PDA

View Full Version : Proposal of new draw rule 9.7



Jesper Norgaard
27-06-2009, 07:30 PM
For some time I have pondered over what I think are some gaping holes in the FIDE rules about drawish material. We had a player with 2 Knights in Denmark capturing the last pawn of the opponent (thus he could never lose on time) and still continuing to play eventually winning the game on time in this quickplay finish of a normal game. In a recent game between Women World Championship finalists they ended up with each having a lone Knight, frantically jumping the poor Cavalry around the stable until one of them lost on time. There was a draw claim before flag fall, but it was correctly denied. The win was first incorrectly rejected after flag fall, but after an appeal granted, and this was the correct result given the current rules of Blitz. This kind of play will truely convert us into Clock-punching Monkeys if I may quote Jennifer Shahade :wall:

My suggestion is to add the following rule 9.7 to the current rules to limit such cases, in all 3 game types (normal, rapid, Blitz). I would be very grateful for comments on the usefulness and the pitfalls on implementing such a rule.

New draw rule proposed as 9.7

9.7 The game is drawn, upon a correct claim by the player having the move, if

(a) the opponent has a lone Queen, and the claimant has a Queen or more
(b) the opponent has a lone Rook, and the claimant has a Rook or more
(c) the opponent has two Knights, and the claimant has any material but pawns
(d) the opponent has a lone Bishop or a lone Knight, and the claimant has any material

For all claims the opponent can reject the claim by demonstrating to the claimant and/or the arbiter that he has a forced win from the position. Rule 9.4 applies (losing the right to claim if making a move).

Garrett
27-06-2009, 07:54 PM
there are some queen vs advanced pawn positions where you allow the pawn to queen and you achieve a position where the opponent cannot stop mate.

This would be claimable as a draw under your rule.

There may be similar problems with your other cases.

Kevin Bonham
27-06-2009, 08:07 PM
There was a draw claim before flag fall, but it was correctly denied.

Actually there was no formal draw claim before flag fall, just the player waving her hands around and saying "draw"; whether to offer or claim being unclear. She should have stopped the clock.


My suggestion is to add the following rule 9.7 to the current rules to limit such cases, in all 3 game types (normal, rapid, Blitz). I would be very grateful for comments on the usefulness and the pitfalls on implementing such a rule.

New draw rule proposed as 9.7

9.7 The game is drawn, upon a correct claim by the player having the move, if

(a) the opponent has a lone Queen, and the claimant has a Queen or more
(b) the opponent has a lone Rook, and the claimant has a Rook or more
(c) the opponent has two Knights, and the claimant has any material but pawns
(d) the opponent has a lone Bishop or a lone Knight, and the claimant has any material

For all claims the opponent can reject the claim by demonstrating to the claimant and/or the arbiter that he has a forced win from the position. Rule 9.4 applies (losing the right to claim if making a move).

Such a rule would only be useful for games that lack an increment, since if there is an increment the player should not lose on time anyway.

Such a rule is not necessary (in the 2009 Laws) for normal, rapid or adequately supervised blitz games, as in practically all these positions a draw claim under 10.2 would be upheld. (Perhaps an exception would be a position with K+Q vs K+Q+Ps where the player with the pawns is in danger of being mated even though the mate is not actually forced.)

However a rule of this sort could have uses for inadequately supervised blitz games, in which case I'd suggest listing it under B3. Better still, however, just don't play inadequately supervised blitz games without increments. I am a little wary of the idea that a player who is material ahead should be able to claim a draw in a position where they are not objectively lost but may still be at risk of being mated.

Along similar lines, in my view B3 should be altered to eliminate all wins on time in cases where a mate is only possible with the assistance of an underpromotion.

Jesper Norgaard
27-06-2009, 08:09 PM
there are some queen vs advanced pawn positions where you allow the pawn to queen and you achieve a position where the opponent cannot stop mate.

This would be claimable as a draw under your rule.

There may be similar problems with your other cases.

Yes it is claimable as a draw, but the opponent (non-claimant) can reject by demonstrating the forced win. There may be other similar problems as you suggest, but this is not one of them as I see it. For instance
White: Qb8, Kg4
Black: Qh1 (after queening), Kg2

1.Qb2+ (now Black can claim the draw, but will be rejected by his opponent) 1...Kf1 2.Qc1+,Kg2 3.Qd2+,Kf1 4.Qd1+,Kg2 5.Qe2+,Kg1 6.Kg3,Qf3+ (hoping for Qxf3?? stalemate) 7.Kxf3,Kh1 8.Qg2+ mate

Other examples?

Jesper Norgaard
27-06-2009, 08:53 PM
Actually there was no formal draw claim before flag fall, just the player waving her hands around and saying "draw"; whether to offer or claim being unclear. She should have stopped the clock.

Actually I disagree, because under Blitz rules she was not allowed to make a 10.2 claim or any other claim. If she had been allowed to make a claim under the rules, I am sure she would have, hence the hand waving.



Such a rule would only be useful for games that lack an increment, since if there is an increment the player should not lose on time anyway.

Such a rule is not necessary (in the 2009 Laws) for normal, rapid or adequately supervised blitz games, as in practically all these positions a draw claim under 10.2 would be upheld.

Agreed, although 9.7 could relieve the 10.2 rule in many specific cases - who knows what it means to win "by normal means" anyhow? There are many players and even arbiters confused about 10.2.


(Perhaps an exception would be a position with K+Q vs K+Q+Ps where the player with the pawns is in danger of being mated even though the mate is not actually forced.)

However a rule of this sort could have uses for inadequately supervised blitz games, in which case I'd suggest listing it under B3. Better still, however, just don't play inadequately supervised blitz games without increments. I am a little wary of the idea that a player who is material ahead should be able to claim a draw in a position where they are not objectively lost but may still be at risk of being mated.

In all of the examples of 9.7 a player is in fact in danger of being mated, since helpmate can be demonstrated. However, are there any real examples where Blitz games have been won and 9.7 could have been claimed? Well I know some games are won by an illegal move in such situations, but these are really odd exceptions. Far more often, one player wins on time after both players have been converted into clock-punching monkeys.

It actually occurred to me in a rapid game that we got Rook vs. Rook with very little time, unfortunately he made the only move that allowed me to win his Rook by force with a check followed by a skewer check. He could have claimed 9.7 if this had existed instead of making the losing move, but I don't really mind that. In a game 30 years ago, R vs. R in a rapid game was denied a draw despite my claim, and I lost on time after some juggling around, in an obviously drawn situation.


Along similar lines, in my view B3 should be altered to eliminate all wins on time in cases where a mate is only possible with the assistance of an underpromotion.

This is far more difficult to overcome because a rule would necessarily have to deal with if the pawn is "blocked" from advance or not. But it is not a bad idea, but it needs an implementation, ie. please suggest a rule. I don't "hear" it yet :whistle:

Garvinator
27-06-2009, 09:16 PM
Along similar lines, in my view B3 should be altered to eliminate all wins on time in cases where a mate is only possible with the assistance of an underpromotion.


I wonder why the last QUOTE is deactivated. Any ideas? Perhaps I can still change my post.
In your attempted quoting of KB, you included a backslash in your starting quote. You had [/quote]. Remove the / and it will quote correctly.

