Very long games in weekend tournaments

[Chris Goodall]

Nick - by normal means. The side with the queen can screw up and lose queen for rook, but it takes deliberately bad play to lose the whole queen.

Roger - yes, any claim of a draw would be upheld without much thought, but with increments there's no opportunity to make the claim. I'm hypothesising about bringing back the ability to claim draws even with increments, but only in positions that are sufficiently one-sided not to require the arbiter's judgement.

[Roger de Coverly]

Roger - yes, any claim of a draw would be upheld without much thought, but with increments there's no opportunity to make the claim.

The player seeking the win with a material advantage would have the option to offer a draw unless the rules of the competition forbid it.

[Nigel White]

I once won a QvR ending in 30 moves from reaching that situation. I think the tablebase analysis I looked at afterwards suggested there was a forced win in about that many moves, but I certainly played sub-optimally, benefiting from a mistake towards the end to allow mate a lot more quickly than it could have been.

[Nick Ivell]

Time was when QvR was considered almost trivial. Then the computers found defensive possibilities, albeit not a draw.

If I remember rightly, the Batsford book on rook endings suggested, in a RPvR position, two possibilities as though they were almost equal: a rook deflection resulting in QvR, and another route resulting in KRvK.

I know which I prefer!

[Chris Goodall]

Perhaps I should rewind and consider what scenarios are likely to lead to a game going on for 8 hours...

1. Neither player has winning chances on the board, both are trying to win anyway.
2. Neither player has winning chances, one is trying to win anyway and refusing the draw.
3. Neither player has winning chances, both are trying to draw in the belief that the other is trying to win.
4. One player has winning chances and is trying to convert them, and refusing the draw.
5. One player has winning chances but is content with a draw, the other (probably much higher rated) is refusing and hoping for a miracle.
6. Both players have winning chances.

I think the only scenario that represents deliberately unreasonable behaviour, as opposed to incompetence, is 5. Hence my assumption that it's the player with winning chances who needs a way of imposing the draw on the other.

[Ian Thompson]

any claim of a draw would be upheld without much thought, but with increments there's no opportunity to make the claim.

There may, though, be the possibility of simplifying to a dead position - in your examples QxR or QxN shouldn't take too long to force; RxR probably wouldn't take too long either if the defender wanted to avoid losing.

What you're suggesting sounds rather similar to the American "insufficient losing chances" rule, where their rule of thumb is that the arbiter should award a draw if they think a player rated 1400 -1600 would be very unlikely to lose the position to an opponent rated 2200.

[NickFaulks]

2. Neither player has winning chances, one is trying to win anyway and refusing the draw.

5. One player has winning chances but is content with a draw, the other (probably much higher rated) is refusing and hoping for a miracle.

I think the only scenario that represents deliberately unreasonable behaviour, as opposed to incompetence, is 5.

Also 2, if the hopes of the player trying to win are truly unrealistic.

[Wadih Khoury]

Perhaps I should rewind and consider what scenarios are likely to lead to a game going on for 8 hours...

1. Neither player has winning chances on the board, both are trying to win anyway.
2. Neither player has winning chances, one is trying to win anyway and refusing the draw.
3. Neither player has winning chances, both are trying to draw in the belief that the other is trying to win.
4. One player has winning chances and is trying to convert them, and refusing the draw.
5. One player has winning chances but is content with a draw, the other (probably much higher rated) is refusing and hoping for a miracle.
6. Both players have winning chances.

I think the only scenario that represents deliberately unreasonable behaviour, as opposed to incompetence, is 5. Hence my assumption that it's the player with winning chances who needs a way of imposing the draw on the other.

I think there are also 2 other scenarios, based on mental and physical endurance:

7. One player believes he has better stamina than the other and is waiting for an increasingly likely error as time goes by. This is sometimes done against very young junior or against senior players
8. One player is effectively blackmailing the other "I will not accept a draw and will force you to go home (I will do the necessary pawn moves to avoid the 50 move rules) at a ridiculous time unless you give me the win"

[Chris Goodall]

There must be only a tiny number of positions where one player can't win, can't be punished for trying to win, AND can push pawns at arbitrary moments.

