World Senior Individuals, Acqui Terme, November 2015

Information and discussion on all matters relating to Seniors Chess.
Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Sun Sep 13, 2015 5:06 pm

I warm up each day by believing in three impossible things before breakfast.
Actually doing the impossible just takes a little bit longer.

Rp is wrong. Use Ra (Rating Average).
You cannot do so for round robins.
If two players are involved in the tie and have met, that game must be discounted.
If more than two players are involved in the tie and some of the games between them have taken place, Ra cannot be used.

There, the impossible didn't take that long.

User avatar
IM Jack Rudd
Posts: 4818
Joined: Tue Apr 17, 2007 1:13 am
Location: Bideford

Re: World Senior Individuals, Acqui Terme, November 2015

Post by IM Jack Rudd » Sun Sep 13, 2015 5:12 pm

The trouble with Ra is that it can be rendered nonsensical by the presence of very weak players in a player's field. (If I beat a 1200 and my closest rival beats a 1000, is that really relevant to our respective performances?)

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Sun Sep 13, 2015 6:27 pm

True. I did point out earlier that a simple variation is to delete the lowest rated opponent. It can be just as misleading if you happen to meet one very highly rated opponent.
Of course averaging of opponent's ratings must, by definition, be incorrect. You are averaging omething that does not have a linear relationship.

User avatar
IM Jack Rudd
Posts: 4818
Joined: Tue Apr 17, 2007 1:13 am
Location: Bideford

Re: World Senior Individuals, Acqui Terme, November 2015

Post by IM Jack Rudd » Sun Sep 13, 2015 6:32 pm

Stewart Reuben wrote:Of course averaging of opponent's ratings must, by definition, be incorrect. You are averaging omething that does not have a linear relationship.
A variation on Ra might be "expected score for a 2600 against your field", or something like that.

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Sun Sep 13, 2015 6:46 pm

Jack >A variation on Ra might be "expected score for a 2600 against your field", or something like that.<
That would be similar to PR. It still wouldn't have a linear relationship.
That would also have the problem of an opponent under 2200.

NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by NickFaulks » Sun Sep 13, 2015 7:16 pm

Stewart Reuben wrote: There, the impossible didn't take that long.
It looks as though Rules Commission should grab this one back, since you've cracked it. I shall stay with my view - which does look vindicated by the fact that you and Jack are already discussing necessary tweaks.
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Sun Sep 13, 2015 7:44 pm

Nick, it is nothing to do with the Rules. Nor, come to think of it QC. It is Technical that deals with tiebreaks.
I didn't write that the system is absolutely correct. I wrote that is is easy enough to write the rules. Ra is better than Bucholz or SB in that the players know where they stand, before the start of the last round. Bucholz and SB are both nonsense once one player does not play one round. Again that is easy to correct, but the people writing the rules are obstinate. Vitual Opponent!

NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by NickFaulks » Sun Sep 13, 2015 11:09 pm

Stewart Reuben wrote:Nick, it is nothing to do with the Rules. Nor, come to think of it QC. It is Technical that deals with tiebreaks.
I didn't write that the system is absolutely correct. I wrote that is is easy enough to write the rules. Ra is better than Bucholz or SB in that the players know where they stand, before the start of the last round. Bucholz and SB are both nonsense once one player does not play one round. Again that is easy to correct, but the people writing the rules are obstinate. Vitual Opponent!
It was noted in Abu Dhabi that universal performance ratings had been kicked on to QC. We shall not do it.

The only advantage of Virtual Opponent is that ( apart from yourself? ) everyone seems to dislike it equally.
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Mon Sep 14, 2015 1:21 am

You mean my opinion, that Virtual Opponent is ludicrous, is mild by comparison with the views of everybody else?

What is 'Universal Performance Rating' supposed to mean? Perhaps the new candidate for President of FIFA will be able to explain.

