Calculating handicap index increases/decreases

[Voyager EMH]

Rounding errors are errors.
Another word for them is tolerances.

An accumulation of rounding errors occurs in the calculation of HI.
The quoting of HI to one decimal place is what the system does. This figure will not be perfectly accurate to one decimal place due to the accumulation of rounding errors.

We must accept our given HI as the reality.
We must accept that this figure is inaccurately calculated. This is also a reality.

 

[IanM]

However many places of decimal you take, adding places of decimal could change the outcome.

Wherever you draw a line, there's something very close, just on the other side!

The calculation is "perfect enough!" I'm losing no sleep over it.

 

[Voyager EMH]

The OP wanted an explanation as to why HI adjustment sometimes did not agree with his own shortcut-method.
I think I gave a good example of how this can happen and why.

I completely and utterly accept the reality of the HI calculation that is in use.
I fully understand it.

If Score Differentials were maintained with two decimal places, then the OP's issue would not have arisen.

It would be better if Score Differentials were maintained with two decimal places, because the HI would be accurate to one decimal place, rather than at present when sometimes it s not.
This is a change that could be done just as a change to CR-Par is one that could be done.

 

[Alan Clifford]

It it is being done as specified by the designers then it won't be incorrect.

I used to think that there must be a secret document specifying where and how to round along the lines of the secret PCC document. But now I'm sure there isn't one. Personally, I don't care how this stuff is done, just tell us the rules and specify the rules to all the WHS sub-systems authorities.

 

[Alan Clifford]

The reason I know HNA use 4 dp is because I had a rather good round and the differential was 7 below my handicap index. But I didn't get an exceptional and I queried it. Doing the calculation with 4 dp resulted in a difference of a micro-tad less than 7.

 

[DickInShorts]

This is a change that could be done just as a change to CR-Par is one that could be done.

I understand from reading this forum that CONGU is introducing CR - Par next year/season

 

[Voyager EMH]

Well, waddayaknow, it has happened again today.

My HI went up by 0.2 to 4.2 after a 3.6 (3.5955) was replaced with a 4.6 (4.5525)

But with 2 decimal places (or more) for the SDs, I would be 4.1.

 

[backwoodsman]

Well, waddayaknow, it has happened again today.

My HI went up by 0.2 to 4.2 after a 3.6 (3.5955) was replaced with a 4.6 (4.5525)

But with 2 decimal places (or more) for the SDs, I would be 4.1.

View attachment 48686

Well, it's bound to happen if you keep trying to calculate your index by the wrong method ...

 

[Voyager EMH]

Well, it's bound to happen if you keep trying to calculate your index by the wrong method ...

It would cease to happen if Score Differentials were kept to 2 decimal places.

The method I have been using is the same method that the authority uses.
I merely use more accurate measurements in an illustration.

 

[Steven Rules]

It's pretty pointless advocating a case on this forum that Score Differentials should be kept to two decimal places. Nobody here can do anything about it.

 

[Voyager EMH]

It's pretty pointless advocating a case on this forum that Score Differentials should be kept to two decimal places. Nobody here can do anything about it.

Fairly similar with regard to CR-Par.
Opinions on here won't count for anything as to when and whether it will be done. Yet we continue to discuss it. (Pretty pointless?) Is that because it is a forum?

I do seem to get right up peoples' noses when I discuss SDs to one decimal place being inaccurate with regard to the resulting HI.
SDs to one decimal place is a fact and a reality. I accept it as such.
No CR-Par today is a fact and a reality.
Next year CR-Par will be a fact and a reality.
Things can change, sometimes.
SDs to two decimal places is something that could change. Maybe it will, maybe it won't. I don't see it as something that should not be allowed to be discussed.

 

[wjemather]

Fairly similar with regard to CR-Par.
Opinions on here won't count for anything as to when and whether it will be done. Yet we continue to discuss it. (Pretty pointless?) Is that because it is a forum?

I do seem to get right up peoples' noses when I discuss SDs to one decimal place being inaccurate with regard to the resulting HI.
SDs to one decimal place is a fact and a reality. I accept it as such.
No CR-Par today is a fact and a reality.
Next year CR-Par will be a fact and a reality.
Things can change, sometimes.
SDs to two decimal places is something that could change. Maybe it will, maybe it won't. I don't see it as something that should not be allowed to be discussed.

Maybe it's not so much what you're saying, but the fact you seem intent of beating everyone into submission by bludgeoning your views onto the forum - I don't have enough fingers to count the number of times you've repeated this stuff about an utterly insignificant decimal place.

 

[Voyager EMH]

Maybe it's not so much what you're saying, but the fact you seem intent of beating everyone into submission by bludgeoning your views onto the forum - I don't have enough fingers to count the number of times you've repeated this stuff about an utterly insignificant decimal place.

Ah, I get it.
Its me that is being objected to, not the points of fact that I bring up from time to time when topics of interest to me arise. I never bother to count those times.

