Welcome back to another edition of the Frida Egg Podcast. My name is Garrett Morrison and this episode is brought to you by ourselves. So we're having a Black Friday sale that is this Friday at our pro shop on our website, so you can find it at proshop dot Thefrida Egg dot com and it's an automatic twenty percent discount off of everything that includes hats, shirts, headcovers, turvice tumblers. But I wanted to talk specifically about our photography prints.
So these are drone shots of great golf courses. Most of the shots were taken by Andy Johnson. You can get them framed or mounted on metal. We've got pictures of bally Neil, Sand Valley, Prairie Dunes, Pasa Tiempo, and we've got a ton of new ones on the way. They look great. I've got one on my wall myself, and they make really good gifts. So this Friday, November twenty seventh, twenty percent off the whole store, including our prints. That's at the pro Shop, pro show dot Thefrida Egg
dot com. All right, so on with our episode. My guest today is Matt Korshane along with his brother Will Korshane Matt founded Data Golf, which you can find at datagolf dot com. So Data Golf is really part of what has been a revolution in golf over the past couple of decades, and that has been this infusion of advanced data, statistics and analytics into golf, especially golf at
the highest professional levels. Now part of that story is Shotlink, which provides a great deal of raw data that simply wasn't available in the nineteen nineties. And another part of it is the advent of strokes gained statistics, which were pioneered by the Columbia Business School professor Mark Brody. And basically, what strokes gained does is it compares a player's performance to the rest of the field, shot by shot, and in this way you can really isolate the value or
quality of each shot a player hits. Now, in strokes gained, with each shot the player gains or loses a certain amount on the field. And what this has done is it's just given us a much better way to analyze the different skills that a player has. We really have a better idea of who the great drivers and iron players actually are because we've been able to zero in
on the quality of each strike of the ball. But you know, strokes gained is like still in its infancy and it's just begun to shift our understanding of golf. I think where we're just scratching the surface, and that's the exciting arena in which Data Golf is working right now. Matt and Will Korshane are very young, but also very qualified,
and they're doing a lot of interesting work. One of the things that I think Data Golf could change is how we view golf courses in golf course design now, I always think that golf course design will be an art and assessing golf course design will be an art as well. Know, no, data, I don't think can tell us if a golf course is good, but it can definitely tell us how. Data can definitely tell us how different tournament venues demand different skill sets from players. And
I find that to be a really compelling question. And so that's what I wanted to talk with Matt Corshane about because Data Golf, the website that he and his brother have developed, has some great tools that use strokes gained data to tell you something about the golf courses on the PGA Tour, specifically this week's host of the Mayacoba Golf Classic, El kamal Ane Golf Club outside of Cancun, Mexico, is one of the biggest outliers in Data Golf's model.
It's such a strange course. It just demands this extremely unusual type of performance from the top players. So I thought it was a good time to get at a question that I've had for a long time with the help of Matt Course Shane, what can statistics tell us about golf course architecture on the PGA Tour. I hope you enjoy I miss the green for example, I'm already upset when I find my ball in the bunker, I'm really upset.
And when I.
Find my ball in a brid egg Friday egg, the dreaded Frida egg fridagg Frida egg egg Frida egg bride egg Lie, I'm about ready to run off the golf course. So, just to find out a little bit about your background in golf, you were a pretty excellent golfer. You got down to a scratch handicap, You played some college golf at Queen's University as a competitive golfer. Did data ever enter your life?
No, it really didn't.
Yeah, I wasn't logging like the Strokes game stats or anything. I was not a and I'm still not. I'm very much a I mean, I'm a vivid head case on the golf course. Now, I think I'm very much a field player. I guess I would say I'm not like a yeah, technical robotic player by any means.
That's an interesting contrast to me. So like if data were you know, some of these analytics were around back then, but they just weren't as commonly used, I guess, especially at the college golf level. Do you think anything would have changed.
For you looking back when I was actually a serious player and didn't have because honestly, the way I think about it now is I just have on the list of the top ten things I need to improve in golf, understanding how to better use data is pretty far down that list. I need to stop like blocking my opening TV all ob before I start worrying about that stuff.
But like, yeah, no, back in the day when I was good, it would be Yeah, it'd be super informalve just to get the basic numbers like the Strokes game stuff, just using Brodie's app, Like just understanding what you're losing strokes is super valuable, and I think it's often for
most people it's counter intuitive. I think probably a lot of people would end myself included, would probably realize they're losing most shots, like on the long game, whenever you play poorly, or always thinking about all these pots you miss, et cetera. But really everybody misses a lot of potts, so I think, yeah, But anyway to answer the question, I think I would have taken advantage of this stuff.
It was more easily accessible in ten years ago.
Well. Hearing golfers talk about it, it's almost like, in addition to providing knowledge, these stats sometimes provide a kind of peace of mind, like you know what's wrong, yeah, or or you can contextualize your bad performances as just like this is what happens. You know, it's not a judgment on me as as a player. This is just what happens.
Yeah.
And it's also I kind of think that is the right way to think about it, because it's also debatable how like actionable this stuff is. Like if I find out that I am lacking, well, if I'm losing small good example of if I'm losing strokes, offer te it's like okay, like, yeah, if I hit.
It further, that'd be great. But I mean there are examples of players.
Who have like doing for Telly last week was getting a lot of attention for he's not a massive guy, and he managed to pick up some club at speech.
So there are things you can do.
But I've always thought, I think my brother and I've always thought that, Yeah, it's more like descriptive. It's nice to know where you're losing strokes and what is the difference between myself and like the number one player in the world.
But I don't know how actionable a lot of it is.
So right now you're doing a PhD at the University of British Columbia in economics. Where did your interest in economics come from?
So my grandfather is, well, he's still alive.
