Is Golf Course Architecture Just a Science We Can't Yet Quantify?

[abmack]

Paul

IMO, absolutely not.  Golf courses aren't about testing data.  What would it mean to apply the scientific method to golf architecture?  A bunch of data is measured to what purpose?  Can the data improve a golf course in aspects other than the nuts n' bolts stuff like drainage etc?  In terms of how a design interfaces with a golfer,  how does one conduct a fair test in an experiment with so many unknown variables at play (ie the weather)?  Can we call an archie a scientist?  Does anybody think of a building archie as a scientist?  It could be argued that in the right hands, course design is better off being less scientific...hence the push for "random" features.    

Ciao

First, economics is a science. I am an undergraduate studying economics at Harvard and have been exposed to many researchers and have conducted my own research. Just like any other social scientists, economists conduct experiments and use data to empirically evaluate relevant questions. The scientific method is key to developing economic understandings.

That the predictions made by economists are sometimes (often) wrong is not evidence of that the field is not legitimately scientific. As many of my professors often admit, economics is a new endeavor; academic methods and consensus is rapidly evolving. When serious economists make predictions and find them to be wrong, they then go back and examine how the actual outcomes differ from the projections. They look at the phenomena and the variables which could have contributed to the deviation. Being "wrong" is a good thing and contributes predictions being better in the future.

Golf Course Architecture
This question which you raise is a very important one. I have long thought that standard economic models have applicability to golf. In the game's most rational form, the player's goal is to minimize his score. Doing so requires that his golf course strategy be consistent with that goal. The psychologists Daniel Kahneman*** and Amos Tversky spawned a body of economic research which focuses on behavior and decision making under risk which I believe can be used to assess golf course decision making.

The first step in conducting this research would be to review the existing literature on behavior and decisions under risk as well as mathematical risk modeling and determine which papers and methods have applicability to golf.

Second, you would have to create definitions and a standard model (fully rational) of golf course strategy and behavior. The full development of this model is likely to be central to any rigorous body of research on golf course behavior.

Third, you would have to use real data on how golfers actually behave in order to test these models. The best existing data is ShotLink data which the PGATOUR has gathered over the past 10+ years.

I can think of hundreds of possible avenues to explore. BUT, the amount of work require to develop and test a framework for golf behavior is MIND NUMBING! Especially the data analysis part. I have experience doing empirical research. My past work has taught me what would be necessary in order to clean and adapt a dataset like ShotLink so that it can be analyzed using standard econometric methods. It is hard for me to overstate how tedious and time consuming this sort of undertaking would be. There are also conceptual hurdles such as accounting for omitted variables and other biases. These endeavors are possible, however.

With respect to the golf course architecture, after developing this framework and the models, one could gain very helpful empirical insights into the design aspects which distinguish good holes. I don't think that it is controversial to assert that there are many different kinds of holes (eg. long par 3s, short par 4s, etc...). Conceptually, it should be possible to show that the best holes within each category exhibit similar risk profiles. This is not to say that they are the same. Instead, what I am saying is that risk-reward-punishment aspects of a type of golf hole relative to its difficulty, yardage, and number of potential strategies are similar among the best holes of each category.

If you have the goal of categorizing the risk/strategic profiles of all holes in golf, I think that you would see that when normalized, there are a lot of common traits. If you disagree with this, think about the fact there is so much heterogeneity in golf courses around the world and there is no accepted standard by which people do normalizations.

After I finish my thesis in march, I am going to work on developing some of the ideas which I have discussed here. I have a professor who has promised to help me think through some of the conceptual difficulties associated with this project. Please let me know what you think.



***Daniel Kahneman won the 2002 Nobel Prize in economics for this work.

[Sven Nilsen]

Andrew:

Supposedly the Freakonomics guys are working on a book based on golf.  A couple of years ago they were recruiting folks to participate in different studies to examine behavior of play on the golf course.

It'll be interesting to see what kind of data set they came up with, and what conclusions they draw from it.

Sven

[abmack]

Andrew:

Supposedly the Freakonomics guys are working on a book based on golf.  A couple of years ago they were recruiting folks to participate in different studies to examine behavior of play on the golf course.

It'll be interesting to see what kind of data set they came up with, and what conclusions they draw from it.

Sven

Really, I have been in touch with Steve Levitt about something else. I should get in touch with him re: golf. Also, his colleague at UChicago, Richard Thaler, is a golf enthusiast who I intend to reach out to.