I can only think of:
B7/8/8/7k/8/6KP/7P/8

White could drag this out for another 550 moves if he wanted.

If I encountered this ploy as an arbiter, I would declare the game adjourned until a time to be confirmed later, hence ineligible for grading and treated as a draw for pairings.

[Roger de Coverly]

hence ineligible for grading

I'm not sure that FIDE have not introduced such a rule banning adjourned games fron FIDE rating, but there is no such rule for ECF rating. There may have been such an attempt (and adjudications as well), but I think it got roundly defaeted.

[Roger de Coverly]

In future we would like to reduce the risk of games lasting more than 5 hours. There are various ways we could do this, none of which is ideal. I would be interested to know what other tournament organisers do.

In the typical schedule of rounds Friday evening, Saturday morning and afternoon and then Sunday morning and afternoon, you could potentially adjourn after 5 hours or more in rounds 1,2 and 4. The point being to finish the adjourned games in the evening or at any other time when both playe rs were free. That is the way it used to be before quickplay finishes were the norm. I can recall a Congress in the 1970s where I adjourned the first two rounds. The Controller had the amusing idea of pairing me against someone also with two adjournments. It worked as I won the round 3 game quickly despite losing a pawn for next to nothing in the opening.

[NickFaulks]

I'm not sure that FIDE have not introduced such a rule banning adjourned games fron FIDE rating

Why not? If you read the rules, you will discover that it isn't there.

[Matt Mackenzie]

There must be only a tiny number of positions where one player can't win, can't be punished for trying to win, AND can push pawns at arbitrary moments.

I can only think of:
B7/8/8/7k/8/6KP/7P/8

White could drag this out for another 550 moves if he wanted.

If I encountered this ploy as an arbiter, I would declare the game adjourned until a time to be confirmed later, hence ineligible for grading and treated as a draw for pairings.

Surely an arbiter would decide White "wasn't making an attempt to win by normal means" and declare the game drawn?

[Chris Goodall]

There must be only a tiny number of positions where one player can't win, can't be punished for trying to win, AND can push pawns at arbitrary moments.

I can only think of:
B7/8/8/7k/8/6KP/7P/8

White could drag this out for another 550 moves if he wanted.

If I encountered this ploy as an arbiter, I would declare the game adjourned until a time to be confirmed later, hence ineligible for grading and treated as a draw for pairings.

Surely an arbiter would decide White "wasn't making an attempt to win by normal means" and declare the game drawn?

Not without a claim to that effect from Black, which can only be made during a quickplay finish, which is not what you're in if you're using increments. Disallowing such claims is one of the selling points of incremental time controls.

[Chris Goodall]

In future we would like to reduce the risk of games lasting more than 5 hours. There are various ways we could do this, none of which is ideal. I would be interested to know what other tournament organisers do.

In the typical schedule of rounds Friday evening, Saturday morning and afternoon and then Sunday morning and afternoon, you could potentially adjourn after 5 hours or more in rounds 1,2 and 4. The point being to finish the adjourned games in the evening or at any other time when both players were free. That is the way it used to be before quickplay finishes were the norm. I can recall a Congress in the 1970s where I adjourned the first two rounds. The Controller had the amusing idea of pairing me against someone also with two adjournments. It worked as I won the round 3 game quickly despite losing a pawn for next to nothing in the opening.

I certainly wouldn't want adjournments to become routine again. Players should feel pressure to arrive at a result by the end of the session. They shouldn't be able to sit on their hands and wait for the computer to save them.

I was under the impression that an adjourned game, in a tournament where the advertised time control was increments but no adjournments, would simply be invalid for grading - or if not, that it would be invalidated by the fact that the adjournment was until an unspecified time, like the 31st of June.

2. The following shall be indicated upon the envelope:
a. the names of the players, and
b. the position immediately before the sealed move, and
c. the time used by each player, and
d. the name of the player who has sealed the move, and
e. the number of the sealed move, and
f. the offer of a draw, if the proposal is current, and
g. the date, time and venue of resumption of play.