He was a pretty stop working but he was a pretty well known economist, so there was always that connection. I actually did my undergrad in biochemistry, which was I mean, I was just a classic. I just went to university not really knowing what to do, and I was like, oh, I mean why not medical school?
Like that sounds reasonable.
It's embarrassing looking back on it, but then I eventually realized I think it's like a few Coon courses and just sort of liked the general approach to things. Just using statistical methods, which are particularly the ones that are prominent in economics, Just using those as a way to understand the world is. It's just changed the way I think about a lot of things, and I've really become I really enjoy looking at the world through that lens.
So at some point you combined your background in golf with your interest in statistics and informed data golf. How did the idea for data golf come about?
Well, I guess the first thing I just say is, so data Golf is run by myself from my younger brother, Will, who's so he's two years younger. Yeah, he was working at that point, and I was just in my second year of the PhDs.
This is like four or five years ago. And then the Yeah, the PJ Tour used to.
Have an academic program where you could access the shot link data for free, and so we got that data and we just started.
Yeah, we just started a Twitter account.
Started a WordPress blog and started doing some basic analyses with the data, trying to just sort of answer.
It's funny to look back.
We still have all these blogs somewhere on our old website. It's funny to look back on them. But yeah, just answering questions like do players in fact, there's the same wording out you shoot a really good round, it's hard to follow it up, So we were sort of just checking the degree to which that's all or true and yeah, and then it's sort of just progressed from there.
Right for sure. So it's really evolved well past the original blog. So the original Data Golf blog was basically a word Press basic WordPress blog, and now it's this very complex website with not only blog posts, but a number of predictive models and interactive tools that kind of make sense of the huge amount of data that's coming out of professional golf right now. And so I wanted
to focus actually on the interactive tools and visualizations. So you know, to someone like me, I find these very useful and very informative just in general, How do you decide what kinds of tools to build?
Yeah, so I think a lot of the tools come naturally out of the model that we have, So really like the centerpiece of the website and what we do with Data Golf is this predictive model. We have of golfer perform, where essentially the output of that model is we're trying to, like, ultimately what matters in golf is strokes gain, how many strokes you're beating the field by. So that's ultimately what we're trying to predict with this model.
And along the way there will be things that I don't know, sort to pop out of the model that we can make into a webpage.
So a good example is true strokes gained.
So true strokes gained is it's regular strokes gain, which is how many strokes beat the feeld by, except we're gonna.
Adjust that for the strength of that field.
So, for example, on the web dot com, if we estimate that those fields are like a shot worse on average than TJ tour field, So if you beat a web dot com field by two shots, that's gonna be worth like a true strokes gin of one because we're saying this field is one shot.
Worse than average. So that's something that sort of comes out of our model.
Another example of a page would be the course history stuff or the course fit tool. We have that output in our in our model, and at some point we say, wow, this is interesting. I think general golf fans would like this. Let's turn it into like an intuitive page that people can can get something out of.
So where that's where a lot of them come from.
So I did want to dig a little bit deeper into what you call the course fit tool, because I think it does have a lot to say about what one of our primary interests at the fried Egg is, and that's golf course design and set up and how design and set up influence play on the professional tours. But I mean, I think maybe before getting into course fit specifically, maybe we should lay the groundwork with this
related but very different notion of course history. Can you tell me what course history is and some of the issues that you've had with it.
Yeah.
So, course istuy in the most general sense, is just how a player has performed at a specific course in the past. And then the way we analyze it is not just how like a player has performed in an absolute sense at a course, because, for example, Tiger, if you just define course a player's course history as just their total stroke seeing at that course, and somebody like Tiger is going to have the best course history everywhere. So course history, the way we define it is performance
relative to expectation or baseline. So and by expectation or baseline that's going to be intuitally, that's like how a player performs across all courses in a given year, that's going to be their baseline. So it might be there might be a plus one strokes gained player. And then if there's a specific specific course, say I guess the National, where they've performed at an average of plus two strokes gained, we're gonna say they have good course history, dick, because
they performed above their baseline. The main issue with course history in terms of trying to use it for actionable things like predicting performance, is that it's generally a small sample.
If a guy has only played four rounds of the course and he's played really well, maybe he won the tournament, which is not not too uncommon, what can we really say from that, Like from our analysis, we've found like if a guy is performing one stroke better than expected at a course, historically, we're only going to improve his predictive performance going forward by maybe like two or three percent per round.
Let's say, so, well, if he's if he's done that for four.
Rounds, we would probably bump up his predicted skill by like zero point zero five if he's been one stroke planner.
So it's a pretty small adjustment, but those things can matter.
Yeah, it seems like your opinion on this has has shifted a little bit over time, but the general critique of how maybe the lay person uses course history is still valid. You know, when when you hear people talking about how a certain player has done well at a course in the past and is using that fact to say this player will definitely do well this week at the same course, that that's a vast oversimplification of how things actually work.
Right, Yeah, it is.
In general, making strong claims about predicting performance in golf is a bad idea, just because there's so much there's so much randomness in golf as we like, even just last week, like like Bryson losing to Lang, right, I guess, so like obviously this is a very unexpected outcome, but those things happen in golf. So but to your point
about us changing our opinion, that's definitely true. I mean when we first started out on Twitter, we were more maybe rash and like combat of I guess to some debdeon and we took a pretty hard stance on course history saying it didn't matter. But I think we've definitely
softened that. There are just examples like Phil Augusta. He now has like he's played like so many rounds there, like like sixty seventy rounds, and he's averaging almost a shot better than than expected Augusta, which is a huge that's a huge difference, and it can't be you can't just wave that away by saying, oh, that's just randomness. That's pretty meaningful. So, yeah, course history, when you get big examples, can matter for sure.