If you are doing multiple section quoting of the same person, I find it easier to copy the opening of the quotes (name included) ie the text inside of [ ] and then keep copying that for each separate quote section.

Jesper Norgaard
27-06-2009, 09:25 PM
In your attempted quoting of KB, you included a backslash in your starting quote. You had [//quote]. Remove the / and it will quote correctly.

If you are doing multiple section quoting of the same person, I find it easier to copy the opening of the quotes (name included) ie the text inside of [ ] and then keep copying that for each separate quote section.

He, he, evidence deleted. I did what you suggested in the first place, but made some manual copy-paste error, and then couldn't figure my error. Now corrected. Thx.

Desmond
27-06-2009, 11:04 PM
For some time I have pondered over what I think are some gaping holes in the FIDE rules about drawish material. We had a player with 2 Knights in Denmark capturing the last pawn of the opponent (thus he could never lose on time) and still continuing to play eventually winning the game on time in this quickplay finish of a normal game. In a recent game between Women World Championship finalists they ended up with each having a lone Knight, frantically jumping the poor Cavalry around the stable until one of them lost on time. There was a draw claim before flag fall, but it was correctly denied. The win was first incorrectly rejected after flag fall, but after an appeal granted, and this was the correct result given the current rules of Blitz. This kind of play will truely convert us into Clock-punching Monkeys if I may quote Jennifer Shahade :wall:

My suggestion is to add the following rule 9.7 to the current rules to limit such cases, in all 3 game types (normal, rapid, Blitz). I would be very grateful for comments on the usefulness and the pitfalls on implementing such a rule.

New draw rule proposed as 9.7

9.7 The game is drawn, upon a correct claim by the player having the move, if

(a) the opponent has a lone Queen, and the claimant has a Queen or more
(b) the opponent has a lone Rook, and the claimant has a Rook or more
(c) the opponent has two Knights, and the claimant has any material but pawns
(d) the opponent has a lone Bishop or a lone Knight, and the claimant has any material

For all claims the opponent can reject the claim by demonstrating to the claimant and/or the arbiter that he has a forced win from the position. Rule 9.4 applies (losing the right to claim if making a move).I think you are going to have to define what you mean by "or more".

Jesper Norgaard
27-06-2009, 11:07 PM
I think you are going to have to define what you mean by "or more".

"Or more" simply means any extra material, e.g. Q+P, Q+Q, Q+R+R+3P etc. If you can find a better phrase, please come forward!

Desmond
27-06-2009, 11:14 PM
"Or more" simply means any extra material, e.g. Q+P, Q+Q, Q+R+R+3P etc. If you can find a better phrase, please come forward!Is N+P more or less than a rook? Does it matter where they are placed; what if the pawn is on the 7th rank? What about R v 2P that are on the sixth? Does the arbiter need to analyse the position to determine if the pawns will queen? Your proposal seems incredibly vague.

Jesper Norgaard
27-06-2009, 11:24 PM
Is N+P more or less than a rook? Does it matter where they are placed; what if the pawn is on the 7th rank? What about R v 2P that are on the sixth? Does the arbiter need to analyse the position to determine if the pawns will queen? Your proposal seems incredibly vague.

Perhaps you are overinterpreting, but it could be vague. You can't claim draw when you have a rook and the opponent has Knight + Pawn, as you can only claim the literally mentioned material. I was thinking about being able to claim against a single pawn, but I think this becomes messy, and I think that if you have a Rook or better, eliminating that pawn will only take a couple of secs. Likewise R vs. 2P can't be claimed.

The arbiter would need to get convinced that there is a forced win, or he could just let play continue. Or the settling of if there is a forced win could be rephrased in 9.7 Any of those would do. But I admit that part is vague, but perhaps not presenting too much practical difficulty. The opponent that has the forced win would probably do a lot of the talking :D

Kevin Bonham
28-06-2009, 12:05 AM
Actually I disagree, because under Blitz rules she was not allowed to make a 10.2 claim or any other claim. If she had been allowed to make a claim under the rules, I am sure she would have, hence the hand waving.

I agree that there is no specific code under which she could have claimed a draw. But a player wanting to ask the arbiter to declare a draw under the 2005 Laws although there is no specific code for doing so (which the arbiter might do by invoking the Preface) should follow Art 6.13, which permits a player to stop the clocks to "seek the arbiter's assistance" provided they have a "valid" reason for so doing.

This is necessary because if a player just says "draw" while gesturing to nobody in particular it is not clear if it is a claim of a draw or an offer.

I think that had she claimed, her claim would probably have been rejected as the situation would be ruled to be clearly covered. But it is still my view that she did not actually claim.


Agreed, although 9.7 could relieve the 10.2 rule in many specific cases - who knows what it means to win "by normal means" anyhow? There are many players and even arbiters confused about 10.2.

Oh, there are many world-leading international arbiters confused about it (because they never have to use it, mainly). But the positions you mention are probably the ones that cause the least contention.


In all of the examples of 9.7 a player is in fact in danger of being mated, since helpmate can be demonstrated.

I was talking about positions with more serious risks (but no forced mate) than just theoretical chances. K+Q+a few P vs K+Q where the player with K+Q almost has a forced mate and the other player must find the right moves to avoid the mate, for instance. Uncommon, but can happen.


However, are there any real examples where Blitz games have been won and 9.7 could have been claimed?

Very probably! For instance, here's a skittles game I lost that almost qualifies (http://chesschat.org/showthread.php?p=156992#post156992) - with two queens and two pawns I walked into a mate by a queen and one pawn. The pawn was immaterial to the mate; take it away and your 9.7 applies.


It actually occurred to me in a rapid game that we got Rook vs. Rook with very little time, unfortunately he made the only move that allowed me to win his Rook by force with a check followed by a skewer check. He could have claimed 9.7 if this had existed instead of making the losing move, but I don't really mind that. In a game 30 years ago, R vs. R in a rapid game was denied a draw despite my claim, and I lost on time after some juggling around, in an obviously drawn situation.

In a rapid game today with rook against rook you could also claim a draw by 10.2 and any remotely competent arbiter would award it on the spot. As Reuben writes "The Law was devised for such positions." But curiously, I once had the same experience as you of losing with R vs R in the days before 10.2 existed, in a situation with incompetent arbiters and an unpleasant opponent. (In my case I did not lose on time, but as no-one was doing anything about my impending loss on time in that position I had to move extremely fast to try to get a 50 move draw or triple rep and while doing so, I walked into a rook skewer.)

Competent arbiters have long known you can just tell people playing on for wins on time in such positions to stop being silly. Anand played on for a while with two knights against king in a rapid; IA Gijssen said he would have just issued a disrepute warning to force a draw to be agreed.


This is far more difficult to overcome because a rule would necessarily have to deal with if the pawn is "blocked" from advance or not. But it is not a bad idea, but it needs an implementation, ie. please suggest a rule. I don't "hear" it yet :whistle:

Well, the 2009 Laws has as part of 6.9:

However, the game is drawn if the position is such that the opponent cannot checkmate the player's king by any possible series of legal moves.

One could add to this:

The game is also drawn if the position is such that it is only possible for the opponent to checkmate the player's king by a series of legal moves in which the player promotes at least one pawn to rook, bishop or knight.

Of course that would not eliminate other "silly" losses on time such as losing on time against a single blocked pawn.