[Jonathan Mallard]

Mike,

I was discussing the concept of soil stabilization techniques - specifically if they were considered for the 4th green at Pacific Dunes. There is a rather large, and expected to grow, undercutting of the cliff. Tom answered publicly that they didn't consider bringing in engineers.

Here's the exchange:

I don't think your participation discourages honest discussion of your work. I also have noted that you won't always respond to questions regarding some aspects of your work. You may not always see them, I get that, but for some of the questions I've asked that are more technical in nature, I don't see much discussion of.

Here's a new topic as an example: The 4th green at Pacific Dunes.

When I played there, I looked to the right of the green, and I thought that there was some undercutting via natural processes or the mechanical properties of the soil itself. I think that left alone for a few more years, a decent size portion of the collar and perhaps the green may fall off to the beach below.

Here's my question: How much does GCA work take into account engineering principles? If a site is analyzed before a design is completed would the tendency be to change the routing, or to work on a solution for the issues present to incorporate the preferred routing.

For the 4th at PD, are there any plans to mitigate the undercutting?

Jonathan:

Well, that's a topic most professionals wouldn't want to touch, because there are potential liability issues and the whole subject raises the possibility of "malpractice," if there is such a thing in golf architecture. 

However, I'll bite.

Most architects would never build a green that close to the cliff edge, even if allowed.  To be safe, they would build it 25 or 50 feet inland, where it didn't come into play so dramatically.  It would be a good hole but would become better many years later, when the edge had eroded a bit and had evolved to be really in play.  Essentially, we decided not to wait.  I told Mr. Keiser that eventually the green would erode, and at that point we would have to build a temporary green to the left of the dune, and carve out more of the dune on the right side, to rebuild the green much like it is today.  That was a dozen years ago, and so far the green has not eroded nearly as much as any of us anticipated it might ... [knock wood] ... though someday it will.

We did not pursue an "engineering" solution for the site.  The green sits on a cliff top 100 feet high, with various layers of sand and sandstone underneath.  I'm not sure anyone would let you work at the very edge, if you asked.

I was familiar with the situation and prepared to take the approach we did, because the site for that golf hole reminded me immediately of the 7th hole at Ballybunion, where they went through exactly the process I'm describing above, after part of the original green fell away in the 1970's.  I don't know if Molly Gourlay planned it that way, or not.  I do know that the replacement green they've built is nowhere near as natural-looking as others on the course, because they are trying to be "smart" and leave a buffer for some future erosion.

http://www.golfclubatlas.com/forum/index.php/topic,58131.msg1362950.html#msg1362950

[Paul Gray]

Andrew,

Economics is not YET a science. Do not believe all you are taught in econometrics.

Nonetheless, your grasp of the concept of the science of architecture does clearly indicate that you appreciate the premise of the question.

Paul Gray
Bsc Economics

[Russ Arbuthnot]

Andrew Mack, nice post! Sounds like you are quite interested in applying economic concepts to golf, and I find it quite fascinating. I enjoyed reading Kanneman's latest book, and I'm also a big fan of N.N. Taleb and Mandelbrot.

I would highly recommend that you check out work done by Rich Hunt in his yearly Golf Synopsis. Rich, a statistician, has been developing ideas similar to some of those you mention for quite a few years now, and he gets better every year. He actually does pour through all that ShotLink data.

Of course, there is also Mark Broady, creator of Strokes Gained, and author of Every Shot Counts. It's likely that you may have heard of both of these guys, but if you haven't, I thought I would mention them.

To get back on-topic, GCA might be a science that we can't yet quantify, but I think it will be a very very very long time before we are able to do so. And I actually pretty happy about that.

[Ian Andrew]

I'm waiting to hear the screaming when GC Architects are deemed to be under the jurisdiction of a Landscape Architecture board, and subjected to their requirements to practice in a given state. Don't laugh, it will happen eventually. 50 states, 50 different licenses required!

Jonathan,

That's a very complex subject that I'm all too familiar with.
About five years ago my wife (what she does professionally) went through all the statutes for me.

In simplest terms (because I don't want to have an open ongoing discussion on this subject):

Some States clearly list golf architecture as a completely different profession than Landscape Architecture in their statutes.
It's been challenged in court, the golf architects won, so you won't see your premise play out.

In a couple of States certain memberships are listed as an exception and write to practice golf architecture.

In some instances, the question of licensing is decidedly "grey"
That is where the Landscape Architecture Boards have made attempts to have it declared Landscape Architecture
To this date it is not listed as Landscape Architecture in any statutes - "but what it is" is still up to interpretation
Interestingly, Civil Engineering License in these States is just as valid as a Landscape Architecture for the same works.

