OT - Did stroke hole allocation method change in the US with WHS?

[Wayne_Kozun]

I always thought that the strokes were determined on holes where they were most needed by a bogey golfer when playing a scratch golfer, but that these were somewhat adjusted so that the highest ranked stroke holes were not at the start or end of a nine, and so that odd strokes holes were on the front and even stroke holes were on the back.

Players often thought that the stroke ratings were based solely on the difficulty of the hole, and this was encouraged by TV broadcasters.  But my understanding was that difficulty of the hole didn't really matter, especially if a hole was very difficult for both low and high handicappers.

But it seems like hole difficulty is now a more major consideration.

This is how the USGA describes it as:

The conclusion was to base the stroke index table on overall hole difficulty relative to par as a starting point, then use the procedures that work well to provide equity in match play, including placing odds and evens on different sides, avoiding consecutive low stroke index holes, and avoiding low stroke index holes at the beginning or end of a nine.

Did this change when the WHS was introduced?

[Simon Barrington]

Possibly this came from the guidance given out to Clubs in the UK & RoW by the R&A for the (previous) CONGU Handicapping Scheme (1985 onwards).

That specifically detailed that difficulty was not to be a factor in assigning "Stroke Index" for each hole (even though the accepted that tradition of old, meant that was mostly the case). They wanted to get a more even spread of strokes given/received across the 18.

The not wanting to have low Stroke Index on Hole 1 (& Hole 9) was apparently to enable a fairer distribution of shots given/received if the Match was to go into Extra Holes.

So a bit of a fudge that probably is the genesis of the change in the WHS (an amalgam of the two schemes, but mostly based on the US System).


Architecturally, there are adverse effects of the new system, in that there has been an explosion of lower handicaps in the UK, Ireland & RoW versus the old CONGU.

These emboldened "better" players have led to a massive amount of pressure for (potentially unnecessary) remodelling across the UK, because they claim that "The course doesn't challenge them enough anymore".

Index has come down for very many sub-5 Handicap Players (on average by c. 3 strokes), but the absolute scoring has not been reduced by anywhere near the same extent.
Arccos data shows that the Average Zero Index (aka "Scratch") Golfer averages 3.1 Strokes above Par now, which would make them 3 Handicap in old money...
England Golf have now had to introduce rules about the percentage of "General Play" rounds included in competitiors Index for National Competitions and have ejected over 60 Applications for these in the last year alone.

Not all this remodelling work has been done with the historical care and attention that one would hope for on the older courses across the UK, especially those designed by ODGs of repute. We are a small (hopefully growing) group of enthusiasts on here, but the wider UK golfing public (especially the younger ones) frankly do not know or care for the architectural merit or heritage as we would like....

[Wayne_Kozun]

Thanks Simon.  I remember seeing an old scorecard from over 100 years ago from a Scottish club with the hole allocation where they were just evenly allocated across the 18 holes without using other criteria.  It was something like - if there was one stroke it would be given on the 9th hole.  If there were two it would be at 5&14.  If there were four it would be 4,8,12&16.


I wonder if that isn't a better way to do it and we are overthinking it.

[Simon Barrington]

There are some Clubs in the UK who had/have the exact same Stroke Indices as exist on the corresponding numbered hole on The Old Course, I quite like the tradition and standardisation of that.

[Wayne_Kozun]

Interesting.  Looking at the stroke indices on TOC it seems to be that they still use this even distribution across the 18 holes.  #1 stroke hole is 14 and #2 stroke hole is 5 - they come in the middle of each nine.  The Road Hole is the #5 stroke hole.

[jeffwarne]

Thanks Simon.  I remember seeing an old scorecard from over 100 years ago from a Scottish club with the hole allocation where they were just evenly allocated across the 18 holes without using other criteria.  It was something like - if there was one stroke it would be given on the 9th hole.  If there were two it would be at 5&14.  If there were four it would be 4,8,12&16.


I wonder if that isn't a better way to do it and we are overthinking it.

Bingo.
yes we are overthinking it, and much like the rules change" simplifications" it's very hard for anyone to stay current with the latest simplifications, so consequently those who once invested the time to learn the rules carefully, or the handicap system carefully , are either confused or required to learn it all again.
Small wonder so few do, and so many misconceptions exist.


At the end of the day, none of the statistical attempts to make a more perfect system etc. come close to being as important as actual honesty or compliance with posting every round.


Under no system do 13 handicaps shoot 74 in a Member-Guest.