As for "bludgeoning my views", I think that is a gross exaggeration.
I don't find arithmetic of this kind to be insignificant. That is my unbludgeoned view.

An HI of 4.1 giving a PH of 4.
An HI of 4.2 giving a PH of 5.
I can not see that one decimal place as being "utterly insignificant".

 

[wjemather]

Ah, I get it.
Its me that is being objected to, not the points of fact that I bring up from time to time when topics of interest to me arise. I never bother to count those times.

As for "bludgeoning my views", I think that is a gross exaggeration.
I don't find arithmetic of this kind to be insignificant. That is my unbludgeoned view.

An HI of 4.1 giving a PH of 4.
An HI of 4.2 giving a PH of 5.
I can not see that one decimal place as being "utterly insignificant".

And to prove you are not bludgeoning, you repeat yourself again!

You may not see it, but it still is (statistically speaking, of course).

Edit: Handicapping is only a (fairly decent) approximation of scoring ability and can never be an accurate enough measure for compound rounding errors of 0.1 to be considered significant - even when they result in Course/Playing Handicap variances - indeed, accounting for all factors (starting with the ratings that underpin the system), the overall variance of the resultant Handicap Indexes is substantially larger than 0.1.

 

[Voyager EMH]

And to prove you are not bludgeoning, you repeat yourself again!

You may not see it, but it still is (statistically speaking, of course).

You repeatedly criticise me rather than address the points that I make by giving an explanation of your reasoning.

The OP asked for an explanation as to why his HI did not change in the way that he expected.
I believe I have given a good explanation, with illustrative examples, of how this can happen.
His opening post was of interest to me.

Perhaps in future I should post only on topics that do not interest me very much. Or cease posting altogether to suit you.
Feel free to use the ignore function.

 

[Voyager EMH]

And to prove you are not bludgeoning, you repeat yourself again!

You may not see it, but it still is (statistically speaking, of course).

Edit: Handicapping is only a (fairly decent) approximation of scoring ability and can never be an accurate enough measure for compound rounding errors of 0.1 to be considered significant - even when they result in Course/Playing Handicap variances - indeed, accounting for all factors (starting with the ratings that underpin the system), the overall variance of the resultant Handicap Indexes is substantially larger than 0.1.

And the "can never be an accurate enough" would be made less disappointingly so by SDs with two decimal places instead of one.
This would not be an "utterly insignificant" change.
It would be a change that would change things.
It would do no harm to change.
It does no harm to debate this issue.

Similar to CR-Par will change things.
Two players at the same course could have the same PH and then with no change to their HIs, could find themselves with differing PHs when CR-Par comes in.
So their calculated approximations of playing ability will change, even though their actual playing abilities haven't changed.

Changes are changes.
I see it as beneficial to make calculated approximations more accurate rather than the "there or thereabouts is OK, not worth bothering with" point of view.

 

[Alan Clifford]

Fairly similar with regard to CR-Par.
Opinions on here won't count for anything as to when and whether it will be done. Yet we continue to discuss it. (Pretty pointless?) Is that because it is a forum?

I do seem to get right up peoples' noses when I discuss SDs to one decimal place being inaccurate with regard to the resulting HI.
SDs to one decimal place is a fact and a reality. I accept it as such.
No CR-Par today is a fact and a reality.
Next year CR-Par will be a fact and a reality.
Things can change, sometimes.
SDs to two decimal places is something that could change. Maybe it will, maybe it won't. I don't see it as something that should not be allowed to be discussed.

And it is not consistent between jurisdictions. That is the thing that bugs me more than the arithmetic methods themselves.

 

[rulefan]

And the "can never be an accurate enough" would be made less disappointingly so by SDs with two decimal places instead of one.
This would not be an "utterly insignificant" change.
It would be a change that would change things.
It would do no harm to change.
It does no harm to debate this issue.

I don't necessarily disagree but you have told us this many times. Isn't once enough?

 

[Alan Clifford]

And it is not consistent between jurisdictions. That is the thing that bugs me more than the arithmetic methods themselves.

Actually the aritmetic bugs me as well.

Starting with some 1 dp numbers you calculate differential as a float. Store it (IEEE 754).
Find the smallest 8 out of twenty.
Average, store as a float.
Compare to an integer (exceptional calculation)
Fiddle about with soft cap calculation and hard cap
Store as a float.
At this point round to 1 dp.

Or, round to 1 dp wherever you like BUT PUBLISH THE STANDARD.

 

[Voyager EMH]

I don't necessarily disagree but you have told us this many times. Isn't once enough?

No.
When someone posts directly to me, then I am inclined to respond.
I am also inclined to disagree with a notion that I should be silenced, so that another person's view can prevail.
There may be two differing views and on this particular topic, neither is right or wrong, they are merely different.
Each view has its merits and weaknesses.
It is a discussion, a debate, a forum.