League Rules regarding Board order by ECF rating

[Anthony Taglione]

Boys, you have to take emergency action. If you can bring yourself to do it, you have to have an opening rep something like this:


White 1 f4 or 1 b3 or some QP stuff such as Torre,London,Colle ( although even these are well-known now and really quite dull)

Black 1 e4 d5 2 exd5 Qxd5 and aim for a pawn structure with...c6,...e6

1 d4/Nf3/c4 Slav-based set up against anything. Sister pawn structure rep, cuts down on time.


If the ageing talents can stay awake using this stuff they might yet beat the theory and computer boom.

Andrew, many thanks for the advice but, frankly, I think I'd rather shoot myself.

I guess I should admit, though, that I did play some of that stuff back in the '70s, along with the Hippopotamus, cabbage, the Englund Gambit and 1.P-QR3.

My favourites were always the Dragon, from either side, and the King's Gambit. It seems that the Dragon is back on the rocks again but that's pretty much where it usually lived.

If I do start playing again, though, it'll be thanks to you and your obit. for Richard O'Brien, which I stumbled across in Google. RIP, Richard. Completely off-topic but does Jane Seymour still play?

[Matt Mackenzie]

The cabbage?? Explain that one again

[Anthony Taglione]

Move in the c-file, the a-file, the b-file, b-file again... It spells out "cabbage". It can be workable unless the opponent realises what one is doing.

Very silly stuff.

[David Williams]

So long as the board order does not create mismatches where there is a gap of more than 40 BCF grading points, you can play in any order you like without altering the expected result. Playing a weak player on a high board reduces his expected score by exactly the same as the sum of the increased expected score of all his team mates.

Strange that this topic generates so much heat.

[Ian Thompson]

So long as the board order does not create mismatches where there is a gap of more than 40 BCF grading points, you can play in any order you like without altering the expected result. Playing a weak player on a high board reduces his expected score by exactly the same as the sum of the increased expected score of all his team mates.

That's true, but you need to allow for there only being three possible results in each game. For example, in a 5 board match, the players are graded:

Team A Team B
200 190
180 170
160 150
140 130
120 110

The expected match score is 3-2 to Team A, but the advantage on each board is only 10 points, so the most likely result of each game is a draw, making the match result 2.5-2.5.

If Team A changes its board order, so the pairings are:

Team A Team B
180 190
200 170
140 150
160 130
120 110

The expected match score is still 3-2 to Team A, but now Team A has a 30 points advantage on boards 2 and 4, which is sufficient for the most likely result of those games to be a win, and a 10 points disadvantage on boards 1 and 3, meaning the most likely result of these games is still a draw, making the match result 3.5-1.5.

[E Michael White]

So long as the board order does not create mismatches where there is a gap of more than 40 BCF grading points, you can play in any order you like without altering the expected result. Playing a weak player on a high board reduces his expected score by exactly the same as the sum of the increased expected score of all his team mates.

This would only be true if the grading system had a firm basis in scientific probablity theory which it hasnt. Unfortunately the theory of accelerated pairings is also based on the same shaky foundations.

[David Williams]

That's true, but you need to allow for there only being three possible results in each game. For example, in a 5 board match, the players are graded:

Team A Team B
200 190
180 170
160 150
140 130
120 110

The expected match score is 3-2 to Team A, but the advantage on each board is only 10 points, so the most likely result of each game is a draw, making the match result 2.5-2.5.

Even if the lower graded player never won a game he'd only draw 80% of the time with a 10 point gap. So the chances of all five games being drawn would be a little over 30%. As the lower graded player will actually win a few the chances of five draws are a lot less than that. And the expected score is still 3-2.

[Tryfon Gavriel]

So long as the board order does not create mismatches where there is a gap of more than 40 BCF grading points, you can play in any order you like without altering the expected result. Playing a weak player on a high board reduces his expected score by exactly the same as the sum of the increased expected score of all his team mates.

Strange that this topic generates so much heat.

Sorry I'm not quite sure this is the full impact. Imagine two teams varying from 150-180 strength:

Team A

Team B

But B also had someone rated say 120. Say the 120 was put Board 1 for team B, this would mean everyone else is pushed down by a board, thus increasing the probability of drawing or winning on all boards except the "sacrificial" board 1.

The more boards in the team match, the more impact such a misplacing by ratings would had.

When I pair leagues in online chess for Chessworld.net I make absolutely sure all teams are ordered by Server Rating before pairing. However, of course online chess is fraught with other biases like the potential for engine cheating, etc. But in theory anyway, by placing both opposing teams by rating, you reduce the bias of a sacrificial lamb placed above people who are higher rated.

[Anthony Taglione]

It's clear that a balance needs to be drawn between the two situations of someone's grading being no reflection of their playing strength, so not particularly relevant to their board-placement, and nefarious manipulation of board-order by an unscrupulous captain out to wangle an advantage by sacrificing a player, who might even get lucky and grab a bonus draw, anyway.

