Surrey's motion on grading for Council

[Jonathan Rogers]

I agree with Richard and the unnamed junior organiser. My first thought in August, when I saw several juniors who seemed blatantly 40 or more points too high, was that this is primarily a disaster for them.

Just imagine, having a grade at (say) 14 years old which you know already will probably be the highest you will ever have, and that you could spend years playing as an adult and never better it? Is that really an encouragement to play? The nice feeling about having a high grade in the here and now will soon rub off - many a time the junior loses badly to someone supposedly 20 points lower (but really, much stronger) there will be sarcastic remarks, if not from the opponent himself then from one of his badly behaved companions. (We have spoken before about how some - few, but some - adults revel in the misfortunes of juniors. The extent of the problem may not be so very serious at the moment, I would guess, but giving the juniors highly inflated grades is sure to make the problem worse).

[Kevin Thurlow]

I understand the problem with the new list and do understand Jonathan and Richard's point -I should clarify that I was thinking back to a regular Rapidplay event where a junior had dominated the U120 section for several months then turned up when the new list had just been published, putting them in the U160 and the immediate reaction was "I'm not playing in that - I might not win". So my problem was more with people only wanting to play when they won... For the record, the junior did play and got 4.5/6 in the U160.

But of course I totally agree that if anyone (not just juniors) has a totally wrong grade, it causes problems. It is ironic that the grades have presumably been fiddled to help juniors and it may not be working...

[Roger de Coverly]

It is ironic that the grades have presumably been fiddled to help juniors and it may not be working...

Talking of which , does anyone know whether the 57% / 10 point observation is before or after junior increments? I'm thinking that if it's after the junior increment that the junior increment itself reduces the observed percentage. Let's suppose we don't have a junior increment. Then you might have juniors of 130 grade but 140 standard playing 120 standard adults. You might expect them to score at least 60% and probably better. So is there a "pulling down" effect if you include the increment in the statistics?

I think it's worth visiting some of the fundamentals of why you have a junior increment. Let's just suppose you didn't have junior increments but that young players come into the system at say 50 points worse than their ultimate stable strength. Lets say it was the 120 to 170 range. Obviously the juniors do a bit better than their published grade bu over time the lags in the system will unravel and they will eventually end up alongside the 170 players of a few years earlier. The problem is that the 170 players are down to 165 and the 120 players are down to 115. So you need a corrective factor in the interests of stability and historic continuity. The original +10 used in the seventies and beyond was felt to be too high by the late eighties and replaced by an age related scale. It's really only because it would have been difficult to use as a two entry table that it was added to the last year grade. I don't think any particularly strong predictive value should be claimed
for the junior increment and it's wrong for the grading team to place so much emphasis on it. In fact they've broken one of the underlying premises of grading systems - hence Surrey's motion.

[Mike Gunn]

No other county has come forward to say that will support the motion. so it is about to fall into the abyss ... 24 hours left to help put the motion on the agenda (i.e. until 9.30am on Wednesday).

[Roger de Coverly]

Another thought on this.

If you always treat juniors as new players, then it makes a junior only event impossible to grade unless later or earlier in the season, one or more of the players happens to play an adult. Even so, it could make the absolute grades of a number of players very dependent on a small number of results - it could be as few as one. Does this account for some of the volatility we are seeing in junior grades? So whilst the relative grades of a school league would be consistent with their results, the absolute grades might depend on a handful of games played by a handful of the players in adult events - so a single blunder could change a dozen or more player's grades by 10 points or more.

[Brian Valentine]

Roger,
You are right. This affect can be reproduced in my spreadsheet accessible on another thread.

[E Michael White]

Another thought on this.

If you always treat juniors as new players, then it makes a junior only event impossible to grade unless later or earlier in the season, one or more of the players happens to play an adult. Even so, it could make the absolute grades of a number of players very dependent on a small number of results - it could be as few as one. Does this account for some of the volatility we are seeing in junior grades? So whilst the relative grades of a school league would be consistent with their results, the absolute grades might depend on a handful of games played by a handful of the players in adult events - so a single blunder could change a dozen or more player's grades by 10 points or more.

You're right on this which is what I referred to as pyramiding in an earlier post, or the grading equivalent of chain letters. Its possible to get very high junior grades by accident eg 300+ or even higher.