Now, so course history can matter a little bit. There's there's a small adjustment that you can make for a player's course history when you're trying to predict their performance in a given week. It seems to me that course fit in a lot of ways makes up for some of the deficiencies of course history. Is that basically right?
Yeah, that's that's definitely right.
You get you sort of with So with course fit, what we're gonna look at is which skills a given course favors. So like the skills, we're gonna focus on our driving distance, driving accuracy and strokes and putting around the green approach, and by focusing on skills, you kind of bypass this sample size problem. So if we know through our fancy statistical methods that Augusta favors people hit
it further. And even if there's a guy who a first time round Augusta, so we have no course history date on him, but if we know he hits it above average distances, we can use course fit to make some judgments about how he's going to perform in that course. So yeah, by focusing on characteristics instead of specific players, which is what you do with course history, of course,
it can be a lot more meaningful. Yeah, because you don't have a we have a ton of data to estimate how how much driving distance is favored at Augusta, we can use every player to understand that.
Right. Yeah, So course fit just basically defined, is is kind of the degree to which a given golf course favors a specific skill set, Right, And those the skills that you measure are you know, maybe it can help me out here driving distance, driving accuracy, strokes gained, approach, strokes gained around the green, and strokes gained putting. Yeah, those are the kind of the five skills that you measure when you're assessing whether a course kind of fits
a particular kind of player. So has anything kind of jumped out and surprised you since you started looking at course fit.
I think the I mean, the main takeaways I think from course fit are like with the data that we have, it seems like the way that court, the way that PGA tour courses differ is along the dimensions of how
much they favored driving distance and driving accuracy. Certainly we can make intuitiveur arguments for why different courses favor putting or around the green, but I think the reality is, like the data, especially the putting, it's just so there's so much variance that it's hard to it's hard to really say whether or not a course favors putting more than the average, and that sort of reflected in the tool.
I think, I guess my takeaway like, I think I was.
A bit surprised. Maybe how still important? Again, there are a lot of caveats here, but how important driving accuracy is just because there is the narrative of I mean, this is how I'm looking at it. Somebody else could look at it and say, oh, I'm surprised how much that distance is the most favored skill, And just for
people who aren't looking at this thing. And basically the hierarchy is the way we have it is basically distance at the average course is the most particular power, along with approaching approach, and then it sort of goes.
I guess driving.
Accuracy, putting around the green are all pretty similar. Around the green is probably the smallest, but that's sort of the hierarchy. And then there's all sorts of interesting things that specific courses that you could get into.
Yeah, and we'll get into that a little bit in a minute here, But first thing, maybe you could just describe the course fit tool a little bit. I think people should go and see if the visualization is really simple and effective, and it would be probably hard to describe over the radio basically what it looks like, but could you just tell me basically what the course fit tool is and how you built it.
So the course fit tool, it's essentially what we're just trying to visualize, as you said earlier, the degree to which each course on the PGA Tour favors each skill, and so what we have on it is a basically have these shapes. Each shape represents the degree to which the course favors skills and so we have the average PGA Tour course on there, and then you can sort of easily see when we overlay another course on top of that whether or not that overlaid course favors distance
more than the average course. Same with accuracy, et cetera. And yeah, I don't know if we want to get into the details of how these numbers are actually calculated. Is it's somewhat complicated, but the intuitive all we're really doing with this tool is like, if you take driving distance, for example, we're gonna before each event let's play on the PGA Tour or each round, we're gonna have an estimate of every player's driving distance and every player's driving accuracy,
et cetera. And to estimating how much a course favors driving distance, we're basically just gonna compare two players, and this is important, two players who are similar in every dimension except distance.
So there's one.
Player, say that hits it ten years further than the other. So we're gonna take those two players, compare their stroke scheme in that round, and then you do that many times and eventually you're able to say, Okay, if you hit a ten yards further than somebody, that translates into whatever point three point four strokes gained around over that person holding everything else fixed.
And it's right.
You know, the visualization makes this very clear because this is an octagon. Basically, right, there are five skills, so there are there are five points on the octagon, and if one point is a little bit farther out than the others, and you can see quite quickly, okay, this this course tends to prefer that skill. And as you were indicating earlier, the most dynamic skills, the ones that seem to vary the most are driving distance and driving
accuracy from course to course. The others are fairly constant, though with some significant exceptions. So this is a it's a really interesting tool. I use it all the time. And to be clear, you know, most people probably use this as a predictive tool, right to say, hey, the course this week seems to be a good fit for player X, All go ahead and place a bet on player X. But I'm definitely not like the usual data golf user. I don't think I'm a lot more interested.
I don't bet. I'm really risk averse that way. So it's like I've never placed a bet on a golf tournament. But I'm more interested. The reason I look at this stuff is that I'm really interested in considering what something like this can tell us about courses on the PGA
Tour and the skills that those courses prioritize. And I know that you know, all this stuff is there's always the caveat that data is noisy, and we can't draw any super firm conclusions about any giving course or any given trend in PGA Tour course design and set up from it. But there are some really interesting things in here that I can't help but think can tell us something about course architecture as it relates to PGA Tour
player's performance. So, you know, just to start at one end of the spectrum, what does the course fit tool tell us about Bethpage Black, which is what's the host of the twenty nineteen PGA Championship. It's on one end of the spectrum for this tool.
Right, Yeah, so best Page according to our stuff, is it favors driving distance more than any other course that the PJ Tour has played, and I guess.
The last five years or so.
