to this:

A client of mine, very much into numbers as a finance guy, told me the tour stats show PGA players make 8 footers for par more often than 8 footers for birdie. Why? He said the same reason human investors ride their losers and sell their winners.  Why? Because they are human.

Rob, bad choice.

This past season, PGA Tour players made an average of 39.8 percent of their birdie putts from 5-to-15 feet while making 52.6 percent of their par-or-worse putts from the same distance. That means that Tour players make a higher percentage of par-or-worse putts than they make birdie putts from the same distance.

I shouldn't have to explain why that's a horrible way to try to look at this sort of thing. FIVE TO FIFTEEN FEET? I'd wager everyone's house here that the distribution of the par putts was closer to the five foot end of the range than the distribution of the birdie putts. Five to fifteen feet is a volatile range… and so even a small difference in the distribution can change the average make rate significantly.

The 2011 study made no adjustments for first or second putts or anything else, really. You could have also cited some of the more recent studies (newer than 2011). Ones that make adjustments that have nothing to do with the score of the resulting putt or ones that don't rely on comparing ten-foot wide ranges with unknown distributions of putts. Or you could have cited Broadie's book, which says:

Even though 11,000 putts might sound like a lot, it represents less than 1% of the putting data in the ShotLink database. Devin Pope and Maurice Schweitzer [the 2011 study] analyzed the ShotLink data for the par-birdie effect. In their 2011 American Economic Review article, they concluded that “professional golfers hit birdie putts less accurately than they hit otherwise similar par putts.” In 2011, I asked Mark Calcavecchia about the effect and he said, “It’s just human nature.” [Like Don contends.] Intrigued by these results, I looked into the data myself. Sure enough, for putts between four and seven feet, pros sink 3.6% more par putts than birdie putts. But that figure doesn’t take into consideration that par putts in this range tend to be second putts, while birdie putts tend to be first putts. We already know that second putts are easier than first putts because of the learning effect.

After controlling for first-putt–second-putt differences, and controlling for uphill, downhill, and sidehill differences, the par-birdie effect is reduced by more than half. Looking at all putt distances, taking into account differences in strokes gained (not just the one-putt probabilities) and the frequency of putts, I calculate an effect of 0.1 strokes per round. But even this computation overestimates the birdie-par effect. A short first putt for par can happen after chipping from off the green, so the golfer gets to see the path of the chip before hitting his putt. Before putting, a golfer often gets to see the putts of other golfers in the group putting along a similar line.

It seems likely that the par-birdie effect is less than 0.1 strokes per round. Far more important, in my estimation, is the performance increase to be gained from going to school. Watch your putts and the putts of others in your group and read the contours of the green, especially near the hole.

Broadie, Mark. Every Shot Counts: Using the Revolutionary Strokes Gained Approach to Improve Your Golf Performance and Strategy (p. 160). Penguin Publishing Group. Kindle Edition.

Can we talk about the actual topic now? I'm sorry the actual data does not align with your anecdata.

[Rob Marshall]

No choice, I simply posted an article on the topic. I am curious as to why a guy from Erie has a southern accent Y’all….

[Ben Hollerbach]

I found the quotes by Lou Stagner to be staggeringly devoid of context. Social media and generalized statistics both share that commonality I suppose.

This behavior has become all too common, especially on social media; make a broad claim, one that may buck long standing commonly held beliefs, but provide little context as to how or why the claim may be true. Generate interest and attention but do little to support the claim.

I appreciate that Erik's commonly used narrative, "Angles matter when the ball is rolling" is slightly more descriptive, but still lacking support.

Outside of a very very few number of shots, at some point of time during a ball's travel,  the ball will roll, even if it is only for a few inches. So that would imply that angles matter on almost all shots. Beyond the exceptional few where  the ball is flown into the hole directly, lands in water or some other hazard that deadens movement upon impact, or gets stuck in a tree.

Of course this is probably not what Erik is trying to say, but without proper context and support how can we be sure?
A better, more comprehensive, pair of statements may be something like:

• On a well struck shot, the approach angle matters to a varying degree based upon factors such as decent angle & spin on the ball and firmness, slope, grass height, & proximity to hazard on the ground at the point of impact.