Kevin Thurlow
Posts: 5821
Joined: Wed Apr 30, 2008 12:28 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Kevin Thurlow » Mon Sep 14, 2015 9:17 am

"What is 'Universal Performance Rating' supposed to mean? Perhaps the new candidate for President of FIFA will be able to explain."

I have a horrible feeling that it might be an extra-terrestrial rating system suggested by the President's friends.

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Roger de Coverly » Mon Sep 14, 2015 9:38 am

Stewart Reuben wrote: What is 'Universal Performance Rating' supposed to mean?
There's the algorithm of 400, although perhaps that should be algorithm of 375. You add up all the ratings, add 400/375* excess of wins over losses and divide by the game count.

https://en.wikipedia.org/wiki/Elo_ratin ... nce_rating

If it's just performance and tie breaks you want to use this for, there isn't any real need to adjust for players rated 400 above or 400 below. It's an implicit tie break in favour of those who faced stronger opposition.

NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by NickFaulks » Mon Sep 14, 2015 10:07 am

Kevin Thurlow wrote:"What is 'Universal Performance Rating' supposed to mean? Perhaps the new candidate for President of FIFA will be able to explain."
With capitals, as you have written, it doesn't exist. I used the general term to describe the mythical performance rating definition which is appropriate for all purposes.

I'm sure we all know that the FIDE President takes no interest whatsoever in such mundane matters. Anticipating the chorus of complaints that this makes him unfit to hold his office, can we be sure that the ECF President devotes much time to them either?
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Roger de Coverly » Mon Sep 14, 2015 10:21 am

NickFaulks wrote: I'm sure we all know that the FIDE President takes no interest whatsoever in such mundane matters. Anticipating the chorus of complaints that this makes him unfit to hold his office, can we be sure that the ECF President devotes much time to them either?
There is a difference. Kirsan is an executive president who leads the organisation. Dominic is more of a constitutional or ceremonial figure, with the actual running of the organisation and the infighting done by Phil Ehr and the Directors.

NickFaulks
Posts: 8453
Joined: Sat Jan 02, 2010 1:28 pm

Re: World Senior Individuals, Acqui Terme, November 2015

Post by NickFaulks » Mon Sep 14, 2015 10:38 am

Roger de Coverly wrote: Kirsan is an executive president
No, he isn't. I've just checked the Statutes, for what little they're worth, in case they say otherwise, but they don't.

I admit that I may not properly understand the role of the ECF President.
If you want a picture of the future, imagine a QR code stamped on a human face — forever.

Stewart Reuben
Posts: 4542
Joined: Tue Apr 03, 2007 11:04 pm
Location: writer

Re: World Senior Individuals, Acqui Terme, November 2015

Post by Stewart Reuben » Mon Sep 14, 2015 12:04 pm

It's really excellent the way this thread has gone so complete off topic.

Roger >You add up all the ratings, add 400 excess of wins over losses and divide by the game count.<
If used as a tiebreak system, I presume its purpose is to separate out players who make the same score. Thus
add up all the ratings and divide by the game count. This is sufficient. Most people might call this Rating Average Ra.
It doesn't tell you what to do if you meet an unrated opponent, or have a bye in a game - wherever it be a win, draw or loss. As Jack pointed out, a very low rated opponent renders the average meaningless. That give two reasons for discounting the lowest rated opponent.
Any remaining byes are best dealt with by taking the average of the games actually played. BUT
If a player has won a game by default, is this not a great advantage and should he not simply be put at the bottom of the list, even though it is no fault of his own?

The BCF/ECF has not had an executive President since 1996. Campo was clearly an executive President of FIDE until 1995. Since then the Deputy President, Makro, has been in charge together with the Executive Director (who has been English educated this century). Makro or Nigel Freeman would understand tiebreaks, if either turned his attention to the matter.

The ECF and FIDE are run differently. ECF elects a number of directors to specific posts. FIDE elects people to the Presidential Board. The various 'jobs' are then parcelled out depending on the interests of the members of the PB.