So yeah, it just means when Bryce in, when any player that goes to bed Page flack, and if they hit it above average distances, instead of that advantage being worth save point four strokes beth Page, it might be worth zero point six strokes or point seven strokes for
a round. So yeah, beth Page is the most extreme in that regard, and it's also yeah, pretty below average and how much it favors driving accuracy the thing also to note with like obviously guys who hit it far, there's almost there's partly a mechanical or partially a mechanical relationship between distance and accuracy, Like as you hit it further, it's just harder to hit more fairways. So there is
a negative, a strong negative correlation between those two. So on our tool, like Bryson, like he ghets at above average distances, so we're getting it, he's getting a big bump for that, But he also is less accurate than average, so he's getting a and he's getting a bump for that as well at beth Page, because essentially you can read it as beth Page penalized accuracy less than the average.
Was, which is not a great way to think about it, maybe.
Because obviously there was a high penalty to miss fairway at beth Page, but all the tool is saying is that the benefit of being a PGA to a golfer who is five percent more accurate than average, that advantage was lessened at that page for whatever reason.
We can speculate on that.
Yeah, I know you don't love speculating about this kind of stuff. I might try to push you out of your comfort zone a few times.
You know.
Part of what makes this suggestion by the course fit tool about what skills Bethpage Black prioritizes, what makes it compelling is that it's kind of counterintuitive. You know. You look at beth Page as it was set up for the twenty nineteen PGA Championship, and you see really high rough you see really narrow fairways, and the assumption that most people would make is that that emphasizes accuracy. You
have to hit the fairway or you're in trouble. But in fact, it seemed to be the opposite, where longer
players had a distinct advantage. And obviously we can't draw strong conclusions from one or two particular results, but it seemed like it really seemed like Dustin Johnson and Brooks Kopka were the only players who had any chance of winning that week, and so is that something that caught your eye as well, that that there was a kind of counterintuitive result from the course fit tool based on what you would assume out of a course that was set up the way Beth Page was.
Well, first, I would say the results I think are this is definitely something that's extreme. Like there's certainly a lot of signal there. There's no doubt that the I think probably there's three events in here, maybe because there's the maybe yeah, the two US opens and then that page had.
A that's right, it's not not just the PGA Championship. Beth Page also had a Northern Trust or something. Yeah, so there is.
Enough data to say to say things, and yeah, I think it was I think it was pretty surprising. Like generally, when you do things like this that are somewhat complicated statistically, you want like eighty five percent of your results to match up with intuition, so then you're like, okay, I'm not doing something completely insane here, and then you have fifteen or twenty percent where it's like, oh, that's like that's countertuitive and interesting.
So yeah, so Beth Page, it.
Does go this way where well, so just to bring up another example that didn't go this way. Firestone has fire Stones, not beth Page. But it's a long course that has narrow fairways with reasonably long rough and it has it favors distance, but it also favors accuracy, which is interesting.
That what that just to reiterate what I said earlier.
Like that means Firestone it's above average in terms of how much it favors distance, holding accuracy constant, but then also holding distance constant constant.
It's still important to hit fairways.
There's an above average benefit at Firestone for that as well. So I bring that up just because that to me is like more yeah, more into if I would have thought beth Page would have been like that. So I don't know what's different about beth Page. Maybe because it was so it was so long, like beth Pages, it was an extreme course like g it's longer than Firestone.
Right. Another thing to mention about beth Page versus Firestone is that beth Page has you know, a lot more movement in its land and the greens are sort of sustained actually elevated, which would seem to you know, give a reward to players who end up closer to the green after their first shot on a on a par four or par five, because you have to get the ball all the way to the green. But I don't know, you know, there's more ground contour at Firestone than people
give it credit for. And I'm not sure you can exactly run shots up at Firestone necessarily, but there might be might be something there, And I mean, I guess these are these are factors. This is where we get into speculation.
Right, Yeah, No, I think it's important to keep in mind, like driving distance is not just when we're looking at a player who bombs it. It's true that he hits it further off at team, but then it's also true that it's just a powerful player in general. So from the rough he's hitting wedges, and like you can think about with driving distance as an app as a skill is going to be beneficial not.
Just on t shots but also on approach shots.
So yeah, that's and at that stage maybe that was playing a role as well.
That's a great point. So on on the other end of the spectrum, we have this week's PGA Tour revenue El Kama Leone Golf Club, host of the Mya cob A Golf Classic. This is on the other end of the spectrum from Bethpage. Black tell me about the craziness of the data that this course produces.
Not only is it the most extreme for in how much it rewards driving accuracy, it's also just of any attribute because again on the default view on the course fit stuff, you can compare across attributes, like you can say if the further out of dot is that means just in an absolute sense, that skill is getting rewarded more strokes gain than the other dots. So so driving accuracy at Alchemeleone is actually rewarded as much as driving.
Distances at beth Page.
And the reason it's not that's notable is because at the average course accuracy is rewarded less and distance, So that means at al chime Leone it's more of an outlier even than beth Page.
Yeah, it's pretty crazy.
And like I know, when we looking at like our predictive model stuff, when we're actually calculating how much we're going to adjust players skill levels, because that's how we actually put to use the course fits stuff. We basically have a player's baseline and then we're going to say, okay, based off course fit we're gonna move off of that baseline by whatever.
Point two strokes or something.
At alchemillion, there's adjustments that are like a stroke, which is insane. So we're moving a player's skill level a shot. I mean, that's like the most extreme guys. But just for context, that's the difference between like the fiftieth ranked player in the world in our rankings and like maybe the two hundredth or something like that. Like, it's a huge I got one shot around is obviously a massive difference. So that to be driven by course fit is pretty crazy.
So yeah, this this week's course is the biggest outlier.
It exerts a huge influence on the results in a certain way. And I guess one way to put it is that if you have a clear best player in the field, you know, somebody whose past performance is really really awesome, a lot better than everybody else in the field, going into the Maacoba Golf Classic, they are a little bit less likely to win there than usual. Would that be fair to say?
It would be fair to say as long as they possess the typical top ranked player skill set Brendan Todd ever to be the best player in the world, then we might actually predict him to be better.
