Tradewise Gibraltar Masters 2017

[Roger de Coverly]

Geurt Gijssen argued that there were very few complaints about the FIDE Rules as they stood, so there was no need to make any changes. His view prevailed.

How about the problem of giving the bye to the lowest ranked player? It's become apparent that can lead to the same player getting the bye in repeated tournaments.

[Roger de Coverly]

In my opinion, this is bad advice. It is much better to pair them together in the early rounds, specifically to avoid pairing them together in later rounds, where the chance of colluding is much higher.

Leagues with the exception of a 4NCL division or two are always all play alls. When a club has two or more teams in the same division, the local rules are usually written so that they play one another as the first match, or failing that, early in the season.

[Roger de Coverly]

My opinion, for what it's worth, is that the arbiters acted prudently and sensibly both in avoiding an Iran v Israel pairing (if indeed they did) and in allowing the other pairings discussed here to stand.

The various people who have parallel run the pairings appear to have detected that the Gibraltar organisers and arbiters had no desire to repeat the Corsica row.

https://en.chessbase.com/post/political ... an-circuit

[Roger de Coverly]

From Facebook:

Stuart Conquest "For complete avoidance of doubt: the pairings were the correct ones, and nothing was done at any time to make Yifan (or any other player) more likely to play women (or men). Her protest - according to the reasons she gave - was therefore entirely without justification."

That doesn't seem to have stopped posters at the chessbase site continuing to flog the deadest of horses. Toxic sites anyone?

[Mick Norris]

As far as I can see there is no such provision in the FIDE rules / Swiss Manager/ the Dutch System of pairing used at Gibraltar and there was no variation to the pairing rules published before the tournament (that I can find).

As has been noted, there was a small variation in Round 1, with the opponents of the top seeds being drawn from within a limited rating band. This was certainly announced beforehand, although possibly only at the event. It has been the practice at Gibraltar for several years.

(Edit: Lawrence has beaten me to it.)

Mick N, I haven't responded to your question because I'm not sure of the answer. I can't add anything to what Mike G has said.

My opinion, for what it's worth, is that the arbiters acted prudently and sensibly both in avoiding an Iran v Israel pairing (if indeed they did) and in allowing the other pairings discussed here to stand.

David

That's fine; my point is that it would be better for the acceptable rule variations to be published to try and save arguments

[Brian Towers]

Why does this manipulation not invalidate the event for norm purposes?

The relevant FIDE Regulation reads:

"Tournaments that make changes to favour one or more players (for example by altering the number of rounds, or the order of rounds, or providing particular opponents, not otherwise participating in the tournament), shall be excluded."

Avoiding a pairing between players from countries actually or prospectively in conflict is not done for any such purpose.

Ok, but is there a list of such purposes that are ok? In other words, can you legitimately avoid Pert v Pert pairings, father v daughter, and where does it stop?

On my FIDE arbiter's course we were taught that the only acceptable purpose was avoidance of known "conflict" and only in cases which did not affect the prize money or major placings. Apart from IRI v ISR pairings other examples of conflict would be of the form "I've run off with your wife and you've publicly threatened to beat me up". If we went into the last round on 1 and 2 then I'd either have to take my chances or forfeit my prize money.

Same family, members of the same club, best pals, etc. are not acceptable reasons. Pre-arranged results are also perfectly possible between players with no such connection between them.

[Alex Holowczak]

This also meant British arbiters didn't engage with the FIDE Swiss pairing rules as much as they could have done, which is probably why many of the arguments about who to float where have been lost - they were several years too late getting to the table.

I certainly don't accept that point.

Stewart Reuben and I raised the issue at the FIDE Congress 1995 in Paris. At that time the BCF Swiss Pairing Rules, as they were then called, were fairly new but seemed to be working satisfactorily, so we suggested that FIDE should consider adopting similar provisions regarding floats.

Geurt Gijssen argued that there were very few complaints about the FIDE Rules as they stood, so there was no need to make any changes. His view prevailed.

It is correct that we were unsuccessful. It is not correct to say that we didn't try.

I didn't say that you didn't try, I said you were too late making the point. This is exactly what you just said, to some extent.

[Mick Norris]

The relevant FIDE Regulation reads:

"Tournaments that make changes to favour one or more players (for example by altering the number of rounds, or the order of rounds, or providing particular opponents, not otherwise participating in the tournament), shall be excluded."

Avoiding a pairing between players from countries actually or prospectively in conflict is not done for any such purpose.

Ok, but is there a list of such purposes that are ok? In other words, can you legitimately avoid Pert v Pert pairings, father v daughter, and where does it stop?

