CSS Farce

[Imurg]

If everybody played 'rubbish', there must have been something difficult about the course that day, such as hard fast running greens or awkward pin positions, even if the weather was benign. We had one qualifier at our club that was like that earlier this summer. Provided you had a large enough field it is statistically unlikely that they all played badly in easy weather conditions, hence the CSS would be expected to go up.

CSS assumes that everyone is playing the same, relative to their handicap. If everyone played badly it is assumed that the course set up is tricky or the weather bad. Likewise if everyone plays well it is assumed to be good weather and/or an easy set up.
On any given day, any of us can play well, mediocre or badly - regardless of the weather, set up or how anyone else is playing.
My handicap is an indication of my ability to play against the course. It's only relative to anyone else if we all play to our handicaps - but we have good days, bad days and mediocre days.
Our initial handicaps have no relation to how anyone else plays. It is based on a fixed number, SSS. It could be a rank, rotton day with high winds but if that's the day you put your best of 3 cards in then that's the card that determines your handicap. Everyone else might NR or shoot a cricket score but that doesn't affect you.
If you put a Supplementary card in, SSS is used. If everyone playing a Comp didn't enter the Comp but entered a Supplementary then SSS would be used for each one.
If the 2 Guys I'm playing with enter the Comp but I put a Supplementary in, their handicap gets adjusted on the comp CSS and mine uses SSS and yet we've played at the same time, under the same conditions on the same course.......
If that doesn't make it a farce then there's no hope...

 

[Ethan]

CSS assumes that everyone is playing the same, relative to their handicap. If everyone played badly it is assumed that the course set up is tricky or the weather bad. Likewise if everyone plays well it is assumed to be good weather and/or an easy set up.
On any given day, any of us can play well, mediocre or badly - regardless of the weather, set up or how anyone else is playing.
My handicap is an indication of my ability to play against the course. It's only relative to anyone else if we all play to our handicaps - but we have good days, bad days and mediocre days.
Our initial handicaps have no relation to how anyone else plays. It is based on a fixed number, SSS. It could be a rank, rotton day with high winds but if that's the day you put your best of 3 cards in then that's the card that determines your handicap. Everyone else might NR or shoot a cricket score but that doesn't affect you.
If you put a Supplementary card in, SSS is used. If everyone playing a Comp didn't enter the Comp but entered a Supplementary then SSS would be used for each one.
If the 2 Guys I'm playing with enter the Comp but I put a Supplementary in, their handicap gets adjusted on the comp CSS and mine uses SSS and yet we've played at the same time, under the same conditions on the same course.......
If that doesn't make it a farce then there's no hope...

Well, it is assumed that with a large data set, more predictable patterns emerge.

The reason for SSS to be used for supplementaries is that there is usually no large group on whom to base a CSS. The problem you describe is that players are allowed to play a supplementary within a comp; leading to an anomaly. Perhaps playing a sup on a comp day shouldn't be allowed. It doesn't undermine the idea of adjusting SSS based on varying scoreability, though.

 

[chrisd]

It doesn't undermine the idea of adjusting SSS based on varying scoreability, though.

But it is precisely this that we are against.

 

[Backache]

Not sure if it has been posted before as I have not been a reader long but the SGU shows some of the research that has been done around handicaps and why the current system although not perfect is probably a lot less imperfect than some suggest.

http://www.scottishgolf.org/wp-content/uploads/Myths-and-Misconceptions1.pdf

 

[Ethan]

But it is precisely this that we are against.

And you are entitled to be against it, but the use of SSS for supplementaries compared to CSS for comps is not a very good argument against CSS. it is an excellent argument about not allowing supplementaries on comp days, though, or using CSS for supplementaries played on comp days.

The whole point of handicaps is to allow comparability between players within one club and between players at different clubs. In order to achieve this, you need to benchmark as many variables as possible, and weather and course set up are two important variable aspects and CSS tries to deal with those. Baseline cause difficulty is obviously another, and the base SSS takes some account of that. Without CSS you would have players who choose to play only on good days with benign scoring conditions becoming separated from those who play any day and their handicaps would start to become incomparable. CSS is one way of reducing this separation. It isn't perfect, but it is better than nothing.

 

[rosecott]

The bottom line is surely that the derivation of the current SSS/CSS system is based on a wealth of actual data and not on the personal ideas/opinions of individual players.

 

[FairwayDodger]

Not sure if it has been posted before as I have not been a reader long but the SGU shows some of the research that has been done around handicaps and why the current system although not perfect is probably a lot less imperfect than some suggest.

http://www.scottishgolf.org/wp-content/uploads/Myths-and-Misconceptions1.pdf

Much like the CONGU report I read, this is a numerical analysis that does not discuss the playing conditions on dates when CSS moves. It makes a sweeping assumption that conditions are "better" during the summer months to conveniently explain a higher proportion of CSS-1 comps. That is flawed for a number of reasons, for example I'd expect people to score better because most members are playing and practising more at that time of year.