But in general, in general, Yeah.
Your statement is true, Like this week, the skilled attribution is going to be compressed just because it tends to be the case that the best players do hit it far. And so the best players on average will be getting a negative bump this week, and worst players on average will be getting a positive.
Yeah, bringing everybody together.
So I guess a way to put is, if Rory were playing this week, then you would give him less of a positive positive adjustment than usual. You would you would be less strong on the idea that he might win than you usually would be at the at the average PGA tour venue.
That's the right way to say.
Yeah, I would probably say relative to his baseline, like it's average still level, we're giving him a negative bump. But what you're saying makes makes sense to It's we're still we still like Rory. We just like him less than normal.
He's still going to be good. It's not like he's all of a sudden, you know, yeah, the worst player in the field or something. Yeah, I mean I find that interesting. Does that mean does that mean that there's just that randomness kind of plays more of a role at a venue like this week's I.
Wouldn't say I wouldn't say that there are some courses where that does seem to be true, but I think it's just the issue is just when I when we say high skill gear, what we mean like high skill on the PGA Tour these days, that generally means you hit it far, just because that's the way that skills are rewarded. So when you go to Camillone, it's true that higher skilled players are getting a negative bump, but
it's it's not necessarily due to randomness. It's just because driving accuracy is now way more important because if you miss a fairway at that course, you're generally taking a drop.
So yeah, I wouldn't say it's randomous.
I would say it's it's a different skill set that's been tested, and so that brings the skill distribution relative to like the typical tour event closer together.
You mentioned what I think is one of the key factors here, and that's that if you miss a fair way at this course, then you're basically in the bush, like it's a lost ball, and so there's a really heavy penalty for missing fairways. But at the same time, you know, I think Sometimes people get confused here because then they say, well, doesn't that mean that penal setups where there's where there's a clear kind of immedia penalty for missing a fairway, doesn't that mean that those would
favor accuracy above all else. But of course we just talked about Beth Page, where there was a real penalty for missing a fairway, but distance tended to be more predictive there. And so how would you reconcile those two? Is it because they're like different levels of penalty, right that you know that taking a drop is a lot different from hitting it out of the rough.
It could be that it could just be a matter of degree, although I think it's more than that. I think it's I mean, maybe it's also has to do with the fact that at death Page, maybe it didn't matter so much if you were five years off the fairway or if you were twenty yards off. I don't even I don't know if that statement is true. It still matters a bit, but alchimately, oh maybe it matters a bit more like the fact that you're two yards
off the fairway. I don't know how many yards you have to deal with there, two yards or three yards, and then you're in the junk.
Maybe that that's part of it, also part of the people what we were.
Saying earlier, where at best page, when you miss a fairway, you are hacking it out of deep rough where bombers still have an advantage, where whereas at UG mileone you're you're just dropping it in pretty light rough. I don't there's not sick rough at this course, so maybe that advantage is gone for distance, But a lot of it's got to just be the fact that this week's course is shorter.
Yeah, I mean, the causes are really hard to identify, but you know, I want to expand this out a little bit, Matt, that el kamal Leone is part of a set of courses on the PGA Tour that I've kind of discovered through the course fit tool on your website that I like to call the web Tour Web with two b's. You know, they're just courses where web Simpson seems to seems to perform particularly well, and players of Web Simson's ILK as in, you know, really good
in every skill except for driving distance. You may also say that Brendan Todd is kind of along this model as well, at least within the past year and a half or so. So in addition to El kamala ayone, you have Harbor Town Golf Links, you have Wyley Country Club, you have TPC, Sawgrass, you have Colonial, and then maybe to a slightly lesser degree, you have Sedgefield Innisbrook, Copperhead, the side of the Valspar Championship. Sedgefield is the side
of the Windham Championship. To an extent, Sea Island, which was last week's venue for the RSM Classic, maybe Sherwood, one time host of the Zozo this year, maybe Merefield Village. I mean those kind of got less and less a part of the Web tour as I went along, but especially that kind of top group of courses Harbor Town, Wyley, TPC, Sawgrass, Colonial,
El kamala Ayone. At all of these courses you see an unusual emphasis on driving accuracy and an unusual d emphasis on driving distance, as in the skill of driving distance is less predictive of success at these courses than average on the PGA tour, and the skill of driving accuracy tends to be more predictive. El kama aone is just the most extreme example of this. But have you
noticed the Web tour before. Have you noticed this class of courses on the PGA tour and if so, what do you what do you make of it.
I'm not gonna say I would have come across this cluster of courses before looking at before doing this analysis, although at the same time I think it does. It does make sense. Like I think most of these courses are are shorter, and I mean they're also there are also differently a few of them. I think of Sedgefield and east Lake is also one that rewards accuracy more like those are pretty like Eastlake is pretty brutal rough Sedgfield. I'm not sure exactly why it favors. It's just it's
a shorter course rough pretty tough there. But I mean again, but as we have at best day, it's not necessarily a simple relationship like thicker rough does not necessarily mean it favors favors accuracy, although I do think the combination of thick rough with a shorter course maybe does. It might just be the interaction of really thick rough with a really long course that gets you that effect where
it distance but not accuracy. Yeah, And Harbordtown, I think is just the course where it just takes driver out of your hands on a lot of holes, which which is fine.
I mean, I think it's honestly what I thought after the when.
Golf came back after the three month COVID break and Bryson was doing his thing and getting all that attention, Like, I was kind of struck by how many courses It seemed like the first few weeks he was playing, he was playing golf courses that didn't really favor bombers that much.
Like I was sort of struck by it.
I was thinking, Yeah, like, obviously we all know distance, like the best players in the game bombit, and it's clearly a huge advantage.
But there are still courses on the PGA Tour, like the.
