Will GHIN website make new World Handicap score adjustments automatically?

[Roadking2003]

Exactly. The only way to compute PCC by time of day is to match score postings with tee times and most courses do not have that information.

[DaveLeeNC]

8 is a small sample so it is going to take a pretty big deviation from expected values to generate even a small adjustment. dave

[Halebopp]

Me neither but I can tell you that the smaller the sample size, the greater the difference between the expected and actual scores would (or should) need to be for the PCC to kick in.

[Shilgy]

Which is why if as few as 8 scores were posted it would be very unlikely to have a PCC adjustment. But if the scores were abnormally high why would 8 scores not qualify for a PCC? Perhaps you understand differently but my understanding is that the adjustment will not really “zero out” the days scores. The adjustment will be more modest than that. If scores are 5 higher than normal, for example, you might get a one or two stroke PCC. If scores are lower than normal, which would be really rare I would expect, the most the PCC adjusts is -1.

[davep043]

I agree with each of the last 3 posts. It seems logical to me that an 8-stroke average deviation from the expected scoring (as an example) might produce a 1-stroke PCC if based on 8 scores, 2 stroke PCC if based on 50 scores, and a 3 stroke PCC if based on 150 scores, or something of that general trend. Those numbers are my guesses for illustration purposes only, and have nothing to do with actual numbers. Its also fair to point out that the PCC has to begin at some fairly arbitrary number of players. I can't say that 7 is too few but 8 is enough, any more than I can tell you that 19 is too few, but 20 is enough.

[jvincent]

I thought I read somewhere that the PCC was limited to a range of -1 to +2 ?

[davep043]

Section III, Rule 5.6, says its -1 to +3. For ease of reference, the Rules can be found at https://www.usga.org/handicapping/roh/2020-rules-of-handicapping.html

[Roadking2003]

All I'm saying is using eight scores to determine playing conditions is statistically weak.

[jvincent]

Got it, thanks.

[Roadking2003]

Yes, it has to start somewhere. But the larger the universe the more accurate the measure of course difficulty.

 

So look at the extreme to illustrate. Let's say two golfers post scores and you determine PCC. What are the odds that those two golfers shot higher or lower than expected due to course conditions? Not too good because our scores fluctuate widely even when the course conditions are exactly the same. So if you only use two golfers and they both had a bad day (not unusual at all) with perfect conditions, you would conclude (incorrectly) that a PCC is needed.

 

Now go to the other extreme. You have 160 golfers playing. Now the odds of all or even most having a bad day are extremely small. So if they post scores above what was expected you could easily justify a PCC.

 

Personally, I wouldn't even compute PCC if there were less than 30 postings.

 

 

[davep043]

As I said in my previous post, I can't defend one specific number over another, I can't tell you that the right "starting point" is 8 players, or 30, or 50. I agree that the smaller the sample size, the more significant the deviation needs to be to trigger an adjustment. I think your example, only two scores recorded on a "perfect" day is completely improbable, but the same could happen with only 8 players. Its obviously much less likely as the number of players increases. There is a significant amount of data available from systems that have done a similar type of adjustment, from CONGU, and from Australia (I believe) so that the folks who decided on the PCC calculation weren't simply shooting blind, which is what I'd be doing if I made those rules. I have a feeling you're in the same boat.

[Newby]

The Australian and European systems were the ones that lead the way for WHS to follow. CONGU uses a different process but with a similar outcome.

[rogolf]

Of course, any PCC adjustment only applies to those players that played that course that day and posted scores for handicap. May not make a huge difference to those few players. I'll just trust the statisticians' work.

[sui generis]

Wasn't it you a few posts ago boasting of your expertise is statistics?

[Roadking2003]

No. You must be thinking of another dude. I like stats and use them often, but have no statistics degree.

[rjp1969]

I would like to know so that I know how much impact it is likely to have. If its max change is 0.1 on a differential, than ignoring it may be the right plan. If, however, it can have a material impact on every differential to the point where ignoring it can have a real effect on an index, than it may drive me to paying for officialness.

 

Also, I attend and set up the comps for golf weekends: competitions with friends who have a variety of official ---> self calculated ----> assigned by us handicaps and I want to understand the new system as well as I can to make that work.

[davep043]

We'll see how it all turns out, but in Virginia the Handicap people are expecting the PCC to apply in something like 10% of the rounds posted. If that proves to be correct, that typically would be 2 rounds in your most recent 20, and its probable that no more than one of those would be used in calculation of your index (best 8 of 20). PCC of 2 or 3 would be substantially less common than PCC = 1, so the likely change to your Index is something like 0.1. From what I've been told, the PCC won't be calculated to bring the overall distribution all the way back to "normal", merely to move it in that direction, which would make those PCC-corrected rounds less likely to be among the 8 best. I don't believe that the PCC will make a big difference in most cases, but a difference of just 0.1 can change a Course Handicap by a full stroke.

Please take all this as my semi-informed best guess, I'm not privy to the formula, nor to the data that was used to develop the formula. I've attended my local association's handicap seminar, I've read what the actual rules say, I have a couple of friends who are more well-connected than I am, and I've tried to pay attention to the posts by more well-informed folks around here. Maybe I'm right, maybe I'll be proven wrong, only time will tell.

[Augster]

I’m pretty sure you’re exactly correct. The PCC is much ado about nothing. A 0.1 difference in index seems about right. And most people I play with just round to the nearest integer instead of taking slope into the calculation. CR-par is going to blow their minds.

I wish the GHIN app would bring back the slope slider. It was really easy to show how insignificant 0.1 really is. Unless the tees you play are right on the “slope rounding line” for your specific cap.

Going forward, for those that don’t want to buy in, figure your daily cap without the PCC (easy) then also figure the cap if you were 0.1 better. If the cap drops an entire shot, play with the lower cap that day.