• On a poorly struck shot, the approach angle matters in relation to the proximity of potential hazards located along the intended flight line, both short and long of the intended landing zone.

At least with statements such as these there are specific & measurable factors in play that can be property considered and qualified when discussing the impact of angles on any given shot for any given player.

[Erik J. Barzeski]

I appreciate that Erik's commonly used narrative, "Angles matter when the ball is rolling" is slightly more descriptive, but still lacking support.

That's been backed up several times before. This is about the tenth time we've had this discussion in the last five years.

Outside of a very very few number of shots, at some point of time during a ball's travel,  the ball will roll, even if it is only for a few inches.

Yeah, cuz that's what we're talking about. Four feet of roll. Or four inches. Oy.

[Ben Sims]

This is about the tenth time we've had this discussion in the last five years.

Erik,


There’s a number of people on this site that have been discussing the same complex and esoteric subjects over and over again for the better part of three decades. I’m not sure why this subject being rehashed is irking you. It’s provocative, misunderstood, and flies in the face of lifetimes worth of common golf knowledge. In your position, I’d be more than happy the subject was coming up yet again. Cheers.

[Ben Hollerbach]

During the numerous times this topic has been discussed on this site I've seen inferences to data and the study on the subject, but I don't think I've ever seen direct references to the actual measured data or to published studies on the subject. Thus the question behind context supporting the claim.

While we have all seen PGA Tour pros fly balls that land on greens and stop on a dime, we've also seen PGA Tour pros hit shots that land and run well away from their intended target. If one shot's landing characteristic minimizes the impact of the approach angle and the other shot heightens the impact of the approach angle, there is a basis for greater discussion on the subject. Potentially an excellent case for study is the different setups of Muirfield Village for the Workday and Memorial held in back to back weeks during 2020. Same course, but the change in setup from one week to another presented a dramatically different challenge to the players.

Did the change in course conditions between the two weeks lead to approach angle having a greater impact on scoring in week 2 vs. week 1? If so, what factors in the course setup contributed the most in that change of impact?

[Ben Hollerbach]

Erik,

There’s a number of people on this site that have been discussing the same complex and esoteric subjects over and over again for the better part of three decades. I’m not sure why this subject being rehashed is irking you. It’s provocative, misunderstood, and flies in the face of lifetimes worth of common golf knowledge. In your position, I’d be more than happy the subject was coming up yet again. Cheers.

You'd think, but he'd rather tell someone that they're wrong than explain to them why he's right.

[Erik J. Barzeski]

There’s a number of people on this site that have been discussing the same complex and esoteric subjects over and over again for the better part of three decades. I’m not sure why this subject being rehashed is irking you.

It doesn't irk me — I just don't see the point in re-hashing everything that's been written many times in much depth. The old posts are still here. The old graphics.

During the numerous times this topic has been discussed on this site I've seen inferences to data and the study on the subject, but I don't think I've ever seen direct references to the actual measured data or to published studies on the subject. Thus the question behind context supporting the claim.

Ben, it's tough to take you seriously when you make statements like: "Outside of a very very few number of shots, at some point of time during a ball's travel,  the ball will roll, even if it is only for a few inches. So that would imply that angles matter on almost all shots." Yeah, cuz that's what people are talking about when they cite the Presidents Cup at Royal Melbourne. That's what people are saying when they're talking about balls rolling. A few inches. Sheesh.

The "measured data" is often proprietary, whether it's ShotLink data that only a few people are given access to, or data from the likes of Arccos, ShotScope, GolfMetrics, etc. There's no "published study" because it's just published data. Lou can post a chart. Mark can post information. Scott can post information. I can post what I see. These aren't scientific papers with a hypothesis and a test and a conclusion. It's a heap of data that's interpreted.

What kind of "published study" do you need to see if someone says "PGA Tour players make 50% of their putts from 8'2"?" That's not something you write a paper about. It's a statistic. The stats that show that angles don't matter (and that there's often a very, very small advantage in scoring to being on the "bad" angle) are out there, and the means by which they're determined are out there. I've shared them, Lou has shared them, etc.