Travelers River Highlands that it's not that it's still it's always an advantage to hit it far, but they are there are courses on tour that, yeah, favor accuracy still, so that I remember being struck by that when the golf came back.
Yeah, and Detroit Golf Club where Bryson won, is not a course that necessarily strongly prefers distance in the way that we see a lot of PGA Tour courses do. Am I right about that?
Yeah, I mean, according to tool, that's true. I feel like I've heard people say, I mean, I'm not sure who I should trust. I guess I've heard people say, like I think I've heard people say before that event that, oh, Bryson's gonna make a mockery of this, of this layout.
And I don't think I watched a shot at that event, so I.
Can't The most exciting thing that happened was Bryson yelling at a cameraman. But I think what people were saying is that he's going to bombit past the fairway bunkers that were put in during the renovation that were put out there to kind of, like, you know, give give tour players something to think about, and he just was
longer than them. I don't know. I mean, there's a lot of these factors that tend to come into play at these courses, and obviously a player like Bryson can do well anywhere, you know, he's he's got other skills. But you mentioned some of the characteristics that these courses
have in common. These courses that actually succeed in prioritizing driving accuracy, and that's something that so many people are interested in seeing come back to the PGA Tour in a bigger way, you know, more of an emphasis on driving accuracy. They're they're trying to figure out how in the heck do we do this? Do we roll back equipment, do we do things to courses? But it seems to me that this set of courses gives a pretty good
example of how to do that. But it doesn't necessarily explain why these courses behave in this way, you know, super clearly. But one thing that I see in common is just course length. A lot of these are shorter courses. So do you think that's like a strong factor.
Yeah, a lot of times you have we can, we can, we can come up with these more complex theories about how courses will favor certain skill sets. But yeah, often those things are hard to like tease out of the data.
But yeah, something like course length. I think it's pretty clear in the data that longer courses, shocking, yeah, favor driving distance, and and yeah, these shorter courses, I mean, I don't know that people would be particularly happy with this set of courses, being like if this is the template that it's like, oh, yeah, these are you want this, these skills to be rewarded in this way, like, Okay, that's your course.
It's like, I don't know if people because because Sawgas is.
Sort of a some of these courses I think of as kind of random or not not rewarding skill that much. But that might also be like a bias because we're used to seeing we're just not like Rory won't play as well at that course, and we're used to thinking of Rory as the best player in the world.
So you have to like.
Maybe come to terms with the reason Rory's the best play in the world is because his language has an advantage.
But the one course, and I'm surprised.
I'm sure that Friday did write some articles about it the course a few weeks back where it was short and it was it played incredibly tough.
The Houston Open, the Sure Memorial Park, Memorial Park, Yeah, I'll.
Be I'll be fascinated to see how that comes out.
That wasn't particularly penal off the tee, right, that was just it was around the greens that were giving people a lot of trouble.
Very much around the greens. Yeah, I mean, I think that the more tournaments that are played there, the better of an idea we're going to get for it. But certainly, you know, since the greens were fairly new, they were quite firm, and so I'm not sure if in future years it'll be the same kind of dynamic. And you know, there are some pros who are frustrated by it, so
that might play into future setups there as well. But I think it is a good example of how, you know, narrow courses with with clear penalties off the tee are not necessarily the only solution, and I think, you know, Wileye probably demonstrates that as well. I don't think of Wildleye as being particularly penal off the tee it. You know, there are some palm trees out there, but it's it's fairly wide open. Other than that, it's just a shorter course.
Yeah.
Wiley is an interesting example though, where if you look at its diagram on the course fit page, it sort of is compressed in every respect, so it doesn't rewards driving distance class and driving accuracy lass. Oh and also slightly approached.
Which I mean again it's hard to.
Say exactly what that means, but in theory that might not be a great thing because it means that you're not rewarding skill as much at Wiley as another course,
but again that's only like part of the equation. I think when we because like, realistically, if you want to reward skin as much as possible in the PGA Tour, you probably would want a course where where greens are super soft and it's just target golf, because yeah, it does take like when there's no randomness in the sense that there's no bounces that can make a a good shot turned into.
A bad shot.
Then yeah, like a soft course like that probably does reward skill the most. But as viewers, like, nobody wants to watch that, so there's other there's other considerations, and so I think, yeah, while I it does level the playing field, it does seem like there's some element of randomness there, but like that could be fine, depends on what your preferences are.
That's maybe not a bad thing. Yeah, I mean, there's an assumption that what we really want are courses that reward skill in a clear way. But I think that once we take that to the logical extreme, what's the course that rewards skill the most that we might not like what's on the other side of that door.
Yeah, maybe maybe that brings us to like top golf, and you just get tour pros hitting shots into baskets, and that's that's the most that's the most skilled thing for rewarding approach shots the.
Most reliable test of skill. Yeah, totally. There's all the other factors have have kind of been leveled. So you know, speaking of course, is that that seem to have high variance is another way to put what we're talking about. You did a really good deep dive earlier this year, I believe on TPC Sawgrass. Tell me about what you found there. What are some of your thoughts about TPC Sawgrass as a course that you know tests certain skills or doesn't test certain skills from PGA tow or players.
Yeah, Sawgrass is a course that rewards accuracy more so than compared to the other four attributes, I think, and any other thing.
About saw Grass is and this isn't really.
Reflected in the course fit tool, is that at Sawgrass, not only is it the case that if you're one shot better than the average player at the average PG two, of course when you go to Sawgrass, you're only going to be point seven or point eight shots better. So it's it is reducing the advantage that skilled players have. And then it's also it's also adding in variants in the sense that Sawgrass is just a course where if the same player plays there one hundred times the variants
in their score. So they're going to average seventy, let's say, but they're gonna sometimes she's seventy five, sometimes.
She's sixty five.