[Thomas Dai]

Understand some U.K. clubs have 2xSI’s. One following the old ‘spread’ method for use in matches and one based on ‘hole difficulty’ for use in stableford. Seems sensible. Don’t know how this fits in with WHS though.
Atb

[Wayne_Kozun]

Understand some U.K. clubs have 2xSI’s. One following the old ‘spread’ method for use in matches and one based on ‘hole difficulty’ for use in stableford. Seems sensible. Don’t know how this fits in with WHS though.
Atb

Isn't Stableford essentially stroke play, but "backwards" as a high score is better?  If so then does hole allocation matter?  Don't you just subtract the handicap off the final 18 hole score and you are agnostic as to the score on any hole?


Or is Stableford used in a match play context?

[Matt MacIver]

For reasons I can’t explain, in our Stableford game a player (say a 10HC) subtracts their handicap from 36 and thus needs to earn at least that many points (26+ here) to share in the pot.   

[Wayne_Kozun]

36 is level par in regular stableford scoring.  You get two points for a par so 18 pars is 36. 


You can convert your stableford points to strokes as Stableford = 108- strokes for a par 72 course.  Or get your points by 108- strokes.  36=108-72

[A.G._Crockett]

I always thought that the strokes were determined on holes where they were most needed by a bogey golfer when playing a scratch golfer, but that these were somewhat adjusted so that the highest ranked stroke holes were not at the start or end of a nine, and so that odd strokes holes were on the front and even stroke holes were on the back.

Players often thought that the stroke ratings were based solely on the difficulty of the hole, and this was encouraged by TV broadcasters.  But my understanding was that difficulty of the hole didn't really matter, especially if a hole was very difficult for both low and high handicappers.

But it seems like hole difficulty is now a more major consideration.

This is how the USGA describes it as:

The conclusion was to base the stroke index table on overall hole difficulty relative to par as a starting point, then use the procedures that work well to provide equity in match play, including placing odds and evens on different sides, avoiding consecutive low stroke index holes, and avoiding low stroke index holes at the beginning or end of a nine.

Did this change when the WHS was introduced?

I could be wrong, and if so I'm sure it will be corrected, but I think the rationale for handicapping holes was changed well BEFORE the WHS changes.  The equalization of a bogey golfer and a scratch golfer isn't really the recommended rationale for handicapping holes anymore; overall difficulty is.

For the rating of holes do be done correctly, the staff (or a committee) at a particular course has to accumulate a lot of data that gives a picture of difficulty for ALL players and ALL tees, rather than using just the back tees and guessing where a bogey golfer will most need strokes to equalize to a scratch.  IMO, it's a better system, BUT it does require some diligence by somebody to actually accumulate and then disaggregate the data. 

[Wayne_Kozun]

IMO, it's a better system, BUT it does require some diligence by somebody to actually accumulate and then disaggregate the data.

Isn't this data very easy to get?  At my club we have had to enter hole-by-hole scores for several years.  So we have the data.  So could any course could gather the data if they enforce the requirement of hole-by-hole score entry.  They can also make this easier with apps - here in Canada there is a Golf Canada app that you can use for keeping your score.  It has GPS distances as well.  Enter your scores on that, click finished after 18 and you are done.  Easy peasy!

And on the hardest hole - 18 could be the hardest hole on a course but it is unlikely to be any lower than the #4 stroke hole due to the other considerations.

[Erik J. Barzeski]

The truth of it all is that… so long as the low numbered stroke holes aren't all clustered at the beginning or the end of the round, it doesn't really matter whole hole is stroke index 1 and which is 18. Just as long as they're kinda spread out… matches tend to end the way they "should."

That's why the current recommendation has the triads thing, with 1 and 2 in the middle triad. (Mostly because they don't want stroke index 1 to be the first hole, should a match go to extra holes.)

[A.G._Crockett]

IMO, it's a better system, BUT it does require some diligence by somebody to actually accumulate and then disaggregate the data.

Isn't this data very easy to get?  At my club we have had to enter hole-by-hole scores for several years.  So we have the data.  So could any course could gather the data if they enforce the requirement of hole-by-hole score entry.  They can also make this easier with apps - here in Canada there is a Golf Canada app that you can use for keeping your score.  It has GPS distances as well.  Enter your scores on that, click finished after 18 and you are done.  Easy peasy!

And on the hardest hole - 18 could be the hardest hole on a course but it is unlikely to be any lower than the #4 stroke hole due to the other considerations.

The data is easy to get IF the club is getting hole by hole scores entered for golfers of all levels and playing different tees.  I suspect that your club is an outlier in requiring that.  I don't think I've seen that requirement anywhere, and Golf Genius type programs for hole by hole entry aren't a great answer, simply because a lot of golfers are resistant to being tied to their phone during a round of golf; some guys are playing golf to get AWAY from using their phone.

[Wayne_Kozun]

I think the requirement for hole by hole scoring is more common here in Canada - friends have told me that it is in place at their clubs as well.  And the Golf Canada app makes it easier, assuming that you are willing to use your phone for scorekeeping.
I agree that some people don't like using their phone while golfing, but I that is likely correlated with age.  Younger golfers are more likely to not mind using their phone for scoring.  Same with golfers who use other technology like Arccos, which I use for tracking my game.