Jesper Norgaard
28-06-2009, 12:51 AM
... I think that had she claimed, her claim would probably have been rejected as the situation would be ruled to be clearly covered. But it is still my view that she did not actually claim.

Exactly my point, or at least I agree 100%.


... with two queens and two pawns I walked into a mate by a queen and one pawn. The pawn was immaterial to the mate; take it away and your 9.7 applies.

But I would bet that you weren't even close to considering a draw claim - until you actually walked right into the mate! It happened to me in a position with my N+4P against B+3P but I overlooked a mate in the middle of the board - Ouch. Naturally under these circumstances I could not have applied 9.7, nor could you in the skittle game, but it was close in your case, I admit that. But is it so bad that you can claim a draw if the risk of losing is present - I much prefer that to enforcing clock-punching stupidities.



In a rapid game today with rook against rook you could also claim a draw by 10.2 and any remotely competent arbiter would award it on the spot.

Yes in a lot of the cases you mention you can invoke 10.2 these days, but not in 5 min. games with no increment (still by far the most common Blitz here in Mexico). Even with 5 min. + 2 sec (another common Blitz) you can't invoke 10.2 unless it was adequately overviewed by an arbiter, which does not happen in most Blitz tournaments. So yes with N vs. B you can instead flash out 70+ moves extra from the time the material was reduced, and then try to convince the arbiter 50 moves have passed since then - not a pretty sight in my opinion!



Competent arbiters have long known you can just tell people playing on for wins on time in such positions to stop being silly. Anand played on for a while with two knights against king in a rapid; IA Gijssen said he would have just issued a disrepute warning to force a draw to be agreed.

Yes and if you can't convince players like that, you can always flash Gijssen's trumph card - the disrespect to the holy chess game! I prefer some rules that leave less room for creative arbiters :wall: ... and where in the rules does it say the players have to stop being silly :whistle:



One could add to this:

The game is also drawn if the position is such that it is only possible for the opponent to checkmate the player's king by a series of legal moves in which the player promotes at least one pawn to rook, bishop or knight.

O I misunderstood you on this one, I thought you would let a player claim draw against a player with a pawn, but in fact you are handling that the player has a pawn, and the other has Bishop or Knight, but that one is already handled by 9.7(d) where any material could be a pawn.



Of course that would not eliminate other "silly" losses on time such as losing on time against a single blocked pawn.
You are right, but I'm afraid it is impossible to construct rules that handles these intiutively "obvious draws" without hurting the genuine intent to win against all odds.

Kevin Bonham
28-06-2009, 01:01 AM
Yes in a lot of the cases you mention you can invoke 10.2 these days, but not in 5 min. games with no increment (still by far the most common Blitz here in Mexico).

In a few days it will be possible to invoke 10.2 in 5/0 games if there is "adequate supervision" (which is defined as one arbiter per game.) Only really likely to be useful in very high level events and in playoffs that use that time control.


Even with 5 min. + 2 sec (another common Blitz) you can't invoke 10.2 unless it was adequately overviewed by an arbiter, which does not happen in most Blitz tournaments.

You can never invoke 10.2 when there is any increment that runs to the end, not even if the increment is only one second.

I have never tried playing something like KR vs KR with nothing but two second increments. Would be an interesting exercise.


O I misunderstood you on this one, I thought you would let a player claim draw against a player with a pawn, but in fact you are handling that the player has a pawn, and the other has Bishop or Knight, but that one is already handled by 9.7(d) where any material could be a pawn.

Correct.

Jesper Norgaard
28-06-2009, 01:22 AM
You can never invoke 10.2 when there is any increment that runs to the end, not even if the increment is only one second.

You are right, I painted myself into a corner on that one.


I have never tried playing something like KR vs KR with nothing but two second increments. Would be an interesting exercise.

Yes and even more interesting with 5 min. + 0 sec. and no arbiter around ... I guess you might be allowed under the rules to stop both clocks at that point and call the arbiter to advice him that you aim to attempt to make 50 moves and then claim a draw - so he at least knows before you actually make the 50 moves claim? Then he could start your clock again to let the clock-punching madness begin. As the rules are now, he would not be able to hand you the draw before 50 moves had passed.

Jesper Norgaard
28-06-2009, 01:38 AM
The Clock-Punching Monkeys quote was in fact from Irina Krush, and not Jennifer Shahade (sorry!), and there was no draw claim possible in her game - my point of view is just we should avoid clock-punching madness with rules as much as possible. My 9.7 rule is an attempt in that direction.

ER
28-06-2009, 01:52 AM
(...) my point of view is just we should avoid clock-punching madness with rules as much as possible. (...)
Yep, let's start a "Save the Clock" campaign,:clap: I 've donated one (clock) to the BHCC! Treat it gently plz? :D

Jesper Norgaard
28-06-2009, 03:00 PM
You can never invoke 10.2 when there is any increment that runs to the end, not even if the increment is only one second.
In fact that is a very good reason to have rule 9.7 as a supplement to 10.2 - most FIDE/WC event games are played with an increment, whether normal games or rapid G/30 + 10 sec. pr. move etc. In these cases you would be able to claim the draw directly with 9.7 which you can't with 10.2 (does not apply) when the opponent has only a Bishop, a Knight, 2 Knights, a Rook or a Queen if you have the "matching material" defined in 9.7. Instead I see no rule the arbiter can invoke at present to prevent a player with Knight vs. Knight insist on playing on for at least 50 moves even though an illegal move would just lead to a correction of the move (not lost game as in Blitz) and even though the position would have been declared a draw on a correct 10.2 claim if there had been a quickplay finish, e.g. no increment. I don't mind an explicit "stop being silly" rule in the Laws of Chess, but it should be explicit. :confused:

Yes there is a rule 12.1 The players shall take no action that will bring the game of chess into disrepute which some arbiters tend to want to use in these cases ("everybody knows it is a draw") but I think it is very dangerous to use such a rule, which was never meant to handle draw claims, for too many unrelated things. A bit like Iran suggesting that you are a traitor and a criminal that can be shot down if you go on the street on the wrong time, or use Twitter or Facebook, or don't wear your burka, or use a cell phone camera, or ...

The 12.1 rule I think was made for cases where a player continuously refuse to follow the Laws of Chess, for instance not writing down moves even when requested repeatedly by the arbiter, or disturbing his opponent continuously. If he brings a stripper to the game, and let the lady strip on the chess board even in the players own time, I'm sure it could be invoked. But it should not be used as the trumph ace to any little problem the arbiter encounters.

The Danish game where a player with 2 Knights against lone King was played until flag fall even if the lone King player invoked 10.2 in the quickplay finish phase, just goes to show that even experienced arbiters differs in opinion to what it can be used for - they checked the rules and confirmed that helpmate was possible, so could not claim 5.2(b) and therefore decided to let the game continue. In the end when flag fall was claimed, the arbiter accepted this and declared a win. Later (much later) the case was adjudicated to the main arbiter of this team event, which decided it was a draw. It was then raised to the highest Danish Chess Court and they decided that although they agree the original decision should have been a draw, the first arbiters original decision of a win had to stand because there is no appeal to an arbiters decision on 10.2 :wall:
I am sure the original decision would have been a draw if 9.7 had existed.