Here's a podcast talking about this one specifically, and dealing with average golfers (not Tour players): https://podcasts.apple.com/us/podcast/is-there-such-a-thing-as-a-good-angle/id1543363295?i=1000596708327 or https://overcast.fm/+mucWG_TXE.

If you disagree with the interpretation, then find the holes or flaws if there are any. And I welcome that challenge, as I enjoy being told "you're wrong" if the person can tell me why. "Ha!" you say? Then it goes to show how very little you know of me. Being told "you're wrong and here's why" is what I call an instant opportunity to upgrade my knowledge. Being told "yeah, I think you're right" doesn't do anything for me. I wouldn't be saying it if I didn't think I was probably right. Being told "you're wrong, here's why" is awesome. It just doesn't happen too often.

If one shot's landing characteristic minimizes the impact of the approach angle and the other shot heightens the impact of the approach angle, there is a basis for greater discussion on the subject.

Hence… me saying "angles matter when the ball is rolling." Angles don't matter when you fly it and stop it relatively close to where it lands. I've further said that this is for shooting the lowest score, on average, not for interest, or challenge, or art, or show, or whatever. I've also said it's generalized data, and when talking about specifics (players, shots, etc.) there are exceptions.

I've talked with my daughter about angles. She's playing college golf, but she's also 5' tall. Pound for pound or inch for inch she hits it really far, but… that means she's hitting a hybrid into a par four a few times a round. Angles matter to her… because her ball is bouncing/rolling.

Show me where I'm wrong, and don't use anecdata or hypotheticals or exceptions.

[Ally Mcintosh]

Look, I’m with Erik: Angles matter a lot less than we - especially on this website - would like to think.


That doesn’t mean we should stop trying to make a difference to golf course strategy using angles. Because when they do make a difference, it really does add to the game.

[Sean_A]

It sounds to me that Eric is saying angles do matter? We can talk about the whos, whens and whys, but it is very clear and obvious that angles matter.

I guess the inference of angles don't matter was a nudge to archies to forget about angles when designing courses? It's hard to imagine how much interest would be stripped out of the game if archies followed that nudge.

Ciao

[Brett Meyer]

My issue with big data on this issue is that it's pooling over very different types of golf holes and giving an average 'effect' when there could be some interesting heterogeneity for holes that are designed different ways. I'd suspect that angles matter more for holes which are designed for angles to matter than ones which aren't.

Arccos' data come from at least tens of thousands of shots over at least thousands of different holes. Maybe the people who use Arccos play better courses on average than the average golfer, but I'd imagine that the average hole isn't setting up too stark a difference in angle from one part of the fairway vs. the other.

But would the results look the same for a sample of holes like the 8th at Pacific Dunes or the 10th at Riviera, where the green is narrow and angled toward one side of the fairway? Maybe not. But it's a strong hypothesis that angles would matter more with this type of architecture and I haven't seen an attempt to address it in Lou Stagner's many Twitter posts. To be fair, it'd be hard to test because you'd have to figure out a way to code holes for architecture.

My suggestion to Arccos would be to run an experiment, or at least half of one. You have the control group data for results from good/bad angles on the sort-of average hole. Now put a bunch of players on the good/bad side of the fairway on holes where a group of experts agree that the design of the hole makes angles more relevant. Do you get the same result? I don't know. But I haven't seen any data so far that convince me that we have evidence that angles don't matter in such a situation...although admittedly I haven't dug around for it.

[Erik J. Barzeski]

That doesn’t mean we should stop trying to make a difference to golf course strategy using angles. Because when they do make a difference, it really does add to the game.

Yeah… it adds visually, it makes people think, it tempts people… and there are ways to kinda throw a wrench into the system for some guys.

I understand it can be tough to play the angles thing with doglegs, etc. but maybe there's something to that… and a different sort of angles. Angles off the tee, because a drive tends to roll out some (unless it's really soft). Force people to try to work the ball or risk running through the fairway? (The difficulty being not everyone hits it the same distance, so that can be tough.)

It sounds to me that Eric is saying angles do matter? We can talk about the whos, whens and whys, but it is very clear and obvious that angles matter.

When the ball rolls, which it does for a LOT of golfers, even very good ones (my daughter got down to a + index competitively for a few months, and is still a 1 or so).

Arccos' data come from at least tens of thousands of shots over at least thousands of different holes.

