WHS working well for me

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
Why? Individual differentials are expressed (ie rounded) to the nearest 10th.
Yes they are, but why are they not rounded to two decimal places?
This would produce a Handicap Index that is correct to one decimal place instead of being possibly inaccurate by plus or minus 0.1.

WHS has built-in this inaccuracy and it could very easily be avoided.
 
Last edited:

rulefan

Tour Winner
Joined
Feb 21, 2013
Messages
14,484
Visit site
Yes they are, but why are they not rounded to two decimal places?
This would produce a Handicap Index that is correct to one decimal place instead of being possibly inaccurate by plus or minus one decimal place.

WHS has built-in this inaccuracy and it could very easily be avoided.
Is 1 dec place in 1 case in 10 really significant, given that CH is calculated to be a whole number anyway.
But for players to make sense of your method, differentials would have to be displayed to 2 dec places, unless they recalculate them themselves.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
Is 1 dec place in 1 case in 10 really significant, given that CH is calculated to be a whole number anyway.
But for players to make sense of your method, differentials would have to be displayed to 2 dec places, unless they recalculate them themselves.

Yes it is significant. A slope of 132 makes a CH of 3 for an index of 2.9 but 4 for 3.0 giving me a whole extra shot which would not happen if differentials were more accurately calculated.

No one needs to make sense of my method. It is not my method. It is WHS method.
I am merely pointing out that WHS method is inaccurate due to the possibility of an accumulation of 8 rounding tolerances.
This inaccuracy would not happen if WHS differentials were two decimal places as that would create a handicap index correct to one decimal place.
The current WHS method will always produce some inaccurate handicap indices (plus or minus 0.1) and some of those will have an inaccurate course handicap (plus or minus one whole shot)
Same method, but more accurate arithmetic, would improve the WHS.
 

rulefan

Tour Winner
Joined
Feb 21, 2013
Messages
14,484
Visit site
Yes it is significant. A slope of 132 makes a CH of 3 for an index of 2.9 but 4 for 3.0 giving me a whole extra shot which would not happen if differentials were more accurately calculated.
Both produce the same PH

No one needs to make sense of my method. It is not my method. It is WHS method.
I am merely pointing out that WHS method is inaccurate due to the possibility of an accumulation of 8 rounding tolerances.
This inaccuracy would not happen if WHS differentials were two decimal places as that would create a handicap index correct to one decimal place.
The current WHS method will always produce some inaccurate handicap indices (plus or minus 0.1) and some of those will have an inaccurate course handicap (plus or minus one whole shot)
Same method, but more accurate arithmetic, would improve the WHS.
Of course the rather spurious accuracy of the Course & Bogey Ratings (and therefore Slope) on a particular day mean that Differentials are only approximations anyway. And penalty scores can skew any indication of normal ability.

Interestingly, one objective of the WHS is "to indicate with sufficient accuracy the score a golfer is reasonably capable of achieving .........". Surely whole numbers is sufficient given that scores are in whole numbers.
 

wjemather

Well-known member
Joined
Dec 3, 2020
Messages
3,018
Location
Bristol
Visit site
Yes it is significant. A slope of 132 makes a CH of 3 for an index of 2.9 but 4 for 3.0 giving me a whole extra shot which would not happen if differentials were more accurately calculated.

No one needs to make sense of my method. It is not my method. It is WHS method.
I am merely pointing out that WHS method is inaccurate due to the possibility of an accumulation of 8 rounding tolerances.
This inaccuracy would not happen if WHS differentials were two decimal places as that would create a handicap index correct to one decimal place.
The current WHS method will always produce some inaccurate handicap indices (plus or minus 0.1) and some of those will have an inaccurate course handicap (plus or minus one whole shot)
Same method, but more accurate arithmetic, would improve the WHS.
Unfortunately, you're starting from a flawed premise that any handicap system provides an accurate handicap. They don't. They merely (aim to) provide the best possible approximation of players' ability in order to allow them to complete against each other as equitably as possible. The margin of error in the resultant handicaps is much greater than 0.1, and probably (for many/most) greater than a full stroke.

As such, a 0.1 difference in the result due to the compound effect of rounding is insignificant to the overall result, even if it does occasionally produce a different Course Handicap.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
I suppose it all depends on whether you prefer correct arithmetic or "its not exactly correct all the time, but a few wrong answers don't matter much."
I see no point in "built-in" inaccuracies, when correct arithmetic makes more sense, is not harder to do, and would be correct every time.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
Unfortunately, you're starting from a flawed premise that any handicap system provides an accurate handicap. They don't. They merely (aim to) provide the best possible approximation of players' ability in order to allow them to complete against each other as equitably as possible. The margin of error in the resultant handicaps is much greater than 0.1, and probably (for many/most) greater than a full stroke.