The other nice thing about doing this is it let's you track your "ringer" score. 

[Matt Schoolfield]

Isn't this data very easy to get?

This is my once-per-handicap-post reminder that, at this point, machine learning could trivially be used to handicap courses automatically, dynamically, and with included seasonal variations. Why the USGA isn't moving to that type of system is beyond me, as it would be a more accurate system, while being both cheaper and easier to implement for everyone involved.

[Wayne_Kozun]

Isn't this data very easy to get?

This is my once-per-handicap-post reminder that, at this point, machine learning could trivially be used to handicap courses automatically, dynamically, and with included seasonal variations. Why the USGA isn't moving to that type of system is beyond me, as it would be a more accurate system, while being both cheaper and easier to implement for everyone involved.

How would this work?  You would have an AI system that runs on a continuous basis that would do handicap, both for calculating an index, as well as the stroke hole allocations?

[Michael Felton]

That sounds like a card-printing nightmare to me. Changing stroke indices during the season might be most "accurate", but it's a lot to deal with.


I think it's worth remembering that everyone's average scores on a given hole are unlikely to be a whole number. For someone who's a 9 index and averages about +12, that person might find that on the easier holes they average +0.3 and the hardest holes more like +1.1. Same thing for a 0 index who averages +3. They might average -0.2 on the easiest hole and +0.5 on the hardest hole. When you play matchplay between two players who are 1 stroke different, it's not like player A is 1 shot better on index 1 hole and the exact same on the other 17 holes. It's more like that they are 0.03 strokes better on some holes and 0.15 strokes better on others and that all adds up to 1 overall. It doesn't really matter when that stroke occurs so long as it's likely to happen before the game is over and not likely on the 19th hole, which is exactly the system.


In stableford and max net double bogey, it does matter where that falls and it should be based on difficulty relative to par. I had heard about some courses implementing a match play stroke index and a stroke play stroke index, but I've never seen it in practice.

[Matt Schoolfield]

How would this work?  You would have an AI system that runs on a continuous basis that would do handicap, both for calculating an index, as well as the stroke hole allocations?

We could use a fairly simple collaborative filtering algorithm. It's machine learning, but hardly "AI". Really more of a clustering algorithm. If we maintained a few courses as our baselines, which would likely be tournament courses that get tons of regional players per year, then we can simply use them as a baseline to set expected values for what the other courses would be. How it works is that how those players with baseline ratings play on other courses (on average), determines what their rating is. The way collaborative filtering works is that other players can fill in the gaps by how they perform in relation to the baseline players. So even though you don't have a ton of data from the baseline courses, it sort of doesn't matter, because everyone is effectively working together to rate the courses by playing them (again, with everything in relation to the players playing baseline courses). From there, we could determine seasonality as average scores increase and decrease on each course annually depending on climate.

It would be an imperfect system, sure, but hardly more imperfect than the existing system, which allows courses to languish for a decade before being re-rated at a non-trivial cost (even if that cost is included in dues). We're already using a system very close to this to calculate PCC. The regularity of changes are arbitrary, you could have new rating/slope once per day, per season, per year, or per decade if you wanted. It's just data being fed into a convoluted math problem.

The benefits are that literally any course could be rated fairly trivially, even someone's homemade backyard course, unrated match play courses, etc. I'm pretty sure you could create a system that doesn't even need baseline courses, but I think that might cause the concept of par and/or scratch to start moving around over time, so I'm not ready to get rid of creating a baseline for everything to be contingent on.

[A.G._Crockett]

Isn't this data very easy to get?

This is my once-per-handicap-post reminder that, at this point, machine learning could trivially be used to handicap courses automatically, dynamically, and with included seasonal variations. Why the USGA isn't moving to that type of system is beyond me, as it would be a more accurate system, while being both cheaper and easier to implement for everyone involved.

The USGA is not involved in any way in handicapping holes at any golf course.  That is completely the responsibility of the individual club.  And it would NOT be cheaper and easier for the individual club, if only because of the “menu costs” of scorecards. 

[Wayne_Kozun]

Isn't this data very easy to get?

This is my once-per-handicap-post reminder that, at this point, machine learning could trivially be used to handicap courses automatically, dynamically, and with included seasonal variations. Why the USGA isn't moving to that type of system is beyond me, as it would be a more accurate system, while being both cheaper and easier to implement for everyone involved.

The USGA is not involved in any way in handicapping holes at any golf course.  That is completely the responsibility of the individual club.  And it would NOT be cheaper and easier for the individual club, if only because of the “menu costs” of scorecards.