That variance is higher at Sawgrass as well, and that's sort of a that's a separate thing to relate this to whyl Whyli is actually a course where even though that skill advantage gets reduced, it's it's a low variance course. So if the same player plays there a bunch, it's going to be a tighter bound around their average score. It's kind of a tricky thing to think through, but the upshot for saw Gas is that, yeah, it's a
it's a very random course. It does I think it does reward accuracy more than average, but that's about it. And it also has this added element of just variance, which I think everybody who watches it what you agree with that, like, there's there are there is potential for big numbers in that course, and there is more randomness in that respect.
So it's Yeah, the takeaway.
With Sawgas is that it's just not only is performance unpredictable, there's also just more variance in general.
And so yeah, I wonder why that. And this is getting into the architecture questions that you know, data is not necessarily gonna tell us clear answers on but I I wonder why that. What design characteristics of TPC sawgrass are at play here? Could it be the really severe penalties at certain places in the course. Could it be the tininess of the targets on the greens? Do you
think these things could have have an influence? I mean not just the greens themselves, but like the sections of the greens where pins might be are super super small.
I think that's it.
I think so yeah, I think you can have variance, and variants can be a good thing or it can be a bad I mean, if you have at sawgrass, like if you have a shot that if it misses, if a player misses this spot by two yards, it ends up rolling down a hill, and it's super penalizing compared to a course that's soft. You miss your spot by two years, you just have a pieto that's two years longer.
It's not a big deal.
So that that can certainly add variance, and that would probably be considered good varians because it's it's making the margin between a good shot and or a marginal shot and a good shot is now that that's creating a bigger difference scores, which is probably what we want. And yeah, I think Sawgrass does that to something to be obviously, seventeen is like a good example of probably variants you might not want.
It's hard to.
Say, like if somebody's playing really well and goes to seventeen and makes a seven, their tournament is all but over because of that, and that's sort of that that's another big source of variance.
Shout out Sergio.
Yeah, and in general, I think it's just because Sawgrass is sagas generally plays super firm too. And like there's speaking of surgery, there's that year where you like eight putted hole or eight the part three and like you just that's probably bad variance.
I think we would agree.
That that's bad rans The greens were just they sort of maybe put them over the edge. And when I think of Sawgas, I just think of it might not necessarily be rewarding skill as much as we like, but it's maybe it is it's hard to say, but it's very entertaining golf to watch.
Absolutely, yeah, I mean, I think it's good to be clear that, you know, when we're asking questions about how PGA Tour venues can reward or not reward different skills and talk about the kind of variety that we might want to see in PGA Tour venues, that's not necessarily
a commentary about good architecture. The question of what makes good architecture and what makes a PGA Tour venue that should be part of the rotation, those are two separate questions, right, And I think that finding that balance between good architecture for the masses and something compelling to watch in a PGA Tour event, that's really something that these courses have to think about, right because they're they're just hosting a
PGA Tour event one week of the year. The rest of the rest of the year, they're you know, open to the membership or open to the public. But personally, you know, what, what I really want to see, you know, I love, you know, these questions about good architecture I'm very interested in and I think that's the highest priority for a golf course. But I'm also somebody. I'm also a PGA Tour fan, and I want to see as
much variety in these venues as possible. I want to see, you know, many different types of courses that test different skills. And that's why I get excited when I see a course like Harbor Town, because it's just not the usual PGA Tour venue. I mean, you're you're a golf fan, do you do you kind of feel the same way.
Yeah? Yeah, I mean I feel the same way. I yeah.
I think there's a clear trade off to be made between setting up a course of rewards skill and setting up a course to maybe not a clear trade up, but there is a trade off between rewarding skill and making the golf entertaining to watch. Like, I really can't watch the PGA Tour play when they play a course where it's soft and it's a diverty fest, it's just very uninteresting to.
Watch for me. Like, I don't necessarily.
Appreciate architecture per se, but I mean I like watching golfers play firm golf courses just because it's tough.
It's just testing players. I like seeing.
I like I like see courses where there's like penalties off the tee, just because I think especially down the stretch, Like just from my own experiences playing golf, I like seeing it's compelling to watch players try and perform under the gun, and hitting a good drive under pressure is I think one of the harder things to do in golf. So I mean that to me is interesting. But yeah, in general, variety I think, is you need it on the PGA Tour like I don't, And I think, unfortunately,
it'd be one thing if there's no variety. But they had settled on firm golf courses as the thing they're going to focus on, but they've settled on on softer birdy fest at least for non majors and non well pretty much just non majors, And yeah, I think that's pretty uninteresting golf for a lot of fans.
And it changes the complexion of the top the set of top players in the world, because presumably if there were more courses on the PGA Tour that had some of these Web Tour characteristics that we've been talking about Web with two b's, there would be different players in the top ten. I mean, it would just be the skill set of the top players in the world might be a little more varied.
Yeah, I think it would be more varied.
But then and then at some point we're just going to get to like this subjective question of how much which skills should be rewarded, because I still like Dustin Johnson. Be able to hit the ball as far as he does is and then obviously an incredible skill that should be rewarded. And then the question is just how much, because yeah, Web's also skilled in various ways. So it's it's a tough if the fine line sort of you have to figure out what exactly your priorities are.
I guess absolutely, and and DJ, it should be said as well as Rory have significant skills across the board, whereas somebody like Web, you know, you could say that he just doesn't really have the skill of driving distance at an elite level, and so and he gets docked for that, and some might say that that's exactly right, you know. So somebody like DJ, his balanced game, which includes that extraordinary ability with the driver, is certainly well rewarded and and should be. I think that's a good
place to wrap up our discussion. Of course, fit you also, I should mention, did a deep dive into augusta National recently for your website before the Masters, and we're not going to go into that in depth, but I just wanted to mention it and just recommend that people go and check that out because I think it's absolutely fascinating. So there's just something that that people can check out
that we won't discuss here. That's a reason to go to datagolf dot com and do some re So I just wanted to to wrap up with it with a few random questions. These don't need to be long in depth answers, but a few points of curiosity outside of this course fit discussion that we've been having so kind of a lightning round. When a player is having a breakout season, are there any key characteristics that you see, like things that change in their game?