Their other analysis evens shows that cat 3 scores are slightly more likely to contribute to CSS-1. Of course this is golf so the analysis has been based entirely on male stats so cat 4 has not been included.

 

[FairwayDodger]

The bottom line is surely that the derivation of the current SSS/CSS system is based on a wealth of actual data and not on the personal ideas/opinions of individual players.

Numerical analysis only that ignores playing conditions on the day. I'm often bemused by the CSS when it bears no relation to difficulty on the day. Plenty of others seem to share my opinion.

The other problem with stats is you need a decent sample size to have any chance of them being meaningful. I'd contend that in women's club golf, in particular, it's pretty unusual to get a large enough field to remove the randomness. 32 people is not the biggest entry we get but it was certainly pretty good, above average I'd say.

 

[FairwayDodger]

Well, it is assumed that with a large data set, more predictable patterns emerge.

But CSS does not require a large data set. Just 7 or more players, I believe, so of no statistical worth whatsoever at that end if the scale.

The calculation of smaller fields than that is even more arbitrary but for a reasonable reason (to reduce the number of RO comps) and at least does not allow CSS-1.

 

[HawkeyeMS]

The bottom line is surely that the derivation of the current SSS/CSS system is based on a wealth of actual data and not on the personal ideas/opinions of individual players.

I don't dispute that it is based on a lot of statistics, I just don't see the point of it.

One of our comps a couple of weeks ago the CSS went up to 71. A quick look at the scores tells me that if we had used SSS about 5 fewer people would have been cut and 8 more would have gone up 0.1 out of about 130 entries - so what's the point?

 

[Backache]

Much like the CONGU report I read, this is a numerical analysis that does not discuss the playing conditions on dates when CSS moves. It makes a sweeping assumption that conditions are "better" during the summer months to conveniently explain a higher proportion of CSS-1 comps. That is flawed for a number of reasons, for example I'd expect people to score better because most members are playing and practising more at that time of year.

Their other analysis evens shows that cat 3 scores are slightly more likely to contribute to CSS-1. Of course this is golf so the analysis has been based entirely on male stats so cat 4 has not been included.

The scores peaked in the summer and although people are probably more practised in the summer than in the spring that is not true compared with the autumn when scores also went up.
They are not making a detailed analysis of every competition but showing trends and to make handicaps equitable it is probably better to have a CSS than a SSS on average, this will not be the case every time but over time it should average.

Your comments about it being an analysis ony up to category three are fair enough , but in order to say it is different for category four you would have to show that worsening weather conditions tend to narrow the returns as lesser golfers are less affected by them. My own limited experience of womens category 4 gofers suggests this is not the case.

 

[rosecott]

Numerical analysis only that ignores playing conditions on the day. I'm often bemused by the CSS when it bears no relation to difficulty on the day. Plenty of others seem to share my opinion.

The other problem with stats is you need a decent sample size to have any chance of them being meaningful. I'd contend that in women's club golf, in particular, it's pretty unusual to get a large enough field to remove the randomness. 32 people is not the biggest entry we get but it was certainly pretty good, above average I'd say.

I would certainly agree that small fields are a problem and that this is something that is particularly problematic in the women's game.

I would, however, be more inclined to rely on the stats of the day rather than opinions on how easy/difficult it was to score well/badly. I know that I am often bemused when I score poorly but a significant number of others score well - and vice versa.

 

[FairwayDodger]

I would certainly agree that small fields are a problem and that this is something that is particularly problematic in the women's game.

I would, however, be more inclined to rely on the stats of the day rather than opinions on how easy/difficult it was to score well/badly. I know that I am often bemused when I score poorly but a significant number of others score well - and vice versa.

As a woman, all my experience of CSS is from women's competitions and my anecdotal comment is that CSS is almost completely unpredictable based on the playing conditions on any particular day.

Our monthly medals are played over two days. Each day gets a separate CSS. Pin positions are the same and, often, weather conditions are comparable. If CSS was accurate and meaningful and successful in its objectives the CSS calculation should give the same number both days (on the occasions when the weather is the same). It often isn't, because it depends on too many random human factors of people playing well or badly just because, well....golf!!

Anecdotally again, I have observed the situation where we've had good weather conditions one of the days and foul weather the other and yet CSS has gone up in the good conditions and down in the nasty ones. Random.

Undoubtedly bigger fields smooth it out but how many players do you need to eliminate the impact of a freak group of scores?

 

[Backache]

As a woman, all my experience of CSS is from women's competitions and my anecdotal comment is that CSS is almost completely unpredictable based on the playing conditions on any particular day.

Our monthly medals are played over two days. Each day gets a separate CSS. Pin positions are the same and, often, weather conditions are comparable. If CSS was accurate and meaningful and successful in its objectives the CSS calculation should give the same number both days (on the occasions when the weather is the same). It often isn't, because it depends on too many random human factors of people playing well or badly just because, well....golf!!

Undoubtedly bigger fields smooth it out but how many players do you need to eliminate the impact of a freak group of scores?