I think what Matt is getting at here is for calculating a golfer's handicap index instead of using the current system of course and slope rating.  You wouldn't need raters to go out and do a physical course rating.  You would just do it statistically once you have a large enough sample of scores posted.

That would be independent of the stroke hole allocation, but you could do a similar method for that as well. although you may only want to change that once per year or so.

[Erik J. Barzeski]

simply because a lot of golfers are resistant to being tied to their phone during a round of golf; some guys are playing golf to get AWAY from using their phone.

I typically enter hole by hole, with stats, at the end of the round. I don't take my phone out much at all when I'm actually playing (I often forget to take photos of courses I've played).

This is my once-per-handicap-post reminder that, at this point, machine learning could trivially be used to handicap courses automatically, dynamically, and with included seasonal variations. Why the USGA isn't moving to that type of system is beyond me, as it would be a more accurate system, while being both cheaper and easier to implement for everyone involved.

The current system is pretty good and has the benefit of being pretty simple and consistent.

It would be an imperfect system, sure, but hardly more imperfect than the existing system

FTFY: It would be a complex system, sure, but only a good bit more complex than the current simple, steady system.

which allows courses to languish for a decade before being re-rated at a non-trivial cost (even if that cost is included in dues).

I think the cost is pretty trivial for many (most?) AGAs.

[archie_struthers]

 




Just opened up a window to my continual disgust with the handicap system. Not counting every score is so stupid if you want to look at the present state of someones' game.  Reducing the number of scores from 10-8 just exacerbated the issue.


Please don't tell me the system is about potential...nuff said  apologize for being off topic (a litte)

[Carl Johnson]

I always thought that the strokes were determined on holes where they were most needed by a bogey golfer when playing a scratch golfer, but that these were somewhat adjusted so that the highest ranked stroke holes were not at the start or end of a nine, and so that odd strokes holes were on the front and even stroke holes were on the back.

Players often thought that the stroke ratings were based solely on the difficulty of the hole, and this was encouraged by TV broadcasters.  But my understanding was that difficulty of the hole didn't really matter, especially if a hole was very difficult for both low and high handicappers.

But it seems like hole difficulty is now a more major consideration.

This is how the USGA describes it as:

The conclusion was to base the stroke index table on overall hole difficulty relative to par as a starting point, then use the procedures that work well to provide equity in match play, including placing odds and evens on different sides, avoiding consecutive low stroke index holes, and avoiding low stroke index holes at the beginning or end of a nine.

Did this change when the WHS was introduced?

Here is my quick take.  As I recall, like Wayne, the former USGA "recommendation" was to calculate the difficulty factor on the "most needed" basis by comparing scratch and bogey player difficulty based on scoring data from competitions.  The Jan. 2020 (new) Rules of Handicapping recommends essentially the same method, except that the relative difficulty for the scratch and bogey player is based on the course rating for each hole.

The 2020 Rules of Handicapping book, Appendix E, gives a simple example of how that works.  To determine the relative difficulty you need the par and bogey rating for each hole.  I thought the principle, with the example, was clear and very helpful to an understanding of the process.

If you turn to the Jan. 2024 Rules of Handicapping book, Appendix E has been changed.  The example of how to determine the relative difficulty for the bogey and scratch player has been eliminated and replaced with the following: "Authorized Associations can provide a report detailing the difficulty factor that has been assigned to each hole derived from the course rating system."  I do not know when the change in the wording occurred, first in 2024 or earlier, but I liked the 2020 explanation much better.

Regardless, the principle of difficulty relative to par remains on a relative basis between the scratch and bogey golfer.


My unsubstantiated opinions. (1) Using the competition data is better than using the rating data, although it takes the Committee more work; (2) None of this means anything unless players properly post scores as they should under the Rules of Handicapping, which most players do not (vanity is a bigger issue than sandbagging); (3) course difficulty changed but course not re-rated can be a problem; and (4) Stoke allocations should be based on the tees you play from, though I'm not sure how this would work when you play from different tees in the game, and it would be more complicated in any case.

[Erik J. Barzeski]

Just opened up a window to my continual disgust with the handicap system. Not counting every score is so stupid if you want to look at the present state of someones' game.  Reducing the number of scores from 10-8 just exacerbated the issue.

I don't have the exact number at my fingertips but it was > 90% of golfers with handicaps saw their indexes change by ≤ 0.2 from the 10/20 * 0.96 to the 8/20 system. If you didn't like the 10/20 * 0.96, that's fine, but to pretend that it was really a change to go to 8/20 feels misplaced.

Please don't tell me the system is about potential...nuff said  apologize for being off topic (a litte)

It's a measure of your "demonstrated ability." Previously they used the word "potential" but since a 24-handicapper's long-term potential might be to win the Masters, they stopped using that term. They meant it more in terms of your "potential" in the next round to shoot a score.