It's like 600 million, I think. (Not all approach shots of course.)

But would the results look the same for a sample of holes like the 8th at Pacific Dunes or the 10th at Riviera, where the green is narrow and angled toward one side of the fairway?

Play around with it yourself:
https://app.powerbi.com/view?r=eyJrIjoiMzczZmUyMWQtYzA0Yi00ZmNlLWFmNmEtZWRlNjViZTU2M2Y5IiwidCI6ImJiNjY5NzU2LWM0YTktNDYwMS1hOWYyLWQyNDRlNTQzNzk3MSIsImMiOjJ9

But you'll probably find… no.
https://twitter.com/LouStagner/status/1350185425923670028

Some other relevant tweets (these are from the pro game):
https://twitter.com/LouStagner/status/1219649236880543744
https://twitter.com/LouStagner/status/1220141484336459776

But it's a strong hypothesis that angles would matter more with this type of architecture and I haven't seen an attempt to address it in Lou Stagner's many Twitter posts. To be fair, it'd be hard to test because you'd have to figure out a way to code holes for architecture.

Yes, unless you're just talking about a specific hole, in which case you either basically have PGA Tour data with a few hundred or thousand shots depending on how long the tournament has been there, or you have to hope you pick a popular golf course to get into the hundreds of shots to make it statistically valid.

My suggestion to Arccos would be to run an experiment, or at least half of one. You have the control group data for results from good/bad angles on the sort-of average hole. Now put a bunch of players on the good/bad side of the fairway on holes where a group of experts agree that the design of the hole makes angles more relevant. Do you get the same result? I don't know. But I haven't seen any data so far that convince me that we have evidence that angles don't matter in such a situation...although admittedly I haven't dug around for it.

I agree that might be interesting. Again, most of the "angles" stuff, because right now it's one of the few ways to do it, has to do with just saying "the flag is within six yards of the right or left edge of the green."

Lou has done some case studies on some specific holes, as have I for some holes. The data is often smaller, and perhaps I chose poor examples (one I chose had a huge/deep greenside bunker guarding the front right, making the left side a "better angle") and… results were basically the same until you got to the higher handicappers, likely because… they could thin the ball or top it or something and it might roll onto the green from the left side, while better players were hitting over the bunker from either side.

Full day of teaching. Have a good day. Even Ben, who is still probably trying to figure out how 18 yards became 35…

[Tom_Doak]

I’m confused.  If strategy doesn’t matter, how do Erik or Lou make money talking about it?  Just hit driver everywhere, and aim well away from the o.b. or water.


Oh, it’s more complicated than that?  Well, so is design.


The most important thing to understand is that architects don’t (or shouldn’t) care what score you shoot.  It’s our job to make the playing field interesting.  Most of that is about visual presentation:  we have no control over where you hit your ball, but we can try to influence where you aim.


If you want to make the game boring and calculate the optimally safe route to save 0.1 strokes per hole, instead of taking on the challenges we present, that’s totally on you.  But that doesn’t mean I’m going to stop putting them out there.

[Ian Andrew]

The most important thing to understand is that architects don’t (or shouldn’t) care what score you shoot.  It’s our job to make the playing field interesting.  Most of that is about visual presentation:  we have no control over where you hit your ball, but we can try to influence where you aim.

I played Riviera with your friend Norm Klopardra (Tom, I hope I have the spelling right). Btw, you set the game up!

On the 10th tee, he spent a few minutes talking about the hole design. He finished by explaining why you should never hit driver on the 10th. It was so effective that I laid up. I think my playing partners, Mark and Tom, played positionally too. He was really convincing. Norm went last, hit driver and made an awful looking six by going right. We were perplexed about why he didn't follow his own advice about percentages and strategy (taking the major risk out of the hole).

He explained that every once "he can't help himself" and will hit driver.


As a general thought, some play percentages and some love risk (that's me). Most are flexible. They take chances based upon the feeling of the moment. Feelings matter. The reward of overcoming something challenging is one of golf's "stickiest" attributes. It provides a rush that we don't get from simple execution. Angles matter because its an opportunity to overcome a risk.

