Nigel Short wishes Gerry Walsh 'Merry Xmas & Begone!'

[Ben Purton]

See now thats a more direct quote...thanks for point that out duck

[Justin Hadi]

Good. It's difficult to trawl through 10+ pages of crap to find the one message you're interested in. This will need support from the other voting blocks if it's to go through. Another point - what is to stop GW standing for CEO if the motion goes through?

Justin

[Justin Hadi]

Good work Ernie. It's nice to see some progress is being made. Next question - who hold the voting rights, who are the proxies (Gerry's mates?) ,and how can they be lobbied? I don't think it's an exaggeration to say that if proportional representation OMOV was in force, and the electorate were informed of the true incompetence of GW given a credible alternative which didn't hit their pockets, there would be a landslide against him. How do we get round the block voting?

Justin

[Sean Hewitt]

Justin,

I posted a breakdown of the voting rights earlier in the thread here.

As you can see there are alot of institutions with a small number of votes each (typically 1 or 2) so there is alot of lobbying to be done.

[Justin Hadi]

Sean,

I meant who individually, rather than the organisation. Who do we speak to?

[Sean Hewitt]

My guess is it's the individual rather than the organisation that usually decides, although some organisations will mandate their delegate to vote a certain way on certain issues.

As far as I know, the only way to identify delegates is to dig out the latest ECF yearbook and trawl through each organisation listed. Thats ok for leagues, counties and unions. I have no idea how you would find out a congresses delegate.

[Mick Norris]

I have no idea how you would find out a congresses delegate.

The ECF have a list

[Sean Hewitt]

The ECF have a list

Of course, but I dont know if they would [be able to] release it.

[Sean Hewitt]

Sean. Can you clarify something. Does a congress organiser have to be a member of the ECF to get a vote at Council or is it simply because he runs a congress he gets a vote?

No council member, whether they be representing a congress, league or county has to be a member of the ECF to vote on council. I guess this is because the ECF is a federation of organisations - not an individual membership organisation.

A congress gets a number of votes dependent on how much game fee they pay, one vote per 1000 results (or part thereof). If the congress is not game fee registered it gets no votes.

Is Gerry as President/Chairman able to hold proxy votes from Congress organisers and use those proxys to keep himself in a job if theres a vote.

If a congress organiser (or league / county / union etc) wishes to appoint Gerry (or anyone else) as their proxy they are perfectly at liberty to do so. They can direct their proxy to vote a certain way on a certain issue if they wish, or they can leave it to the proxy to vote whichever way they choose. Its entirely a matter for the delegate appointing the proxy.

[Sean Hewitt]

With an increasing frequency of appearance of rating lists, I fully understand that there is a sound mathematical argument for increasing the K factor - although whether it should it be to 12, 15, 20, or 30 is anybody's guess.

Nigel is absolutely correct - so what should the K factor be increased to?

In my opinion, you have to understand what the factor is seeking to achieve before you can answer this. I believe that its purpose is to control how quickly a players rating converges with his current playing strength. Put it another way - how many games should a player have to player at a given playing level before his rating is equal to that level? Answer that question, and you can work out what K should be.

So, if a player rated 2400 starts playing at 2500 level how many games should he have to play before he is rated 2500?

[Maxim Devereaux]

With an increasing frequency of appearance of rating lists, I fully understand that there is a sound mathematical argument for increasing the K factor - although whether it should it be to 12, 15, 20, or 30 is anybody's guess.

Nigel is absolutely correct - so what should the K factor be increased to?

In my opinion, you have to understand what the factor is seeking to achieve before you can answer this. I believe that its purpose is to control how quickly a players rating converges with his current playing strength. Put it another way - how many games should a player have to player at a given playing level before his rating is equal to that level? Answer that question, and you can work out what K should be.

So, if a player rated 2400 starts playing at 2500 level how many games should he have to play before he is rated 2500?

Unfortunately, I don't think this is the only question that needs to be considered.

Yes, it's true that the higher the K-factor, the more quickly a players rating converges with his current playing strength. However, chess players are not automata, and frequently play at levels a good few hundred points above/below their 'actual' playing strength, certainly for individual games, and even at times for whole tournaments. Using myself as an example, I have probably been about 2350-2450 strength over the past couple of years, but have posted 9+ round tournament performances of between 2280 and 2550 during this time.

Higher K-factors for players whose ratings are roughly constant, but whose performances suffer from these natural fluctuations would only serve to increase the variance of their ratings and actually reduce the 'correctness' of the system. Getting the right K-factor for the new and improving players while not making it too high for the bulk of existing and stable (or only slowly changing) ratings is not an easy matter.

I don't believe the following has been suggested before, but it seems to me that what is actually required is some system whereby a player's K-factor in any period is dependent on how active they have been in a number of preceding periods (say 2 or 4), and how much their rating changed during these periods. A player with a small number of games would then play off some default K-factor; a player with a large number of games but a very small rating change (delta) would play off a correspondingly LOWER K-factor, and a player with a high delta (which therefore implies a high difference between current rating and performance) would be assigned a higher K-factor for subsequent periods. This would seem to satisfy the goal of getting active new and improving players to the correct rating as quickly as possible, without adversely increasing the variances of existing players.

Now I realise already that this system is also imperfect, and open to exploitation in similar ways to the current system (e.g. by sandbagging ones rating down over several periods to not only start from a low point, but also gain a high K-factor, which would then allow the rating to be raised very quickly in a single period), but it would at least seem to address some of the limitations of the existing system in regard to improving and stagnant players.

Thoughts?

[James Coleman]

Max, that suggestion or something along the lines of it, strikes me as being eminently sensible. I know many players that play one international tournament a year so for those players it makes no difference how often the rating lists get published etc

[Sean Hewitt]

So, if a player rated 2400 starts playing at 2500 level how many games should he have to play before he is rated 2500?

Unfortunately, I don't think this is the only question that needs to be considered.?

To be fair, it seems that we agree that this is the question that needs to be answered, but what you are also saying is that the answer may well be different for different groups of players. This seems to be a reasonable point of view, and indeed FIDE concur given that they split players into two groups and assign different K factors (those who have been over 2400 and those who have not).

I do think though that any system of diffentiating K factors should be as simple as possible.

To me, an age related K factor supplement could well have some merit.

[Carl Hibbard]

As to personal abuse on this thread, it strikes me that far, far worse than anything than has been said about GW's mother is the observations about Ingrid Lauterbach's attractiveness or otherwise. Why men think they are entitled to post this sort of thing on forums is beyond me, but I suggest the censor stop worrying about GW's reluctance to say exactly who pays for his mother, remove the posts in question and ban the relevant posters.

Well said, John.

At this point I don't want to ban the offending parties - it's a bit late on this one as it's carried on past the point of no return now...

Clearly it was well out of order and won't be accepted a second time

[Justin Hadi]

As far as I know, the only way to identify delegates is to dig out the latest ECF yearbook and trawl through each organisation listed. Thats ok for leagues, counties and unions. I have no idea how you would find out a congresses delegate.

If anyone has a list of ECF delegates I would be happy to e-mail mine asking for a change of leadership and strongly suggest anyone else who is unhappy with Gerry Walsh does the same.