I doubt that there is any algorithmic answer to this problem which will be satisfactory to all concerned. As such, we have officials and arbiters to hear complaints and justifications. Team captains need to have discretion within their own team but opposing captains also need to have a mechanism available to punish unreasonable practices.

[Neill Cooper]

We had a recent away match where the home team had to put a sub in on board 1. However there was never a sense that the board one was expected, nor panic at his not appearing. (in their previous 2 home matches they had defaulted board 1 once and used the same sub on the other occasion)

Yes, we could have logged a complaint as I think it affected the result of the match (grades and result below) but I did not want to waste my time. If they are so keen to win a division 3 match then they just lose my respect.

Board Grade Grade Score
1.....121.....179.....0-1
2.....151.....141.....1-0
3.....143.....131.....0.5
4.....140.....129.....1-0
5.....130.....125.....0.5
6.....125.....102.....0.5
7.....120.......93.....0.5

[Neill Cooper]

In contrast we had a home match a few weeks later where 6 of the 7 boards were equally graded (within 4 points!). I had already warned the tournament organiser that board 6 (a junior) was under-graded by 20 points!

Surely for many chess players this is one important aspect of most inter-club chess, that you send the evening play someone of a similar ability to you and get a good game.

Board Grade Grade Score
1.......147.......150.......0-1
2.......137.......137.......1-0
3.......135e......134.......0.5
4.......131.......127.......1-0
5.......129.......126.......0-1
6.......102.......125.......1-0
7........93........94.......0-1

Match drawn 3.5 all, but white beat black 6.5 to 0.5!

[Trevor Davies]

Board Grade Grade Score
1.....121.....179.....0-1
2.....151.....141.....1-0
3.....143.....131.....0.5
4.....140.....129.....1-0
5.....130.....125.....0.5
6.....125.....102.....0.5
7.....120.......93.....0.5

Match result of 4-3 is very close to the "expected" result of 3.9-3.1 (and close to the expected result if the teams had played in correct order of 3.8-3.2). So the Board 1 "sacrifice" didn't achieve much, other than incense the opposition.

A similar situation in our league resulted in the "miscreants" (the winning team, who were judged to be negligent in overstimating an ungraded player's strength) being penalized the (2) match points for their win, but their opponents (who were routed) got nothing. So the match point score was 0 , 0.

In addition to this penalty, to compensate the injured team, there was a range of views from:
(a) declaring all games below the disputed board lost. Hence, a match score of -2 , +2.
(b) shifting 0.5 game points towards them from the other team (which in a close match would have had a knock-on effect on match points).

In the event, the league rules did not allow for such compensation.

How do other leagues deal with such incidents?

[Neill Cooper]

How do other leagues deal with such incidents?

If I had successfully objected I think the main possibility (Surrey League) would be to declare board 1 ineligible, so lose any points he gained (zero), and also remove 1 point on top. So it would have been a 3 - 3 draw.

[David Pardoe]

Some very interesting discussions about gradings and about board order..
Yes, I agree that captains should have reasonable scope to juggle board order...my maxim would be 15 ECF grading points as a general rule..possibly allowing up to 25 grading points in exceptional circumstances. But to say that clubs should seek advice/opinion from league secretaries when operating outside reasonable/normal boundaries.
Where rougue captains are percieved to be blatantly flouting rules/spirit of the competition, it should result in the straight award of the match to the opponents.
Player grades....Yes, I`ve long held that a 10% rule should be a general rule of thumb (ie, treat grades as +/- 10% of actual grade). Grades should not be regarded as `cast in stone`.....
Returning players.... they should forget all about these grading issues, at least for the first season.
Instead, it is imperative that they report immediately to the local club of there choice, and present themselves for selection...the lucky club able to receive such players would be highly delighted to see them.
They should then commit themselves to enjoying there chess and in rediscovering long lost skills, maybe some long lost friends...and hopefully hours of good chessing. Maybe such players should be provisionally rated after the first 5 games, after an initial guestimate, based on any relevant information available. My suggestion is that players who have been out of the game totally for 10 years might initially be rated no more than 15 - 25 grading points lower (but factor in also, the recent grading rebase).
Players who had played informally, or used the web for practice, etc..might even merit a slight initial rerating!

Just my opinions.....

[David Williams]

So long as the board order does not create mismatches where there is a gap of more than 40 BCF grading points, you can play in any order you like without altering the expected result. Playing a weak player on a high board reduces his expected score by exactly the same as the sum of the increased expected score of all his team mates.

Strange that this topic generates so much heat.

Tryfon Gavriel disagrees with me, but his example involves putting a 120 against a 180 instead of a 150. Both Neill Cooper's examples support what I say, in that the only unfairness arises when a 121 substitute plays a 179.

For me the only real issue is substitutes (and defaults) on high boards. These do make a difference, and it's not easy to ascertain whether there is a genuine reason or not. In passing I feel I should mention a local third-team player, grade 77, who was called on twice in one season when his first team top board failed to show up. Already short of time, he found himself playing Nigel Davies and Stewart Haslinger. By all accounts he acquitted himself well.