On my FIDE arbiter's course we were taught that the only acceptable purpose was avoidance of known "conflict" and only in cases which did not affect the prize money or major placings. Apart from IRI v ISR pairings other examples of conflict would be of the form "I've run off with your wife and you've publicly threatened to beat me up". If we went into the last round on 1 and 2 then I'd either have to take my chances or forfeit my prize money.

Same family, members of the same club, best pals, etc. are not acceptable reasons. Pre-arranged results are also perfectly possible between players with no such connection between them.

Brian

That's useful, reasonably simple and would be good if it was published formally

[Angus French]

For tournaments like the Gibraltar Masters (essentially 250 players and 10 rounds, though there may be other factors to consider) how easy would it be to modify the FIDE pairing rules (assuming permitted by FIDE and supported by pairing software) such that the following pairing types were avoided:
- players of conflicted countries (as recognised by FIDE or some other world authority);
- same immediate family (partners, parent-child, siblings);
- greater than x% female players (where 'x' is a function of the % of participants who are female and might, for example, be 50 if 20% of players are female);
- same club?

[Roger de Coverly]

how easy would it be to modify the FIDE pairing rules (assuming permitted by FIDE and supported by pairing software)

There's already the "secret" flag that prevents ISR-IRI and similar pairings. I'd not imagine it would be difficult to introduce other such hacks. But should you do so? In the limit, David Howell announces in the British Championship that he doesn't want to be paired against Adams or Jones.

[Alan Walton]

For tournaments like the Gibraltar Masters (essentially 250 players and 10 rounds, though there may be other factors to consider) how easy would it be to modify the FIDE pairing rules (assuming permitted by FIDE and supported by pairing software) such that the following pairing types were avoided:
- players of conflicted countries (as recognised by FIDE or some other world authority);
- same immediate family (partners, parent-child, siblings);
- greater than x% female players (where 'x' is a function of the % of participants who are female and might, for example, be 50 if 20% of players are female);
- same club?

I would agree with the first two points, which seem logical; in a world of equality any female exclusive rule to me seem unnecessary, and keeping track of clubs internationally would be difficult

On a separate issue, countries in UEFA qualifying for the 2016 championship had rules for conflicted countries

For political reasons, Azerbaijan and Armenia (due to the disputed status of Nagorno-Karabakh), as well as Spain and Gibraltar (due to the disputed status of Gibraltar) could not be drawn in the same group. Unlike the previous UEFA qualifying tournament, Russia and Georgia had agreed to play each other if they were drawn together

I wonder what happened to all the Spaniards refusing to play in Gibraltar due to the disputed status

[Alex Holowczak]

assuming permitted by FIDE

There's nothing - for now - saying that you can't use whatever pairing system you like, so long as you declare it in advance on the entry form/terms and conditions. If it were particularly radical, it might be a courtesy to tell QC before you're going to use it, particularly in a title norm event.

[Ian Thompson]

assuming permitted by FIDE

There's nothing - for now - saying that you can't use whatever pairing system you like, so long as you declare it in advance on the entry form/terms and conditions.

The rules do say, though, that "While reporting a tournament to FIDE ... the Arbiter has to submit the rules of this [not officially recognised] system for checking by the Swiss Pairing Committee."

What they don't say is what the Swiss Pairing Committee check and what happens if the pairing system fails the check.

[Brian Towers]

how easy would it be to modify the FIDE pairing rules (assuming permitted by FIDE and supported by pairing software)

There's already the "secret" flag that prevents ISR-IRI and similar pairings. I'd not imagine it would be difficult to introduce other such hacks.

The "secret" flag isn't so secret. I'm not familiar with all the FIDE approved pairing systems but in Vega it hides in plain sight under "File/Round Manager/Avoid Pairs". This gives a page of options allowing individual pairings and pairings between groups of players to be prevented up to a given round.

But should you do so? In the limit, David Howell announces in the British Championship that he doesn't want to be paired against Adams or Jones.

That's a much more serious point. With the exception of conflict pairings I think you need the active consent ("I really want" rather than "I don't care") of all parties concerned. Easy enough for family members, not so easy for larger groups. Apart from the first 2 or 3 rounds I don't think you can have this when both players are in the top x% where x can be argued about.

How could this have helped in Gib? Well, suppose half the women players agreed they'd rather not play each other in the first 3 rounds. Using Vega, these players would have their "club" field set to the same value, say "woman", and the option to avoid same club pairings set for the first 3 rounds. Yes, I know they use Swiss Manager, but surely it must have similar functionality? Of course Yifan could still have been paired against a woman player in the first 3 rounds who opted out but it would be an improvement from her point of view and she would at least understand better what was going on.

[Brian Towers]

What they don't say is what the Swiss Pairing Committee check and what happens if the pairing system fails the check.

I thought the answer to the first part was that the pairing system must not favour any player and to the second that the tournament would not count for norm purposes.