If golf was simply "bowling" with all the trouble at the sides and strictly about execution, I wouldn't play the game. It's not enough for me. I need to roll the bones every once and a while and feel the thrill of making the carry.

[Ira Fishman]

Erik points to an interesting point when he mentions angles of the tee and doglegs. A dogleg creates a different kind of angle than hazards around a green. I do not know what that stats say, but it seems intuitive that if you end up with a longer second or third shot because you ended up on the side away from the dogleg, your score likely is to affected.


One of the reasons that I like centerline bunkers is that they create multiple options off of tee that consist of different angles. When well done, they also can present a risk reward dilemma even for longer hitters. Plus for those hitters, a centerline bunker minimizes the ability to just play away from trouble.


Ira

[Ben Sims]

Oh, it’s more complicated than that?  Well, so is design.


The most important thing to understand is that architects don’t (or shouldn’t) care what score you shoot.  It’s our job to make the playing field interesting.  Most of that is about visual presentation:  we have no control over where you hit your ball, but we can try to influence where you aim.

If you write another book, make this the inside of the dust jacket. Without shining too much ass, this is among the best dozen or so things I’ve ever read or heard you say. I’ll take partial credit for being the OP that prompted it, obviously. 


I said something earlier about data analytics (for the avg golfer) becoming mainstream enough that it changes what architects do to provide interest, tempt, or confuse. This quote makes me feel a better about how that will work out going forward.


This thread has been enormously informative from a number of perspectives. I can’t imagine a world where I’d stop playing golf to have fun. Shooting better scores is part of that, but taking on and succeeding against poor odds is a bigger part of what makes golf great.

[Kalen Braley]

I think Brett did a nice job laying out many of the issue I have.

Actually "proving" that Angles actually don't matter is a very difficult to thing to do given the near endless variables at play:

Length of hole, size of green, undulation of green, amount of bunkers and where their positioned, fairway width, depth of bunkers, rough height, fairway height, length of approach, type of grass, weather conditions, strengths and preferences of golfers, etc, etc, etc,

The burden of proof lies with those making the claims, to clearly and succinctly lay out the methodology, controls, processes, and data to arrive at said conclusion, not with others to disprove the hypothesis.

The Sagan Standard could certainly be applied here: "extraordinary claims require extraordinary evidence"

[Erik J. Barzeski]

Oh, it’s more complicated than that?  Well, so is design.

Thing is… it's really not much more complicated than that, no. I'm sure design is WAY more complicated than strategy. Just figuring out how to drain water is probably an order of magnitude more difficult.

The most important thing to understand is that architects don’t (or shouldn’t) care what score you shoot.  It’s our job to make the playing field interesting.  Most of that is about visual presentation:  we have no control over where you hit your ball, but we can try to influence where you aim.

Of course.

If you want to make the game boring and calculate the optimally safe route to save 0.1 strokes per hole, instead of taking on the challenges we present, that’s totally on you.  But that doesn’t mean I’m going to stop putting them out there.

Doing that doesn't make the game boring. You've still gotta hit the shots. You've still got four hours outside with your buddies. You've still got a beautiful golf course to look at.

On the 10th tee, he spent a few minutes talking about the hole design. He finished by explaining why you should never hit driver on the 10th. It was so effective that I laid up.

Laying up is generally not the best strategy there.

Erik points to an interesting point when he mentions angles of the tee and doglegs. A dogleg creates a different kind of angle than hazards around a green. I do not know what that stats say, but it seems intuitive that if you end up with a longer second or third shot because you ended up on the side away from the dogleg, your score likely is to affected.

Naturally, yes.

When I think of doglegs, for example… the 17th at Pine Needles. even if you carry the left-hand bunker, it can be tough to stop your ball from rolling through into the trees on the right. Play right of the bunker, and you leave yourself a really long approach.

I said something earlier about data analytics (for the avg golfer) becoming mainstream enough that it changes what architects do to provide interest, tempt, or confuse. This quote makes me feel a better about how that will work out going forward.

That, and the fact that this "statistical driven strategy" stuff is only ever really known to a tiny percentage of golfers. Heck, everyone here is a golf nerd, and most of y'all don't believe me.    So, there are always going to be plenty of people who think angles matter in times when they really don't (for scoring, not for interest/whatever), etc. 