As such, a 0.1 difference in the result due to the compound effect of rounding is insignificant to the overall result, even if it does occasionally produce a different Course Handicap.

I disagree.
I am not starting with any premise whatsoever.

I am pointing out that differentials calculated to one decimal place only produces an inaccurate calculation of an average of 8 differentials.
There is no premise there. It is a statement of fact. The WHS arithmetic of "average of best 8" is flawed.

I have made no premise or judgement about whether the whole system produces a realistic handicap.

I have pointed out inaccurate arithmetic.
 

wjemather

Well-known member
Joined
Dec 3, 2020
Messages
3,018
Location
Bristol
Visit site
I disagree.
I am not starting with any premise whatsoever.

I am pointing out that differentials calculated to one decimal place only produces an inaccurate calculation of an average of 8 differentials.
There is no premise there. It is a statement of fact. The WHS arithmetic of "average of best 8" is flawed.

I have made no premise or judgement about whether the whole system produces a realistic handicap.

I have pointed out inaccurate arithmetic.
Despite your assertions, the arithmetic is neither inaccurate nor incorrect; it is merely subject to various rounding errors at several stages of the process. Not rounding at one step does not undo the rest of the rounding.

Firstly, the the ratings upon which the differentials are derived are themselves subject to rounding (this is probably the most significant); then score differentials are rounded; then the Handicap Index is rounded; then the Course Handicap is rounded; and finally the Playing Handicap is rounded. You are focussing on the step where rounding possibly has the least significance.

When a 0.1 (or even 0.01 ;)) difference in index regularly results in the same one stroke difference on the course as a 1.9 difference in index, and the average player can easily shoot ten strokes different from one day to the next, the overall significance of such rounding errors is inconsequential.
 

harpo_72

Journeyman Pro
Joined
Feb 20, 2013
Messages
5,503
Visit site
Well result of the match play came in the low chap was 1 over gross and -1or 2 to handicap and lost 5&3. Other chap played out of his skin ???
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
Despite your assertions, the arithmetic is neither inaccurate nor incorrect; it is merely subject to various rounding errors at several stages of the process. Not rounding at one step does not undo the rest of the rounding.

Firstly, the the ratings upon which the differentials are derived are themselves subject to rounding (this is probably the most significant); then score differentials are rounded; then the Handicap Index is rounded; then the Course Handicap is rounded; and finally the Playing Handicap is rounded. You are focussing on the step where rounding possibly has the least significance.

When a 0.1 (or even 0.01 ;)) difference in index regularly results in the same one stroke difference on the course as a 1.9 difference in index, and the average player can easily shoot ten strokes different from one day to the next, the overall significance of such rounding errors is inconsequential.

This where we will differ. I believe this step is very significant.
It does produce an accumulation of 8 rounding errors so that the average will not be accurate to one decimal place every time.
If differentials were always two decimal places then the average of 8 would be accurate to one decimal place every time.
The arithmetic is flawed and inaccuracies occur which could be easily avoided.
I do not believe this step is made to be insignificant by the fact that course ratings and slope ratings are involved in the calculations of the differentials.
 

Swango1980

Well-known member
Joined
Jul 22, 2019
Messages
10,643
Location
Lincolnshire
Visit site
This where we will differ. I believe this step is very significant.
It does produce an accumulation of 8 rounding errors so that the average will not be accurate to one decimal place every time.
If differentials were always two decimal places then the average of 8 would be accurate to one decimal place every time.
The arithmetic is flawed and inaccuracies occur which could be easily avoided.
I do not believe this step is made to be insignificant by the fact that course ratings and slope ratings are involved in the calculations of the differentials.
Are you not assuming you only round on one direction all 8 times? Sometimes you'll round down to the one decimal place, sometimes up.

It is not right to say WHS is ALWAYS 0.1 out every time. In fact, it is likely to be very rare.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
Are you not assuming you only round on one direction all 8 times? Sometimes you'll round down to the one decimal place, sometimes up.

It is not right to say WHS is ALWAYS 0.1 out every time. In fact, it is likely to be very rare.

I have not assumed anything and I did not say the calculation is ALWAYS out every time.
I said it would not be accurate every time.
Occasionally there will be an inaccuracy.
There will be some plus rounding errors and some minus rounding errors.
When there is far more of one than the other, then the average of 8 calculation is prone to the inaccuracy of plus or minus 0.1.