So in a nutshell, it won't happen.

BTW, between a third and a half of the golf architects I know are registered Landscape Architects in the US

[Sean_A]

Andrew,

Economics is not YET a science. Do not believe all you are taught in econometrics.

Nonetheless, your grasp of the concept of the science of architecture does clearly indicate that you appreciate the premise of the question.

Paul Gray
Bsc Economics

Paul

In the case of economics we are waiting for the science to be stronger...I think much stronger (in other words the assuptions have to be drastically reduced) before we can say its a proper science.  Maybe that day will come, but I fear if more intricate models for human behaviour and policy are developed nobody will understand what an econ junkie is talking about! 

Not sure how architecture fits in though.  How can the artistic side of the coin be quantified and codified?  Perhaps more importantly, will that help architecture in any way even if it is possible to do so?

Ciao

[Paul Gray]

Andrew,

Economics is not YET a science. Do not believe all you are taught in econometrics.

Nonetheless, your grasp of the concept of the science of architecture does clearly indicate that you appreciate the premise of the question.

Paul Gray
Bsc Economics

Paul

Firstly, sorry to go all Mucci.

In the case of economics we are waiting for the science to be stronger...I think much stronger (in other words the assuptions have to be drastically reduced) before we can say its a proper science.  Maybe that day will come, but I fear if more intricate models for human behaviour and policy are developed nobody will understand what an econ junkie is talking about!  

The fact that it ISN'T a science is exactly what I was saying to Andrew. If a degree in economics taught me anything, it's that the subject is most certainly NOT currently a science.

Not sure how architecture fits in though.  How can the artistic side of the coin be quantified and codified?  Perhaps more importantly, will that help architecture in any way even if it is possible to do so?

It is conceivable, however distant in the future, that we will be able to quantify what appeals to an individual. We can very roughly predict now who will like what. It is not beyond the bounds of possibility that, as our understanding of the mind increases, our ability to predict what appeals to you vs me vs Golfer One has the potential to increase. At such a point we have repeatable, quantified formulae which can be GUARANTEED to produce a certain response. I'm not saying for one second that such a demystification process would be either desirable or beneficial. As with most sciences, we'd probably assume far too early that we had all the answers and remove individual interpretation from the process long before we were able to replace it with solid formulae. Actually, economics is a perfect example of just that!
Ciao

[Steve Lang]

  GCA a science? maybe, economics a science? maybe, but just practicing scientific methods and technology around a subject doesn't make something a science... proven repeatabe quantifiable things do.

Even something like industrial chemistry is full of uncertainty along with the kinowledge and facts, folks just know and have generally learned how it works. But deductive and inductive reasoning are still alive and continue to be developed and cross fertilize each other... the paradigm

[V. Kmetz]

Hi,

Is cooking (or more grander, creating a restaurant based in one's ability to cook) more of an art or a science? Certainly, elements of both underpin the final result.

Whatever your opinion/answer, golf course design/architecture is likely to fall into a similar category for your viewpoint.

cheers

vk

[Jonathan Mallard]

In some instances, the question of licensing is decidedly "grey"
That is where the Landscape Architecture Boards have made attempts to have it declared Landscape Architecture
To this date it is not listed as Landscape Architecture in any statutes - "but what it is" is still up to interpretation
Interestingly, Civil Engineering License in these States is just as valid as a Landscape Architecture for the same works.

So in a nutshell, it won't happen.

BTW, between a third and a half of the golf architects I know are registered Landscape Architects in the US

Ian,

Thanks.

One brief comment - not intended to draw this out.

Most states write the rules that leave it up to the practitioner to determine his/her own level of competence in a given area.

Short answer is: don't get in over your head!

This of course (pun intended) applies more to the construction, (earthwork, irrigation systems, drainage, cartpaths, bridges, any buildings) than the location of the golf design features.

[Doug Siebert]

A GCA must understand drainage, traffic patterns, agronomy and so on, not to make a great course, but to avoid screwing up what could otherwise be a great course.

Art sets the bar for how great a course can be, but a lack of science can lower that bar.

[Jonathan Mallard]

A GCA must understand drainage, traffic patterns, agronomy and so on, not to make a great course, but not avoid screwing up what could otherwise be a great course.

Art sets the bar for how great a course can be, but a lack of science can lower that bar.

Which proves my point.

Or, to paraphrase, Engineering is essential, but not necessarily required.