This thread has been enormously informative from a number of perspectives. I can’t imagine a world where I’d stop playing golf to have fun. Shooting better scores is part of that, but taking on and succeeding against poor odds is a bigger part of what makes golf great.

Golf is still ridiculously difficult. You're ALWAYS battling poor odds, even if you have a "perfect" strategy. I reject out of hand the notion that playing strategically makes golf "less fun."

I like to use Tobacco Road as an example. You can play it a number of ways. You can go after everything, letting yourself get baited into going for every dumb shot out there. You can shoot 83, but pull off ONE of those shots, and you remember that. And if you enjoy golf that way, cool! Great. I'm glad, as ultimately… do what you want. Do what makes you happy.

What makes me happy is deciphering how best to play the hole, then pulling off the shots to do so. A "boring" 69 at Tobacco Road is more interesting to me and more fun for me than a 78 where I go for everything even though I know it's the "wrong" play.

Golf is still plenty hard enough. You've still gotta hit good shots.

[archie_struthers]

 


The idea that the tour players have a lower score from the "wrong" side is really eye opening. Kind of reminds me of Jack Nicklaus playing the safe shot , almost never flag hunting unless it perfectly fit his shape. This despite having more talent than almost anyone who ever lived!

[Brett Meyer]

But would the results look the same for a sample of holes like the 8th at Pacific Dunes or the 10th at Riviera, where the green is narrow and angled toward one side of the fairway?

Play around with it yourself:
https://app.powerbi.com/view?r=eyJrIjoiMzczZmUyMWQtYzA0Yi00ZmNlLWFmNmEtZWRlNjViZTU2M2Y5IiwidCI6ImJiNjY5NzU2LWM0YTktNDYwMS1hOWYyLWQyNDRlNTQzNzk3MSIsImMiOjJ9

But you'll probably find… no.
https://twitter.com/LouStagner/status/1350185425923670028

Some other relevant tweets (these are from the pro game):
https://twitter.com/LouStagner/status/1219649236880543744
https://twitter.com/LouStagner/status/1220141484336459776

Thanks for posting this Erik. I did what you said and played around with the 10th hole at Riviera.

But first of all, Stagner's post above about the 10th hole at Riviera didn't address the issue of angles at all. He compared layups to attempts at the green. To test whether angles matter, you'd want to compare average scores for those in the left side of the fairway vs. those in the right side of the fairway, maybe just to pins on the right half of the green. If scores are lower from the left side, that's some evidence that angles matter.

To test this, I subsetted the Riviera data to shots under 250 yards to pins on the right half of the green (pins 3 and 4). Here's what that looks like:



What I'd really want to do is see the average score for those left of pin placement 2 vs. those to the right of it. It doesn't look like I can do that. But using the eyeball test, it looks to me like there might be a few more birdies and a few fewer bogeys relative to pars left of that line vs. right of it. I suspect that the scoring from left of the line is slightly lower than from right of the line, but maybe not a huge difference.

But that raises another issue that I have with using big data: even if the scores from the left half of the fairway here were lower than from the right have, that still would be a bit lacking as evidence for my angles theory. Why? Because there's a good chance that players who hit it in the right side of the fairway were playing worse than those who hit it in the left half of the fairway. I don't think that anyone would intentionally aim at the right side of the fairway here if laying up. So our 'effect' of angles might be polluted by an effect of playing worse.

One way to help address this would be to control for the player's scoring or, better yet, strokes-gained driving to this point in the round and maybe even for a broader set of recent rounds. Now Arccos has this data and they could do this test. I'm sure they do this type of analysis in other scenarios. They should do it here too and Lou Stagner should post some of that on Twitter, or maybe in a Substack.

Ultimately the problem with big data is that if you have any type of bias in the data generating process, you can get very misleading results. Arccos has the data to address some of this with more sophisticated data analysis. But you can't get around the fundamental fact that different trending players are playing from the good/bad side of the fairway and that you don't have as many of the good-trending players playing from the bad side and the bad-trending players playing from the good side. So angles aside, you're likely not getting a second shot of the same quality from those on the good side vs. the bad side and ultimately there's no way to control for that...meaning that you really need an experiment to test this.