Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Council meeting
#31
"2244 - significantly higher than live grade."
6 rating points is significant?? News to me!
Reply
#32
Phil Thomas Wrote:Let us do a mind experiment. Suppose Murad played another 4 congresses and scored identical results to the first 4. His strength would then be evaluated from these 8 numbers.
1663, 2065, 1654, 2180, 1663, 2065, 1654, 2180.
The correct way to do this would be to average all 8 – result is still 1891.
The incorrect way to do this is to average the highest 4 and get 2123.

Where did the extra 232 points come from ?
Its not magic its a serious mathematical flaw.
Cherry picking data this way gives, in effect, a set of corrupt raw data.

If you are going to only use 4 results out of the 8 it would be better to use the middle 4 results rather than the highest 4 results. This gives 1864.

There are pros and cons to using all of the results or only some of them. On the plus side using all of the results may be more "accurate". On the minus side taking an average(arithmetic mean) is vulnerable to being excessively influenced just by one very low or very high result.

Perhaps you should do the same as is proposed for the selectors' estimates and exclude the lowest and the highest and take the average of the rest - just using 4 out of the 8 results is probably ignoring too much data.

On the subject of very low or high results there are problems using TPRs. Most congresses are probably only 5 rounds whereas taking a lead from the CS grading system at least 8 would be better. This is less of a problem though if the average of several TPRs is taken. There are also problems with TPRs not existing or not being meaningful if someone scores either 0% or 100% cf. the CS grading system requiring a new player to score at least 1/2 point (and presumably not winning all their games although it does not appear to say this anywhere.) While 0% is probably not a problem for the players under consideration 100% could well be.
Reply
#33
hamish olson Wrote:"2244 - significantly higher than live grade."
6 rating points is significant?? News to me!

You are absolutely right Hamish.
I used the 2175 figure in the motion rather than the 2138 figure on the notice board.My mistake


Taking my thesis a little further

(1) If you include all congresses and not just some
(2) If you weight tournament results in proportion to the number of games played.
(3)If you include league results again weighted according to the number of games played.
(4) If you also include any other graded games played in that season.

Then you have reinvented the grading system.


Why stop part way through part 1 ???
Reply
#34
Ian Jamieson Wrote:
Phil Thomas Wrote:Let us do a mind experiment. Suppose Murad played another 4 congresses and scored identical results to the first 4. His strength would then be evaluated from these 8 numbers.
1663, 2065, 1654, 2180, 1663, 2065, 1654, 2180.
The correct way to do this would be to average all 8 – result is still 1891.
The incorrect way to do this is to average the highest 4 and get 2123.

Where did the extra 232 points come from ?
Its not magic its a serious mathematical flaw.
Cherry picking data this way gives, in effect, a set of corrupt raw data.

If you are going to only use 4 results out of the 8 it would be better to use the middle 4 results rather than the highest 4 results. This gives 1864.

There are pros and cons to using all of the results or only some of them. On the plus side using all of the results may be more "accurate". On the minus side taking an average(arithmetic mean) is vulnerable to being excessively influenced just by one very low or very high result.

Perhaps you should do the same as is proposed for the selectors' estimates and exclude the lowest and the highest and take the average of the rest - just using 4 out of the 8 results is probably ignoring too much data.

On the subject of very low or high results there are problems using TPRs. Most congresses are probably only 5 rounds whereas taking a lead from the CS grading system at least 8 would be better. This is less of a problem though if the average of several TPRs is taken. There are also problems with TPRs not existing or not being meaningful if someone scores either 0% or 100% cf. the CS grading system requiring a new player to score at least 1/2 point (and presumably not winning all their games although it does not appear to say this anywhere.) While 0% is probably not a problem for the players under consideration 100% could well be.


Ian just one thing I wish to add to your posting. It has been discussed recently elsewhere but is relevant so no apologies for repetition.

If a player benefitting from the 200 up rule beats a player of grade 400 below his/her grade then the higher rated player will loses grading points. The fair thing to do would be to ignore that game when calculating the grade change of the stronger player. However, if I remember that debate correctly this not an easy thing to do - to have the game graded for one player and not the other.

You may see some examples of this anomaly in the Primary Individual results from 2013. To reduce the number of such cases we ran the primary individual as an accelerated event for the first 2 or 3 rounds this year.
Reply
#35
Phil Thomas Wrote:For Monica on the grading database I can see, today ,4 tournaments with grading results of 1663(Ayr) 1347 (Edinburgh) 1621 (Glenrothes) 1375 (Girls Championship).


Please don't bring individual cases into this, we don't want anyone's name to be dragged through the mud.


Firstly, with regards to this bad data as such I disagree. Obviously, the data becomes bad when you make it up to be such but, in general, higher rating performances come as the player becomes older and better. As selectors, we do take into account when a performance is recorded. This isn't such a bad idea to measure a current junior's strength, you can think of it as a live rating over a smaller period.

Secondly, live ratings aren't the only criteria, of course.