I have stated nothing but facts about the average of 8 calculation. No assumptions.
8 figures to one decimal place that have rounding errors of plus or minus 0.05 can and will create an average that has an accuracy of plus or minus 0.1.
The vast majority will be spot on.
Some will be in error by +0.1 and some by -0.1.

All this could be corrected by retaining two decimal places for every differential and rounding the average to one decimal place.
 

Swango1980

Well-known member
Joined
Jul 22, 2019
Messages
10,643
Location
Lincolnshire
Visit site
I have not assumed anything and I did not say the calculation is ALWAYS out every time.
I said it would not be accurate every time.
Occasionally there will be an inaccuracy.
There will be some plus rounding errors and some minus rounding errors.
When there is far more of one than the other, then the average of 8 calculation is prone to the inaccuracy of plus or minus 0.1.

I have stated nothing but facts about the average of 8 calculation. No assumptions.
8 figures to one decimal place that have rounding errors of plus or minus 0.05 can and will create an average that has an accuracy of plus or minus 0.1.
The vast majority will be spot on.
Some will be in error by +0.1 and some by -0.1.

All this could be corrected by retaining two decimal places for every differential and rounding the average to one decimal place.
You said, direct quote:

"It does produce an accumulation of 8 rounding errors so that the average will not be accurate to one decimal place every time."

I read that as every single time, the average will not be accurate. What you probably meant was there will be occasions it may not be accurate.

I believe others have more than countered your debate, given the much bigger inaccuracies within handicap calculations. Personally, I am guessing out of a few hundred million golfers worldwide, you may be the only one concerned by this particular issue. You could email WHS support to voice your concerns, maybe they'll change it if they think you have a point. I reckon there could be higher things up the list when they next get round to making any changes.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
You said, direct quote:

"It does produce an accumulation of 8 rounding errors so that the average will not be accurate to one decimal place every time."

I read that as every single time, the average will not be accurate. What you probably meant was there will be occasions it may not be accurate.

I believe others have more than countered your debate, given the much bigger inaccuracies within handicap calculations. Personally, I am guessing out of a few hundred million golfers worldwide, you may be the only one concerned by this particular issue. You could email WHS support to voice your concerns, maybe they'll change it if they think you have a point. I reckon there could be higher things up the list when they next get round to making any changes.
I have no concerns so I will not be voicing any. I believe all I have done here is state the facts of the arithmetic in order to inform others of these facts.
I do not accept "countered your debate" when I have not debated - I have stated facts.
Should anyone wish to ignore or disbelieve these facts, then that is entirely up to them and this also does not concern me.
 

Backsticks

Assistant Pro
Joined
Aug 7, 2012
Messages
3,685
Visit site
Whilst your facts are correct, they are irrelevant and their effect is lost in all the other calculations of handicapping. Handicaps are a good example of 'false precision'.

Hear hear. Thats the bottom line really. Talk of 'precision' to a decimal point of a rating of a persons golfing level relative to a field is pointless. It is so inaccurate anyway, that concerns about a possible occasional computation 'error' of 0.1 has no real influence on the outcome. That it might change the result is not the point - the point is that the possible result change is within the handicap accuracy noise range anyway.
 

Voyager EMH

Slipper Wearing Plucker of Pheasants
Joined
Mar 14, 2021
Messages
5,141
Location
Leicestershire
Visit site
I have an official handicap index of 3.0.
I know that with more accurate calculation of the average of best 8 differentials (no other calculations involved) this would give the correct answer of 2.9.

At my course with a slope rating of 132, I have a course handicap of 3 if my index is 2.9 and a course handicap of 4 if my index is 3.0.
A whole extra shot because of the poor arithmetic of the average of best 8 calculation.
There will be others in this position around the country.
This could cause trophies to be wrongly won. (Is that irrelevant?)

Differentials retaining 2 decimal places would remove these errors in the average of 8 calculation.

All of the above is factual. It is not irrelevant.

Anyone referring to other the other calculations is obfuscating.

An arithmetical error is an error.
If you choose to ignore this, then that is up to you.
I prefer arithmetic that does not give wrong answers.

I have no problem accepting my official handicap. This would also be true if it were lower than it should be.
If all the facts I have set out are irrelevant, then so are all the responses. I do not believe the previous sentence to hold any truth.
Thank you for all your responses and comments. Most stimulating.
 
Top