Kevin Bonham
28-06-2009, 05:07 PM
In fact that is a very good reason to have rule 9.7 as a supplement to 10.2 - most FIDE/WC event games are played with an increment, whether normal games or rapid G/30 + 10 sec. pr. move etc. In these cases you would be able to claim the draw directly with 9.7 which you can't with 10.2 (does not apply) when the opponent has only a Bishop, a Knight, 2 Knights, a Rook or a Queen if you have the "matching material" defined in 9.7.

Yes, although hardly any opponent would bother playing on in these positions with any increment above maybe two seconds.

By the way I've noticed that if there are armageddon deciders in the upcoming FIDE world cup they are going to have three-second increments from move 60.


Yes there is a rule 12.1 The players shall take no action that will bring the game of chess into disrepute which some arbiters tend to want to use in these cases ("everybody knows it is a draw") but I think it is very dangerous to use such a rule, which was never meant to handle draw claims, for too many unrelated things.

The vagueness of the disrepute rule has been a cause for concern on this board before - I started a thread about it here (http://chesschat.org/showthread.php?t=7920) but posters here are not aware of too many practical examples of it being used.


The 12.1 rule I think was made for cases where a player continuously refuse to follow the Laws of Chess, for instance not writing down moves even when requested repeatedly by the arbiter, or disturbing his opponent continuously. If he brings a stripper to the game, and let the lady strip on the chess board even in the players own time, I'm sure it could be invoked. But it should not be used as the trumph ace to any little problem the arbiter encounters.

Actually a player who continues to refuse to follow the Laws can be punished under 12.8 "Persistent refusal by a player to comply with the Laws of Chess shall be penalised by loss of the game" so the law is probably there to deal with behaviour that is severely inappropriate but not clearly specified elsewhere in the Laws.


The Danish game where a player with 2 Knights against lone King was played until flag fall even if the lone King player invoked 10.2 in the quickplay finish phase, just goes to show that even experienced arbiters differs in opinion to what it can be used for - they checked the rules and confirmed that helpmate was possible, so could not claim 5.2(b) and therefore decided to let the game continue. In the end when flag fall was claimed, the arbiter accepted this and declared a win. Later (much later) the case was adjudicated to the main arbiter of this team event, which decided it was a draw. It was then raised to the highest Danish Chess Court and they decided that although they agree the original decision should have been a draw, the first arbiters original decision of a win had to stand because there is no appeal to an arbiters decision on 10.2 :wall:
I am sure the original decision would have been a draw if 9.7 had existed.

An interesting and unfortunate case. Yes, there is no appeal against 10.2 so unfortunately the player was the victim of a clearly incorrect decision and the highest-level ruling that the decision had to stand is (alas) correct.

Jesper Norgaard
29-06-2009, 03:12 PM
9.7 The game is drawn, upon a correct claim by the player having the move, if

(a) the opponent has a lone Queen, and the claimant has a Queen or more
(b) the opponent has a lone Rook, and the claimant has a Rook or more
(c) the opponent has two Knights, and the claimant has any material but pawns
(d) the opponent has a lone Bishop or a lone Knight, and the claimant has any material

For all claims the opponent can reject the claim by demonstrating to the claimant and/or the arbiter that he has a forced win from the position. Rule 9.4 applies (losing the right to claim if making a move).
Perhaps this could be improved a little bit with this version, just trying to make the grammar a bit more precise (Quote begin)

9.7 The game is drawn, upon a correct claim by the player having the move (the claimant), if

(a) the opponent has a Queen only, and the claimant has at least a Queen
(b) the opponent has a Rook only, and the claimant has at least a Rook
(c) the opponent has two Knights only, and the claimant has any pieces but no pawns (possibly a lone King)
(d) the opponent has a Bishop only or a Knight only, and the claimant has any pieces and/or pawns

For all claims the opponent can reject the claim by demonstrating to the claimant and the arbiter that he has a forced win from the position. The arbiter may let the game continue if he is unsure whether there is a forced win. Rule 9.4 applies (losing the right to claim if making a move).

Quote end. Note that if in 9.7(d) the claimant has a lone King the game is already a dead position as per 5.2(b).

Saragossa
29-06-2009, 06:24 PM
I am of the opinion that if you voluntarily play blitz or other non-increment chess then you have to be prepared to suffer annoying losses due to you flagging, if you don't agree with this then don't play.

Rather then make new drawing rules they should simply abolish the whole armageddon rubbish, and if they are keeping it and you disagree just don't play in the tournament. If enough strong players don't turn up then the organisers will have to revise the system.

Jesper Norgaard
29-06-2009, 09:37 PM
I am of the opinion that if you voluntarily play blitz or other non-increment chess then you have to be prepared to suffer annoying losses due to you flagging, if you don't agree with this then don't play.

If you play Blitz tournaments this is almost a guarantee, you will suffer losses due to flagging even in a drawish situations, however if there is an increment the risk of these annoying losses is much less. But here in Mexico almost 80% of the tournaments (Blitz,Rapid,Normal) are played without an increment, the increment is almost non-existent in Blitz and Rapid, and in Normal games the rate is probably 50%-50%. I am just trying to mend a less-than-perfect situation. And I think the FIDE Rule-making is partly to blame. If we had 10.2 available in all possible modes of play, these problems would perhaps diminish. But I think the FIDE rules are made by arbiters, who are very reluctant to accept the extra work of all kinds of 10.2 disputes. In fact choosing the right rules would be better in democratic elections between all chess players, not just chess arbiters.

I respect your opinion but I don't think many players will refuse to play a tournament just because the ultimate tie will be determined with an Armageddon game. And I would much prefer to have 9.7 in that situation, and it also would be nice to have 9.7 in the die-hard local Blitz tourney (5/0).

Jesper Norgaard
01-07-2009, 07:09 AM
In a local Blitz tourney (5/0) a player got Rook vs. Rook in a clarified position (no immediate threats) but he had less time than his opponent. He stopped the clocks and claimed a draw to the arbiter. He did not mention 10.2. The arbiter ruled that the game should continue. When he did not accept this ruling we had to quicly scramble together an Appeal Committee, and I explained as a member of this that although this situation should be deemed a draw in a quickplay finish from a Normal game or Rapid game using rule 10.2, this could not be applied in a Blitz game.

I did have a bad taste in the mouth because I thought the opponent should have shown some sportsmanship and accepted the implicit draw offer. I would of course have preferred the use of the 9.7 rule to impose the draw as there was no forced win in the position, and a claim for draw was made.

Another anecdote where 9.7 would have been useful, although in the end it seems an arbiter made a wrong decision: In a game of Sub 20 or perhaps Sub 18 championship of Mexico, that would eventually decide a single player to earn participation in a later international Sub x event, there was a tie between Manuel Leon Hoyos (now a GM) and Uriel Capó Vidal (now an IM). Hoyos was the favorite, but it had not cropped out in the final position of the tournament. The rules specified the tie to be broken with 2 G/15 games with no increment. It is not clear if the arbiter thought it was Blitz games, which would explain his later ruling, and the limit between Blitz and Rapid games is in fact 15 minutes, so he might have mixed up between these modes. It is clear that even in 1997 rules G/15 was also a Rapid game. For such an important decision as to who would represent Mexico, it is incredible there was no increment, but alas such were the rules. Long Armageddon games!?