In statistics the answer is never a given number, it is a number to give an x% chance to be within y% of the mean or something like that.
In golf you get bad breaks the alternative problem is that if you do not have a CSS people who wish to protect their handicap don't play in bad weather and you get biased handicaps that way.

I don't think anyone is suggesting that there are no vagaries in the handicapping system, just that alternative methods suggested tend to result in bigger vagaries.

 

[nickjdavis]

I don't dispute that it is based on a lot of statistics, I just don't see the point of it.

One of our comps a couple of weeks ago the CSS went up to 71. A quick look at the scores tells me that if we had used SSS about 5 fewer people would have been cut and 8 more would have gone up 0.1 out of about 130 entries - so what's the point?

13 people (10% of the field) would leave the club with higher handicaps.

5 people who perhaps might have been working hard on their game would be denied the reward of a cut.

8 potential sandbaggers would have moved another 0.1 closer to a higher handicap in readiness for that "big open comp that's coming up in a couple of months".

 

[chrisd]

Tattoo

13 people (10% of the field) would leave the club with higher handicaps.

5 people who perhaps might have been working hard on their game would be denied the reward of a cut.

8 potential sandbaggers would have moved another 0.1 closer to a higher handicap in readiness for that "big open comp that's coming up in a couple of months".

But only as a result of that competition, next weeks could have very different results

 

[FairwayDodger]

In statistics the answer is never a given number, it is a number to give an x% chance to be within y% of the mean or something like that.
In golf you get bad breaks the alternative problem is that if you do not have a CSS people who wish to protect their handicap don't play in bad weather and you get biased handicaps that way.

I don't think anyone is suggesting that there are no vagaries in the handicapping system, just that alternative methods suggested tend to result in bigger vagaries.

Hard to see anything producing a bigger vagary than a system that makes an arbitrary adjustment to your handicap based on how well other people played!

I realise one reason in support of CSS is to encourage people to play in bad conditions but in many ways that's a separate issue. If individuals self-select to only play in good weather that's up to them and plenty choose that option anyway, with CSS and handicaps the least of their worries.

Those who do it to protect their handicap are only hurting themselves when they are forced to play some competition in bad weather and they are not accustomed to it.

 

[nickjdavis]

Tattoo

But only as a result of that competition, next weeks could have very different results

Two wrongs don't make a right.

At the end of the day the handicapping system is what it is... its not perfect, but its not a system that was plucked out of thin air by some old buffoon who never plays golf... there is some logic and structure behind it.

If CSS varies from SSS on a regular basis....I hear some folks, not just here but at other clubs, tell of it varying almost weekly, then Id suggest that there is something amiss with the handicaps at that club. There are statistical models that accurately represent the scoring patterns expected of a field of golfers... sure a group of high handicappers will have a much more varied scoring distribution than a similarly sized group of Cat1 golfers but it should be a rare occurrence where a large proportion of them shoot the lights out in the same competition.

It will be interesting to see what 2016 brings in terms of how the Cat 4 scores will affect CSS...but one thing I am certain of is that the "lookup tables" used to determine the CSS will have been tweaked to take account of the scoring variability of this group.

 

[FairwayDodger]

Two wrongs don't make a right.

At the end of the day the handicapping system is what it is... its not perfect, but its not a system that was plucked out of thin air by some old buffoon who never plays golf... there is some logic and structure behind it.

And a lot of flawed assumptions.

If CSS varies from SSS on a regular basis....I hear some folks, not just here but at other clubs, tell of it varying almost weekly, then Id suggest that there is something amiss with the handicaps at that club. There are statistical models that accurately represent the scoring patterns expected of a field of golfers... sure a group of high handicappers will have a much more varied scoring distribution than a similarly sized group of Cat1 golfers but it should be a rare occurrence where a large proportion of them shoot the lights out in the same competition.

Not necessarily the handicaps, might just be that the average field size is too small to eliminate natural variance in players' performance. Somebody smarter than me would have to determine the minimum field size needed to compensate for that but it's obviously much higher than we tend to get.

Remember, you don't need to shoot the lights out to affect CSS - just play in your buffer.

It will be interesting to see what 2016 brings in terms of how the Cat 4 scores will affect CSS...but one thing I am certain of is that the "lookup tables" used to determine the CSS will have been tweaked to take account of the scoring variability of this group.

That should already be known since cat 4 players are already used in calculating the CSS for women's comps. I suspect you'll find men's CSS will start to become more volatile. As for us, if Cat 5 handicaps are going to count, I dread to think!

 

[FairwayDodger]

It occurred to me that we have a few days on which both men's and women's competitions take place so a comparison of each CSS might be informative. If CSS was a fair and accurate representation of difficulty then surely the difference between SSS and CSS for both men and women would tend to match?

From the results still accessible to me on the website, there were 8 such days in the last year. This image shows the analysis, I haven't drawn any conclusions from it but it seems things worked out the same only 3 times out of the 8. Interesting but fairly inconclusive without a much larger sample set, I think.