Live ratings aren't particularly accurate in terms of a ranking for junior players, their strength can be badly hidden by a low/high rating. I can explain a couple of decent reasons for this.

It's possible to get a 2000 performance by scoring 5/5 vs 1600s or 2.5/5 vs 2000, with the latter being harder.

Playing abroad also gives an inflation as these guys tend to have higher ratings for the same ability. This can be seen very clearly when comparing results of players up to a very high (23/2400) level. Note that there are quite a few top players with discrepancies between their Fide, usually achieved abroad, and their Scottish rating. This effect gets even more significant lower down the rating scale.

While I don't really want to get into selection policy I feel something must be said considering there has been some, unjust in my opinion, criticism of policy and selectors.
Reply
#36
Calum MacQueen Wrote:
Phil Thomas Wrote:


Please don't bring individual cases into this, we don't want anyone's name to be dragged through the mud.

Calum
Please note that I am quoting from a motion that has been put forward for discussion at the council meeting.
And given the complexity of the council meeting agenda it is a motion that it likely to be discussed again at the agm

The real question here is why are individual players and their results mentioned in the motion ??
Reply
#37
Phil Thomas Wrote:
Calum MacQueen Wrote:
Phil Thomas Wrote:


Please don't bring individual cases into this, we don't want anyone's name to be dragged through the mud.

Calum
Please note that I am quoting from a motion that has been put forward for discussion at the council meeting.
And given the complexity of the council meeting agenda it is a motion that it likely to be discussed again at the agm

The real question here is why are individual players and their results mentioned in the motion ??


True enough but this obviously is just an example. I feel the real question is why you've jumped on it to have a go at selector's impartiality?
Reply
#38
Calum,

I have attempted on this thread to attack the mathematics of the motion and not to enter into the topic of selection processes that have already happened.

To use the method outlined in the motion, as I stated earlier, is within the remit of the IJD and as such does not need permission from council or agm.

Going back to the maths. I hear from a reliable source than the tournament performance in the worked example for Murad the numbers used are not real tournament numbers - they are merely examples to show how the new system works.

That is the mathematical equivalent of Nick Park stating that the moon is made of Wensleydale cheese; and quoting as evidence his 1989 film.


As for why me?

I respond with another question Why not? It could have been anyone from amongst the many notice board readers who is in the subset that feel competent enough to post on the mathematics of grading systems.
Actually to do so I need also be in the subset of those who feel able to handle potential notice board abuse. To misquote David Deary it takes guts to post on this board and it takes guts to complain to moderators about unacceptable posts. To help David's thought processes rather then blame the board moderators perhaps that anonymous person has a high strike rate because the posts attacking him/her are more worthy of deletion?

For the record Calum I find your posts polite, totally valid, and a positive contribution to notice board debate. They reflect your character away from the notice board.

A notice board that is far from dead. I counted yesterday 16 threads that had been active within the last 7 days. Many more and this board will start to look like the ECF forum - now would that be a good thing? - new thread anyone?
Reply
#39
No CS director should threaten legal action against another - it's only chess we're playing - not life or death
Reply
#40
Phil Thomas Wrote:Going back to the maths. I hear from a reliable source than the tournament performance in the worked example for Murad the numbers used are not real tournament numbers - they are merely examples to show how the new system works.

That is the mathematical equivalent of Nick Park stating that the moon is made of Wensleydale cheese; and quoting as evidence his 1989 film.

Or, an alternative view is that it is an example... Not using data in the grading system... You know, that thing you complained about here.

Phil Thomas Wrote:The real question here is why are individual players and their results mentioned in the motion ??

You're unhappy when real data is used and unhappy when data is made up. Makes about as much sense as choosing the 10th highest rated player as an indicator of age group performance.

Phil Thomas Wrote:I respond with another question Why not? It could have been anyone from amongst the many notice board readers who is in the subset that feel competent enough to post on the mathematics of grading systems.

Firstly, I'm not sure how the general Chess Scotland public feel about this one. Do you really think the majority of people don't understand how using the mean as an average works??? Because that is exactly what the method is.

Secondly, why you? It was you, it wasn't anyone else. And it is always you. I've lost count of the amount of people you've fallen out with by directly attacking/criticising them. Considering you've shown blatant disdain towards Chess Scotland by withdrawing your membership, I'm amazed you find the will to come on the board and complain about things such as this.

If you don't want to help Chess Scotland move on as an organisation then leave the volunteers who are giving plenty of their time and effort to help. If you do, get a membership and say something positive about something for once. We recently had your wife ranting about how an extra £5 per head was bad value despite the new digs for the Liverpool tournament didn't have a tramp staying there and the beds had that added bonus of not being laced with the previous visitor's public hair. Trying to criticise everything is just a terrible policy and one we cannot have in a community that is trying to pull together to get the grant back.

Just as a little post script, do you not think it's deeply, deeply inappropriate to bring up this example considering the email that recently went around. The one you clearly know about considering the only person to be CC'ed into the original email was your wife?
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)