Hoyos got an advantage in both games, but had problems converting them to full points. In the first game he won the exchange but it later reduced to only his Rook against Uriels Knight - which should normally be a draw. Hoyos was maneuvering around, but did not obtain anything concrete. He kept on trying to win until he had 5-10 seconds left, then stopped the clocks telling the arbiter "Okay, draw!" and the arbiter asked why, under which rule, Hoyos told him "he can't win" when the arbiter showed the position similar to White:Kh1,Rg1 Black:Kh3,Nf2++ Hoyos responded that was not going to happen, and the arbiter told him that was not argument, and the game should continue, and pressed the clock. Soon after Uriel was declared winner on time, even though Hoyos setting up a suicide Rook situation should not be to difficult, but I think he was not aware he might lose this on time, and was caught unprepared how to react.

In all of this it is clear that Hoyos did not mention rule 10.2 directly but said it was a draw because the Knight couldn't win, but it is not clear to me if the arbiter rejected the claim because Hoyos did not mention 10.2 explicitly and therefore his draw claim imprecise, or because the arbiter thought it was essentially a Blitz game. Of course Hoyos was devastated from this incorrect decision. The next game showed an advantage to Hoyos again, but he could not concentrate because of his outrage from the first game, and the second game fizzled out to a draw, eventually qualifying Uriel as champion and international participant.

I am sure that if a 9.7 rule had existed, the loss on time in game 1 would never have resulted, but in the end the arbiter was wrong. If the allotted time had been 14 minutes and 59 seconds per player, the arbiter would have been correct though.

Interestingly if it had been Rook vs. Bishop Hoyos could calmly have played on until flag fall, because it is not possible to construct a helpmate with this material for the Bishop. As it were he should perhaps have tried to settle the draw himself before the eventual draw claim with "no time on the clock", by the "suicide Rook" for instance White:Kf1,Re1 Black:Kc8,Nc7 1.Re8+!,Kd7 2.Re7+!,Kd6 3.Rxc7 because in all cases where the Rook is captured, it is an instant draw. It is not forced that the Rook must be captured, but it will probably be difficult to avoid pretty soon for the player with the Knight.

Kevin Bonham
02-07-2009, 02:07 AM
In a local Blitz tourney (5/0) a player got Rook vs. Rook in a clarified position (no immediate threats) but he had less time than his opponent. He stopped the clocks and claimed a draw to the arbiter. He did not mention 10.2. The arbiter ruled that the game should continue. When he did not accept this ruling we had to quicly scramble together an Appeal Committee, and I explained as a member of this that although this situation should be deemed a draw in a quickplay finish from a Normal game or Rapid game using rule 10.2, this could not be applied in a Blitz game.

Your decision is correct because while it was open to the original arbiter to find some way to cause the game to be drawn, it cannot be said that his decision to continue was clearly incorrect. Therefore it would be wrong to overturn it on appeal.


For such an important decision as to who would represent Mexico, it is incredible there was no increment, but alas such were the rules. Long Armageddon games!?

Well FIDE have been happy to allow armageddon games without increment for even higher stake fixtures but fortunately they are now moving towards better ideas such as increment from move 60.


Hoyos got an advantage in both games, but had problems converting them to full points. In the first game he won the exchange but it later reduced to only his Rook against Uriels Knight - which should normally be a draw. Hoyos was maneuvering around, but did not obtain anything concrete. He kept on trying to win until he had 5-10 seconds left, then stopped the clocks telling the arbiter "Okay, draw!" and the arbiter asked why, under which rule, Hoyos told him "he can't win" when the arbiter showed the position similar to White:Kh1,Rg1 Black:Kh3,Nf2++ Hoyos responded that was not going to happen, and the arbiter told him that was not argument, and the game should continue, and pressed the clock. Soon after Uriel was declared winner on time, even though Hoyos setting up a suicide Rook situation should not be to difficult, but I think he was not aware he might lose this on time, and was caught unprepared how to react.

When someone loses in blitz in this fashion I think it is self-inflicted. If a player insists on playing for a win in something like KR vs KN then they should be willing to accept the risk of loss on time. But I agree the game was not blitz anyway.


In all of this it is clear that Hoyos did not mention rule 10.2 directly but said it was a draw because the Knight couldn't win, but it is not clear to me if the arbiter rejected the claim because Hoyos did not mention 10.2 explicitly and therefore his draw claim imprecise, or because the arbiter thought it was essentially a Blitz game.

It is not necessary for the player claiming a 10.2 to explicitly state they are claiming 10.2. If in doubt as to what the claim is the arbiter should ask (and indeed must ask according to 6.13d if the clock is stopped):


If a player stops the clocks in order to seek the arbiter's assistance, the arbiter shall determine if the player had any valid reason for doing so.

In my view if a claim is made and it is unclear what is being claimed then the arbiter needs to find out what is being claimed to determine if there is a valid reason for the claim.

The new 10.2 does not require the player to stop the clock. In my view the requirement to stop the clock should have stayed.


Interestingly if it had been Rook vs. Bishop Hoyos could calmly have played on until flag fall, because it is not possible to construct a helpmate with this material for the Bishop. As it were he should perhaps have tried to settle the draw himself before the eventual draw claim with "no time on the clock", by the "suicide Rook" for instance White:Kf1,Re1 Black:Kc8,Nc7 1.Re8+!,Kd7 2.Re7+!,Kd6 3.Rxc7 because in all cases where the Rook is captured, it is an instant draw. It is not forced that the Rook must be captured, but it will probably be difficult to avoid pretty soon for the player with the Knight.

Correct. If it is blitz then at a certain point the player needs to draw a line and start playing for a draw by giving up the rook.

I wonder if anyone has ever mistakenly got themselves mated by the knight in KN vs KR. Suspect it must have happened to someone sometime.

Jesper Norgaard
03-07-2009, 01:48 AM
I wonder if anyone has ever mistakenly got themselves mated by the knight in KN vs KR. Suspect it must have happened to someone sometime.
I set out to check some statistics of this even though I am aware that my method can't be called 100% error free and will probably show some false positives. But I think it is interesting to check how many games where 9.7 could have been used, in fact turned out as genuine wins over the board. I did this with a big database with 2.3 million games, where a good percentage is grandmaster games. Still it is probably not well dimensioned to what we want to check, especially if we want to see Blitz games (about 2% are Blitz games perhaps?). All the same I think it has some relevance.

For Rook vs. Knight I found 521 examples with 135 wins, as expected the number of wins is quite high since the Knight must stay close to the King or risk being trapped to one side of the board and getting eaten. Of these I found 3 games where the Knight won, and looking closer clearly 2 were false positives, since the Knight was getting eaten in these 2 cases. However, the third one perhaps was a correct win (on time) since the Rook was apparently about to give itself up for the Knight, and it says (Zeit) which is German for time. But still not close to a checkmate with the Knight.

For Bishop vs. Bishop there were 84 examples which were all drawn.

For Bishop vs. Knight there were 162 games all drawn.

For Knight vs. Knight there were 181 examples and 2 wins, which I suspect are false positives or win on time (which is outrageous if nobody has ever won that over the board). In Neuliger-Raffalt 1990 White:Kf4,Ng4 Black:Ke6,Nc6 it says 1-0. I suspect on time, or perhaps the wrong result is reported.