Generally, if a player is playing well recently, and all I know is that they're playing well, I haven't looked at deeper into the data, it's probably due to the putting around the green stuff, just because that's the stuff that varies a lot round around. But if they're truly having a breakout season, one where our model is actually gonna say, okay, this guy's being elevated to a new level. I mean I think it's usually it's usually driven by either off the t stuff or approach stuff.
Yeah.
The easiest way is, like for Telly, like somebody who's just gained ten yards and that matters. That's an easy way to do it.
Who is the best player on the PGA Tour without a major? H? And why is it? John Rahm?
Oh right, wrong?
Yeah, obviously is wrong. Yeah, obviously obviously it's wrong. Yeah, I guess I was thinking, do I want to say, who's like, who deserves one at this point?
And not like the best career, not the best career, yeah, but like more you know, who is currently you know, based on their recent performance? What would it clearly be wrong? I mean I don't know, is that the clear answer to the question.
I mean yeah, to me as someone who's looking at I don't really think of players who perform well in majors versus.
Those who don't. I just think of golfers who are good versus golfers who are not.
So I'm obviously rom has not had that many chances to win majors, or maybe any good chances. But he's rom Is We've been saying wrong. Ram has been underrated he's probably not underrated anymore, given that he's the number two is replaying in the world. But he's just he started his careers as close to Tiger in the last twenty years as anybody else has in terms of just strokes gained and number of top fives, and he just
didn't have a transition period. He just came came out on the PJ Tour and was already playing like the top five planet world. So he's going to get a major soon.
Yeah, this game is unbelievable. Yeah, it can't it can't belong. People are very high right now in Dustin Johnson's ability to win multiple additional majors. What would you say to that all that excitement, Yeah.
I would say the best player in the world has Maybe you don't even have to trust us. You can trust Betty Markets, trust anything. They have about a nine or ten percent chance, if they're clearly the best player in the world, a nine or ten percent chance to win a major. So that means if a guy plays, if DJ plays ten majors and he's aging, that means
we've expect him to win one. So still it's like you can't assign majors to too many golfers because there's only four of them each year, so it's it's always unlikely to win majors.
Yeah, yeah, there's there would be a good hobby in listening to golf podcasts and tabulating the amount of future majors that are given out.
Yeah yeah, we'd be in the thousands.
Yeah, exactly. So you know you're on golf Twitter on occasion. So I'm sure you've seen how people use statistics to try to win arguments. What is the what was common mistake that you see when people use stats in this way.
I think the biggest one is confusing correlation with causation. I have an economics background, so that's I am very sensitive to correlation versus causation. And I mean a good example, I saw AUGUSTA somebody posted just on number three the scoring average for players who had hit it more than three hundred yards on the whole versus the scoring average
for players who had hit less than two forty. And the conclusion they were trying to say was, oh, look, you should be hitting driver here to go for the green because the scoring average is better for the guys who hit it further. But that's the flawed for many reasons, and the biggest one is that the guys who are hitting at two forty on that whole were like Larry Mize, Samuel Aisle, among others, and the guys who are hitting at three hundred, like Bryce and Rory the best players
in the world. So that's correlation versus causation because you're picking up we want to know the causal effect of hitting driver on that hole versus not hitting driver. But when you do that analysis, you're picking up all sorts of things, like the fact that Rory is just a better golfer in every way than my So you're, yeah, that's Twitter is not a good space for statistical arguments.
You're you're comparing Rory McElroy to a retired golf pro yah, yeah, which which is meaningless. Yeah yeah, all right. So you know, this is maybe a big question, but I'm just curious, like when you're looking at the future of data golf and what kinds of new questions you want to answer. What do you think the new frontiers are in golf stats? What do you think the questions are that that we haven't answered yet, and maybe we could find methods to answer them.
Yeah, I think I actually have a good answer, because we're hopefully for next year we're going to be we're going to have official access to shot link data. We're the actual shot level stuff, or sort of, because right now everything on our site is just round level stuff just because you're not allowed to without an official license, you can't display shot link data.
So if that hurdles passed.
I think yeah, I think the next step for golf analysis is just building on Brody's stuff with strokes gained and trying to if we really trying basically, because right now in the PGA Tour, the way strokes gain is calculated is just you have a generic baseline function that says, from this distance and this lie, we expect to a player to take this many shots and then you get
strokes gain from that. What we would like to do is have a model that and it would be difficult, but it would have a different baseline function for every hole.
So like, maybe there's a hole where going this would be very subtle and potentially not possible, but maybe there's a hole where when you're two hundred yards out in the right side of the fairway, that's actually significantly worse than being two hundre yards out in the left side of the fairway and right now with strokes gained, that's treated the same.
But if you had again it would be difficult.
But you couldn't theory be able to say, no, I guess from this angle or from this spot, players take whatever three point four shots to get down from this spot it's three point one. And I think once you have that whole specific baseline function, there's all sorts of things. The course, your your assessment of course fit could totally change. You could drill down to a whole level. You could say yeah, you could say a lot of a lot of interesting things. So I think that's sort of the
next direction. That then that has implications for prediction and golf and for yeah, for for a lot of things.
So I think that's the next one.
All right, Well, Matt, good luck with your with your future work. And thank you so much for talking to me today.
Yeah, thanks a lot for having me on. Enjoyed it.
Hmmm.