That’d be a really fair way to do it until we see more data etc.

All of the suggestions above assume the course is rated correctly. If a course rating and slope are significantly off, the PCC is going to kick in daily. Hopefully that’ll get flagged at the state level and they can adjust the CR/slope remotely or come out and do an “emergency” re-rating of the course or tees played.

[davep043]

@Augster says
"All of the suggestions above assume the course is rated correctly. If a course rating and slope are significantly off, the PCC is going to kick in daily. Hopefully that’ll get flagged at the state level and they can adjust the CR/slope remotely or come out and do an “emergency” re-rating of the course or tees played."
I'm not so sure this will happen, especially at private clubs. A player's handicap is established playing a course, using the CR and Slope for that course. As long as nothing really changes, he'll continue to shoot the same scores, so his differentials (based on that handicap) will fall right into the "expected range". For example, if the CR/Slope is too low, he'll have an artificially high handicap. When he has a reasonably normal game, his net scores will be artificially low, which will coincide with the "low" Course Rating. That same thing will hold true for a population of golfers. A course's rating will only look off if you can compare players' differentials while playing numerous courses.

[rjp1969]

Thanks! This is a perfect answer (both DaveP and Augster) for my purposes. It's a good addition from a completeness point of view, incorporating the CSS concept, but unlikely to have a material impact very often.

It is annoying that this makes it impossible to have a properly calculated homebrew handicap, though.

 

also noticed that DaveP and Augster have almost exactly the same number of posts (4566 vs. 4565)!

[DaveLeeNC]

 

Regarding the effect of PCC being kind of an 'auto adjust' WRT possibly mis-rated courses...

Elsewhere in this forum that point has been brought up. It seems that the basis for PCC adjustment will be 'historical scoring norms' for a given course rather than simple/actual differential vs Slope/CR adjusted predictions (based on HI). So even if there is no 'home course bias' which would be the case where I play where there are one club with 9 courses, that effect won't exist.

Given ANY CR/Slope assignment to a course there will be some 10 HI golfers (just to pick a #) who will play that course better than other (legit) 10 HI golfers. I have a brother who is younger than me and his scoring vs mine (when I was his current age) were/are similar. He is a good bit better than I was from outside 150'ish yards, but I was always a lot straighter off the tee. He would have a big advantage over the 'younger me' on long/wide open courses. He had little chance on courses with lots of trouble off the tee.

dave

[mey123]

I have this email from TheGrint from early in Janurary. Hopefully it helps others. I linked my Ghin # to the site (i bought it from a local amaetur golf association) and it links directly into TheGrint. I am not sure i like the new setup, so may be taking a closer look at Golfshot or something.

You can also now buy a Ghin # through them.

But heres the email text:

"WHS Integration

We are excited about the progress of our integration with the World Handicap System since we launched on Jan 15th. The reception has been overwhelming. In fact 80%+ of people who have logged in, have created or linked a WHS# account. 

However, we have also gotten an enormous number of emails that our Customer Support team is carefully reviewing and responding to. 

So, if you wrote to us but have not received a response as of yet, please accept our sincerest apologies. Please know that we will get to your email and respond as soon as possible.  

The following are some important issues and clarifications that we initially had that have since been resolved:

1) We solved an issue where some users were purchasing the PRO membership however, it was not displaying properly. 

2) When you create or link your WHS# we have to show "NH" for one day until the revision of your handicap is performed daily at midnight. 

3) Those of you who have linked an old Handicap ID (GHIN or equivalent), please make sure that Handicap ID is active. If not, your number will show "NH". We are making adjustments to the app at this time to make this more obvious.  

Thanks again for the tremendous response and even more so for your patience as we make this transition.

Cheers, 

TheGrint Team"

[davep043]

@DaveLeeNC says:
"Elsewhere in this forum that point has been brought up. It seems that the basis for PCC adjustment will be 'historical scoring norms' for a given course rather than simple/actual differential vs Slope/CR adjusted predictions (based on HI). So even if there is no 'home course bias' which would be the case where I play where there are one club with 9 courses, that effect won't exist."
I'm not sure I understand what you're saying here with the term "historical scoring norms for a given course". As I read the Rules (Interpretation 5.6/1 here https://www.usga.org/handicapping/roh/2020-rules-of-handicapping.html), the PCC will be based on actual differentials for play on the day, as compared to expected scoring range for each player who submits a score. Here are the steps outlined:
Calculate the expected score for each eligible player.Calculate the expected standard deviation of Score Differentials at the golf course, incorporating all applicable Slope Ratings.Establish how many players scored better or worse than expected on the day.The proportion of players submitting a score equal to, better than or worse than their expected scoring range determines whether a PCC adjustment is required.If an adjustment is required, determine how much harder or easier the golf course played that day.Based on these calculations, determine any final PCC adjustment required for play on that day.The only "history" that would influence this, it seems to me, is the Handicap Index of each player, which in turn determines the Differentials for each player. Now in your situation, where your scores are spread over 9 courses, you really might see indications that one of those is mis-rated. The amount of non-member play could cloud the issue a bit, or could actually make the distinction more pronounced. Its probably possible to evaluate scoring records from only the members to look for trends.

[DaveLeeNC]

Dave, maybe I will go dig up the discussion. But some months the answer was that the basis for adjustment was not JUST posted diff's vs HI's. I specifically said that the PCC will have a tendency to adjust possibly mis-rated courses, and some knowledge-able person (don't recall who it was) said that this was not the case. If there was some course where posted diff's were typically 5 strokes above HI's, and others where posted diff's were typically 3 strokes above HI's, that would be the basis for each course.

But it was several months ago and I don't recall what thread it was in.

dave