For 2 Knights against a single pawn (for which you can't claim a draw in 9.7) there were 129 examples with 56 wins, quite a high result but expected.

Relevant for 9.7 there were 39 examples (perhaps some are also part of the former pool) where 2 Knights against lone King resulted, and all were drawn.

Just for fun I checked also Knight and Bishop against lone King, and as expected quite a high number of draws even though this is a forced win. Lenier Dominguez at rating 2170 couldn't make it, likewise Gilberto Hernández at 2525 missed it againts Anatoly Vaisser and Vladimir Epishin at 2567. I managed it myself in my 2 games where I got this advantage, but in the second one my opponent just resigned without testing me!!! I was flabbergasted, but Okay it is his shot. I still had to pick up 2 lose pawns of his before being able to continue with the mating.

For 9.7(c) (2 Knights) and 9.7(d) (Bishop or Knight only) it seems therefore there are really no examples whatsoever of winning even if this could be possible on time - an even more convincing case than I would have expected. Still note that we have at least 3 examples where this was won on time (Søren Bech Hansen, Uriel Capó and Women WC Armageddon). Only the last was Blitz, and would today be able to be handled with 10.2 because the game was (heavily) supervised.

For 9.7(a) (Queens) and 9.7(b) (Rooks) I expected to find quite a number of wins since the defense can be screwed up pretty easily. However, statistics as found here show that only few cases are real. First of all I had to browse through all the wins and got to a large part of them were based on a genuine win that emerged naturally from the position, for instance when Queening in the corner there are often a theoretical win if the King is close enough.

For Queen vs. Queen I found 897 examples with 75 wins. Of these 72 were normal wins that raise no eyebrows, and if a 9.7 draw would have been claimed then the opponent would be able to demonstrate a forced win. There were 3 wins where probably one was a win on time, while 2 others follow here:
Keskisarja - Paakonen White:Kg6,f7 Black:Ke5,h2 65.f8Q,h1Q 66.Qf6+ and now it just says 1-0? Well it is true that 66...Ke4? 67.Qc6+ wins, but 66...Kd5 must be a draw.

González Velez 2400 - González Casado Sant Boi 1997
White:Kf4,a7,h5
Black:Kf6,b2
53.a8Q,b1Q 54.Qf8+,Ke6 55.Qe8+,Kf6 56.Qe5+,Kf7 57.Qd5+,Kg7 58.Qd7+,Kh6 59.Qe6+,Kxh5 (alas the last pawn goes, but it might be best chance) 60.Qg4+,Kh6 61.Qg5+,Kh7 62.Qh5+,Kg7 63.Qe5+,Kf7 64.Qc7+,Ke6 65.Qe5+,Kf7 66.Qc7+,Ke6 67.Qc6+,Kf7 68.Qc4+,Ke7 69.Qe2+,Kd6 70.Qd2+,Kc5 71.Qe3+,Kb4?? 72.Qb6+ 1-0.

For Rook vs. Rook I have 1749 games with 50 wins, where the majority were forced wins coming naturally from the position. Only in 9 cases play was initially a drawn position that was eventually won over the board. I didn't count my own win in Rapid by the way :P
A typical example is Marion-Liset U18 2001
White:Kd5,Rh2
Black:Kd7,Re6
63...Ke7?? 64.Rh7+ +- (they went on until checkmate too!)

It seems therefore that the chance of a win for Queen vs. Queen from a drawn position is 3/897 e.g. percentage 0.3344%. The chance for a win for Rook vs. Rook is a little higher, 9/1749 e.g. 0.5145%.

Statistics on this from a Blitz-game-only material would be nice, but I have nothing remotely representative. If anybody has a large DB of Blitz games only, or know where to get it, that would be very useful for me.

In conclusion, the real winning chances in all positions of 9.7 are so slim that it doesn't make sense not to try and make rules to stop the clock-punching nonsense, even in Blitz games without increment. Eventually this means to stop granting wins in positions that would never have won except for the time guillotine. Okay, that is my opinion, and not everyone will agree.

Kevin Bonham
03-07-2009, 03:43 AM
Interesting stats. Sometimes in databases the results are recorded incorrectly.

I assume that quite a few of the rook vs knight cases were forced wins. (In my one experience of this ending, I had a forced win for one move only and missed it costing myself an =1st).

Would be interesting to know this if it could be determined: of the wins in ridiculously drawn positions, how many were played before 10.2 was introduced?

I imagine that if your 9.7 existed, as soon as Q v Q arose the player at risk of losing on time would immediately claim the draw with the idea that if it does happen to be a forced win (and some Q v Q forced wins are not obvious) then the other player might not see the forced win in time. So maybe the opponent should get five minutes to see if they can find a forced win or not.

Jesper Norgaard
03-07-2009, 06:15 AM
Interesting stats. Sometimes in databases the results are recorded incorrectly.
Indeed. But I think that the more important issue is that the DB is not Blitz games only, more like 2% are Blitz. That will lead statistics in the wrong direction. I tried to correct for this in the wins I saw, but naturally did not look at all the draws.


I assume that quite a few of the rook vs knight cases were forced wins. (In my one experience of this ending, I had a forced win for one move only and missed it costing myself an =1st).

Probably but that one I didn't check because I was only looking for the Knight winning ... :whistle: even checkmating :wall:



Would be interesting to know this if it could be determined: of the wins in ridiculously drawn positions, how many were played before 10.2 was introduced?

Of the ridiculously drawn positions I only count Single Bishop/Knight or 2 Knights as that. Then the count is almost zero, only a Knight vs. Knight game springs to mind, where pieces and Kings were far from each other, and result might be plain wrong. From the Queen and Rook examples, high percentage of these were games were from recent Sub x groups particularly below U14, so the weaker the players the more examples I guess. I think a third of games are pre-1993, but I don't know the 10.2 rule birth.



I imagine that if your 9.7 existed, as soon as Q v Q arose the player at risk of losing on time would immediately claim the draw with the idea that if it does happen to be a forced win (and some Q v Q forced wins are not obvious) then the other player might not see the forced win in time. So maybe the opponent should get five minutes to see if they can find a forced win or not.
Yes no point in not claiming. Let the opponent prove he has a win and let the arbiter be inclined to postpone a decision and perhaps award you the draw even after your flag fall. If the arbiter has just the slightest doubt he should let the game continue, so most arguments of "I have a forced win" should be accepted unless they are pure fantasy. If the other player has no time I'm sure he will happy for the implicit "draw offer" and take that. More likely he will have more time than our hero and thus I don't think he should get more time after a syntactically correct claim. If it is not syntactically correct, e.g. he claims it when material is not correct, for instance Rook and pawn vs. Rook and pawn (the second part incorrect) clearly the opponent should be given a time bonus, perhaps two minutes, and our fallen hero with the incorrect claim should perhaps be penalized with 1-2 minutes. Don't know, I prefer to leave that part of the rule making to the arbiters :eh:

It should not be necessary that the arbiter is a strong chessplayer for this rule to work - the burden of proof is on the player that wants to win (on time! :doh: ) so a red flag should be if the claimant is in check, then it is more likely there is a forced win. But Q vs. Q often wins with a series of checks, and then the silent threat that chokes the defense. Luckily an arbiter needs no experience on the "theoretical draw", which is also quite diffuse like the definition of "intelligence" - he only needs to know the rules and use common sense, as in 10.2. So he needs no rule of what is a forced win. He should let the reasonable doubt benefit the player that wants to win. And since that is the case, I feel that there should be no penalty on the claimant either, unless blatantly wrong.