[Tom_Doak]

I have long thought that standard economic models have applicability to golf. The golf's goal is to minimize his score and makes decisions which can examined using models of expected probability and risk.

Andrew:

The problem with your theory of golf is the same as the problem with economics:  the assumption that all the players always act rationally is just completely wrong. 

Hardly anyone really understands the odds involved in their risk-taking.  In golf, 98% of players over-estimate their ability and take more risks than would be rational.  In economic spheres, systemic risk is underrated or ignored, even though it produces catastrophic results once every few years.  Of course, the same people build beach houses in Florida, knowing they will have made enough money in between disasters to rebuild everything, with the help of low-interest government disaster loans.

P.S.  One of my interns studied game theory; I'd guess that has some application to golf design, too, but you have to correct for the fact that 90% of people are not out there grinding to shoot their lowest score every time.  They just want to have fun.

[Doug Siebert]

I have long thought that standard economic models have applicability to golf. The golf's goal is to minimize his score and makes decisions which can examined using models of expected probability and risk.

Andrew:

The problem with your theory of golf is the same as the problem with economics:  the assumption that all the players always act rationally is just completely wrong.  

Hardly anyone really understands the odds involved in their risk-taking.  In golf, 98% of players over-estimate their ability and take more risks than would be rational.  In economic spheres, systemic risk is underrated or ignored, even though it produces catastrophic results once every few years.  Of course, the same people build beach houses in Florida, knowing they will have made enough money in between disasters to rebuild everything, with the help of low-interest government disaster loans.

P.S.  One of my interns studied game theory; I'd guess that has some application to golf design, too, but you have to correct for the fact that 90% of people are not out there grinding to shoot their lowest score every time.  They just want to have fun.

I imagine the ideal of the golfer who makes perfect decisions to minimize his score is a fantasy just as much as the ideal consumer.  Game theory is a lot more applicable to a lot of economic decisions than the oversimplifications that economists make*, and probably to golfers as well.

I don't think the idea that not everyone wants to shoot the lowest possible score is a knock on using economic theory in golf.  In Economics the ideal consumer wants to maximize utility, in golf I want to maximize fun.  How does one measure fun?  Well, it is subjective, how I play a shot to maximize fun may be different than you maximize fun.  Likewise how to "best" spend an extra $1000 I have laying around to maximize utility is probably different from how you would spend it.  So I guess the two subjects may be similar to the degree Andrew suggests, but not for the reasons he suggests...


* My dad earned his PhD in Economics from Berkeley and taught Economics for over 40 years, so I feel I know the subject well enough to criticize the "science's" failings.

[Rich Goodale]

I would love to see the results of a scientifically designed and administered study using ShotLink of the intentions, risk preferences and outcomes of a weighted randomized sample of all golfers (from McIlroy to Mr. and Mrs. Havercamp), which could develop and analyze a statistically valid data pool which also accounted for the weather conditions of the day (temperature, humidity, wind strength and direction, etc.) and the mental and physical health of each golfer (overall and on the day).  It would be nice to have this study conducted over real golf courses (whose relevant variables would also be determined and calibrated) but I'll settle for a driving range while the kinks in the system are unkinked.

Not completely sure what I would do with the data if I were an Archie, but it would be fun to see.

[Paul Gray]

I would love to see the results of a scientifically designed and administered study using ShotLink of the intentions, risk preferences and outcomes of a weighted randomized sample of all golfers (from McIlroy to Mr. and Mrs. Havercamp), which could develop and analyze a statistically valid data pool which also accounted for the weather conditions of the day (temperature, humidity, wind strength and direction, etc.) and the mental and physical health of each golfer (overall and on the day).  It would be nice to have this study conducted over real golf courses (whose relevant variables would also be determined and calibrated) but I'll settle for a driving range while the kinks in the system are unkinked.

Not completely sure what I would do with the data if I were an Archie, but it would be fun to see.

Rich,

That is very much what I was alluding to above:

As with most sciences, we'd probably assume far too early that we had all the answers and remove individual interpretation from the process long before we were able to replace it with solid formulae. Actually, economics is a perfect example of just that! 

[Jeff_Brauer]

I have long thought that standard economic models have applicability to golf. The golf's goal is to minimize his score and makes decisions which can examined using models of expected probability and risk.

Andrew:

The problem with your theory of golf is the same as the problem with economics:  the assumption that all the players always act rationally is just completely wrong. 

Hardly anyone really understands the odds involved in their risk-taking.  In golf, 98% of players over-estimate their ability and take more risks than would be rational..........