Kevin Bonham
03-07-2009, 05:05 PM
I think a third of games are pre-1993, but I don't know the 10.2 rule birth.

10.2 was introduced in 1997 though I understand it had precursors (of sorts) in some of the previous Laws for various time controls before then.

Bill Gletsos
03-07-2009, 08:21 PM
I believe this discussion needs to have a look at the historical context of these issues and how the Laws regarding them have evolved over time.

Now Jesper's proposed new rule reminded me of a previous FIDE Rapid rule from the late 80's early 90's.

Back in the late 80's FIDE introduced specific rules for "Rapid (60 minute) and Active (30 minute) Chess".

Of significance to the current discussion is what was part (b) of Article 12 of those rules and what was considered positions without "practical winning chances".

The 1988 version of Article 12 of the Rapid (60 minute) and Active (30 minute) Chess rules was as follows:

A game may be drawn, but only before the claimant's flag falls, and supported where necessary by a competed scoresheet,
(a) if the player demonstrates a perpetual check or a forced repetition of position (if this claim is found to be false, the appellant is penalised by having five minutes added to that used on his clock, with the provisio that he be left with at least two minutes);
(b) if his opponent has no practical winning chances (if this claim is found to be false, the appellant is penalised by having five minutes added to that used on his clock, with the provisio that he be left with at least two minutes);
(c) if both kings are in check and a player announces that he is correcting the irregularity, then the last move shall be established and the correct position set up. If this is impossible to do with the last played move of one player, then the game shall be proclaimed a draw.

The following shall be considered positions without "practical winning chances" (provided that there is no forced way to win):

(1) if claimant has queen vs. queen (or rook or bishop or knight or pawn),
(2) if claimant has rook vs. rook (or bishop or knight),
(3) if claimant has bishop (or knight) vs. bishop (or knight),
(4) if claimant has bishop vs. pawn, knight vs. pawn, rookvs. pawn - in all cases provided there is no forced win for his opponent,
(5) king vs. a or h pawn and bishop on the "wrong" diagonal, providing that the king is controlling the promoting square,
(6) in all cases claimant could have additional pieces.

(d) claiming a draw according to points 12(b) or 12(c) is permitted only before claimant's flag falls.

Note even back at this time in 1988 the rules of Lightning chess only allowed a draw claims under the following circumstances:

10. A game is drawn:
a) if one of the kings is stalemated
b) by agreement between the players during the game, not before or after the game
c) if the flag of one player falls after the flag of the other player has already fallen and a win has not been claimed
d)if a player demonstrates a perpetual check or a forced repetition of position under the conditions of Article 10.5 of the Laws of Chess
e) if both players have insufficient material for a checkmate (only king vs. king, king and bishop vs. king, king and knight vs. king, king and bishop vs. king and bishop on diagonals on the same colour)
f) if one player has insufficient material for a possible checkmate as described in 10e) and his opponent's flag falls first.


Note back at this time there were no time controls played with increments.

From the above we can see that FIDE were quite prepared to allow without "practical winning chances" to be used as a means of obtaining a draw in Rapid but not in Lightning.

By the introduction of the Laws of Chess that came into effect on 1st July 1997 all this had been removed by the 1996 FIDE Congress in Yerevan where all the laws (normal, rapid and lightning) were consolidated into the one set of laws.

The essential part of the above Article 12(b) became Article 10.2 of the 1997 Laws and continues today. The Rules Commission meeting at the 1996 FIDE Congress after much discussion decided that 10.2 would not apply to Lightning/Blitz games.

I believe the decision to omit 10.2 from applying to Lightning falls under the traditional spirit in which Lightning has long been played.

Jesper Norgaard
05-07-2009, 05:28 PM
Quite and interesting read, thank you Bill for the "rule making history" (you can read that 3 different ways, as the history of rule making, and as that particular rule making it to the history books, or as a nice little anecdote about making rules ...) :D

I can see that the 1988 rules for Rapid and Active are actually quite close in thinking to the rules 9.7 I made. I see that they are getting away with opponent having a single pawn too, or the claimants lone King controlling the promotion field from an a-pawn or h-pawn plus wrong colored Bishop. Also Queen against lesser piece is included.

As a more general comment time management always has been and always will be an important part of chess - even correspondence chess. There are pitfalls both in using too much time and too little time. No clock management can prevent that there are occasional losses on time, nor can any design of time periods or increments prevent them. There is a certain element of "hitting the wall" when you have only 1 minute for the next 20 moves, that will bring the most skillful player down. Those who enjoy playing bullet games of 1 minutes for the whole game is pushing adrenaline to extremes in these games, and get a kick out of "mission impossible". But I would rather like the opposite, to see time management being the least obtrusive to the players in the games, and let Chess itself become the only argumentation. Let's not make Chess a Circus Act and reduce ourselfs to clock-punching monkeys. As such it is clear that the guillotine after a fixed amount of time, is a constraint that can't be overcome in any way once a player has managed his time badly without having outright forfeited. The rule 10.2 is meant to relieve this in that you can then ask the arbiter to have mercy on you because you would never have lost this if you had had a tiny bit of extra time ...

So the increment per move, from move 1, is really the only way to overcome that. And I think I prefer to play having always a time increment for each move. There is much reason to blame the time control on Radjabov's almost-loss on time in a winning position against Smeets, going up to first control with too little time left, and since he did not have any time left in the last move he just hit the clock right after his move even though he was aware he had knocked over other pieces. It was the only way to survive, and then it still needed the pity from Smeets and the arbiters to work. In fact he should count himself lucky to get a draw out of this mess. But even I think it looks so silly that this is a complete road-stopper when in fact he will be given a new 30 min. or 60 min when reaching move 40. The round may not have finished at all because other players have passed move 40 and are given the extra time. Couldn't he borrow a little from the time he is given in the end of the game, for use right now?

We can actually get rid of 10.2, the 1988 Rapid and Active definitions, and my 9.7 if we enforce an increment in any move of any game. It is a revolution I know, but FIDE has gotten the train rolling on this, and should perhaps go further. All of these drawish situations are only played on because there is a hope that the flag will fall, and again if there are no more chess arguments why my opponent should win the game, then I can defend with as little as 1 second of increment per move, until my opponent gives up "being silly" because he knows he will not win. :clap:

I am quite happy fulfilling this by playing Blitz games of 3 minutes + 2 seconds per move, or 4 minutes + 1 second per move, both with my current chess clock OTB and on the internet. End of discussions, if you think you can win, prove it on the board, no need for the appeal committee for what the players themselves can determine on the board. If you just have to juggle around your King in the corner to prove it, I'm fine with that. However even though FIDE has seen the light, this is far from happening in Mexico and Denmark (my natal country). I collected some statistics to show that.

In FIDE events with Normal games there is no surprise that of the ones I could find from 2008, all 23 events had at least 30 seconds increment per move, some from move 1, others from move 61, but none with no increment.