P.S.  One of my interns studied game theory; I'd guess that has some application to golf design, too, but you have to correct for the fact that 90% of people are not out there grinding to shoot their lowest score every time.  They just want to have fun.

TD, of course, that raised the question of who do you design for?  I get the impression that most of design centers around what some tour pro (presumably more logical than most) might do if he played.  Or even, the 2% of good ams who might play the course.

I doubt most of us would design the course for the total unthinking player (at least most courses) preferring to have some strategy in there for the few who actually get it.

Another thought - I have always wondered what part of golf is strategy and what is skill. No matter how you cut it, skill comes out way on top, near the 90+% range.  And thus, for the majority, why think?  Either you hit the shot well or you don't?

[V. Kmetz]

Hi,

JB, on that last point...skill vs strategy...I've come to compare courses as ones that are either:

Match Play or Medal Courses; and that the essential difference between the two is that the latter more reward "skill" and the former more rewards "fortune."

Certainly this is broad, maybe even superficially so, as a definition.

But when I think of the classic GB&I links courses, early American "importation" of templates, and Scottish transplants designing courses on wild hillsides and other alien types of property, I think of courses and holes that are sometimes so audacious...in their judgment over blindness, or in their green contour, or in their "arcane" features...that they subject opponents or fellow competitors to starkly different "luck" ("fortune") outcomes for shots that are "similar"...thereby leveling off the differences, to some degree, in "skill" that exist between competitors or match opponents.

My point is you place me (10-11 HCP) and a plus player on a blind 150 yard Punchbowl approach and I've got a better chance to halve or win the hole than I do on the 17th at Sawgrass, or most things in between.  The more the course is laden with such features, blindness...wild kicks near good spots...nasty little pots that suck up otherwise good shots...huge green shoulders...the more that better player and I will come together... the more the course requires "skill-sets" (like prodigious driving, first class wedge and sand play) and does so with intimidating penalty, the more the course favors the better player.

Because the game originated in match play, and holes/courses were "discovered" in an effort to fit a match purpose, I think that the further and further we have gone from those origins (naturally so - I'm not judging good or bad here), the more that the elicitation of skill has evolved into the paradigm goal for good and/or championship architecture.

I guess I'm saying that at one time, very far back in the game's origins, the course (or hole, or feature) itself was more like the modern handicap stroke...that make everyday golf enjoyable between people of divergent skill...

What would be the point of playing with a person who's far better or far worse than you, and known to be so ahead of time? For camaraderie, perhaps, but in terms of the execution of a game itself...I think the "unfair," (now) "arcane" features of the older "match" style of architecture performed this equalizing function on a fundamental basis

...But then, for many years of the 20th century, the ethos increasingly became "perfection," and "man-sized" courses that are made for skill... in predictability of playing grounds...and thus in a dozen meaningful ways we sought to serve the better player, because golf was thought to be best realized by the exploits of the elite players and the commercial adjuncts of all that comes with that.

So, my intermediate conclusion to the topical thread question is that GCA was MORE an Art, that we have tried to penetrate, understand and replicate with Science.

cheers

vk

[Rich Goodale]

Every game of golf is an act of art which is expressed over a canvas created by an architect, presented by a greenkeeper and subject to weather.

[Mark Pearce]

Don't we need to understand what science is first?  It seems to me that we don't have agreement on that.  FWIW I don't see either GCA or economics as a science.  That said, I have a BA in Physics.  My wife a BSc in Sociology and Criminology and my eldest son is soon to start studying for a BA in economics.  Two of the other universities he had applied for offered a choice of BA or BSc!  It would seem that even academia is unsure whether economics is an art or a science.  I'm quite sure GCA is less a science than economics, so I think the answer to the OP is no.

[Rich Goodale]

No disrespect, Mark, but how does one get a BA (Bachelor of Arts) in Physics?  My university never offered me the option of getting a BSc in English Literature, even though the dissecting of the poetry of TS Eliot or JV Cunnigham was far more difficult, at least for me, than solving differenential equations......

[Mark Pearce]

Rich,

Nearly every degree at my alma mater was a BA.  Only very modern degrees gained a BSc.  No doubt a tradition lost in the mists of time.

[Rich Goodale]

Mark

Your alma mater should never have subdued to the siren call of the BSc.  If they had stood their ground, later maters might have followed you and stopped the concept of "science" in its tracks.  We all have bits and pieces of knowledge that we access from time to time, with various degrees of interest and capabilities of integration, but nobody can claim to be a "scientist."  IMO

Rich