For Rapid games, I found 5 FIDE events with usually 10 seconds increment per move, in Beijing Mind Sports it was 5 seconds, and in only one event Czech Open there was only 30 minutes fixed for the whole game, no increment.

For Blitz events I only found 4 FIDE events, and when Blitz games were used as a tie-break they usually had increment, but there were two events where Armageddon were used as the tie break. In EU Rapid Championship in Warsaw there was the interesting construct with 5 min. for White, 4 min. for Black, and each player getting 1 second added per move. If a draw resulted, the Black player would win. A drawing of lots determined a winner as the one to choose side.

In Mexico 2008 I found 88 events from which 24 had increment, but 64 had no increment in Normal games.

In Rapid games there were only 1 tourney with increment, and 95 that had no increment.

In Blitz games where were 16 tournaments, all with no increment. I *did* play a Blitz tournament with increment in Mexico, after giving simultaneous display in the "Zocalo" main square of Mexico City, there was a Blitz tournament after with 5 min. + 2 seconds per move. I liked it a lot, but perhaps more so since I got third place and enjoyed the ride more than I normally do in Blitz tourneys. But alas it was a lonely swallow.

The argument for not using increments is that some people have clocks that do not contain an increment mode, and many tournaments can't afford to supply everyone with a clock, they need people to bring their own clocks. But to me this can be handled easily by having the default time as X time with Y increment, and then the alternative time as X + 60*Y. For instance when using 90 min. + 30 seconds per move the alternative is 90 + 30 = 120 or 2 hours per player for the whole game. This would guarantee that if I have a digital clock myself, I will always get games with an increment - well in theory. But in practice tournament directors are very reluctant to accept the increments here in Mexico. Perhaps they are afraid that rounds will be delayed when there is an especially long game etc. Or maybe they are just conservative and wants everything to stay as it was.

In Denmark the overview was just as depressing. There were an overwhelming majority of games with no increment in 2008.

For Danish Normal games, I counted 9 tournaments out of 100 that had increment. For Danish Rapid games, all 28 tournaments were without increment. For Danish Blitz tournaments, the 3 I could find had no increment.

I just know that Matthew Sweeney will hate me for wanting to kill all types of guillotine games. :evil:

My defense is, it is only eutanasia. The guillotine games are already dead, they just don't know it yet ... in 100 years nobody will play guillotine games I think. People with a compulsive desire for guillotine should rather go out and play some Russian Roulette. Or perhaps "chicken" with another hot-head going head-on motorcycle to motorcycle to see who is the chicken. I'd rather play Chess.

Bill Gletsos
05-07-2009, 06:03 PM
I just know that Matthew Sweeney will hate me for wanting to kill all types of guillotine games. :evil:I suspect you mean antichrist, not Matthew Sweeney.

Jesper Norgaard
05-07-2009, 11:24 PM
I suspect you mean antichrist, not Matthew Sweeney.
Yes you are right, I meant AC. Perhaps I should have left out that comment alltogether, as it is becoming personal in an unhelpful way, and besides I want to convince all the conservatives in Mexico and Denmark that we need the increment for all moves in all chess games, so there is no need to search for unrelated enemies (Bush-style) :doh: :doh: :doh:

How about "the increment is like the safety belt for safe driving", you don't really need it until you really need it. :clap:

I think that Radjabov might agree with his 1 second on the clock in the middle of an OTB game ... :owned:

That said I have already lost an OTB game on time not being able to handle the 30 seconds only, but the situation was so difficult that perhaps not even Rybka would have found a defense, so I don't really regret that. Instead I have seen countless games (especially Blitz games) where a perfectly playable position hits "the wall" of the unforgiving guillotine. I have won many many OTB games (Normal or Rapid) where it is not clear I would have won if there had been an increment - I had an advantage, but subsequent play would have shown if it would really win. In these situations 10.2 help little because there is too little time or to much play left to demonstrate a defense. And with some pawns left for both sides, and no continuation left from the 10.2 claim to figure out if there was progress, an arbiter has to reject a 10.2 claim. Then there are the countless Blitz games where 10.2 could be considered, because the position really is half-dead, but you can't claim it because FIDE left out that possibility in 1996 after much discussion.

In conclusion I like 3 minutes Blitz games with 2 seconds increment per move better than 5 minutes Blitz games with no increment, even in the rare case where you actually have adequate supervision, with 1 arbiter per game. The players should finish the games, not the arbiters via 10.2. :clap: Unfortunately I can't find anybody except on the Internet that wants to play it with me, much less tournaments where it is used. Also there are virtually no Rapid tournaments here in Mexico where the increment is even allowed (as opposed to enforced in every game of the tournament).

Jesper Norgaard
18-10-2014, 02:59 PM
A lot of water has passed under the bridge since last commenting on this thread, but I went ahead and made the following proposal to Geurt Gijssen in the "An Arbiter's Notebook" October 2014 column on www.chesscafe.com.

In the mean time it has become a member supported portal, meaning you have to pay a fee to view the material. In other words it is copyrighted material. However, I do not think that bringing my own question and then Geurt's answer in a subsequent post is outside "fair use" of said material. If Chesschat has a problem with the copyright issue, please go ahead and delete the post with Geurt's answer.

See http://en.wikipedia.org/wiki/Fair_use
___________________________
Dear Geurt

In Blitz games without increment, you can often see players continue to try to flag the opponent by aimlessly shuffling pieces around. I have often
wondered why this way of winning games, without showing any game plan, can be allowed. The old 10.2 (now G.5) was designed to avoid that in all types of positions, but is only available in Standard/Rapid games without increment, or if all the Blitz games are supervised by a dedicated arbiter. In the final Armageddon game Socko-Foisor here https://www.youtube.com/watch?v=jr86xZcJAaM between the 2 players of the 2008 World Women Chess Championship, N vs. N occurred on the board, and Foisor tried to appeal to the arbiters for a draw with hand gestures. The arbiters after flag fall declared the game a draw because "you cannot obtain a position where defender cannot avoid mate in 1". However, this was a very old rule that was no longer active. Instead an appeal by Monika Socko determined that indeed she should have won, and she was awarded the match against Foisor.

In Standard/Rapid/Blitz games with increment, the increment assures that all the winning attempts any player may want are possible. But it doesn't give a solution to what happens if a player continues winning attempts beyond all reason, in a position that cannot win, trying to move around to avoid recognizing the inevitable draw. In this case all that can be done is for the player to claim a draw based on 3-times repetition (9.2) or 50 moves rule (9.3) when these are possible. If the player is also not inclined to claim, the arbiter must claim a draw based on 5-times repetition (9.6) or 75 moves rule. Even if a player cannot really hope to "flag" the opponent, there is also no solution until the above claims can be made. This might well disturb the schedule of a tournament, especially if the increment is 30 seconds and the player is very stubborn. It is also possible for the organizers via G.2 that Standard/Rapid games without increment are not allowed to apply Appendix G, so a draw cannot be claimed.

These considerations have inspired me to suggest the following new Article

9.8 The game is drawn, upon a correct claim by a player having the move, if the opponent does not have a forced win, and
(1) the opponent has a lone queen, and the claimant has at least a queen
(2) the opponent has a lone rook, and the claimant has at least a rook
(3) the opponent has two knights only, and the claimant has any material but pawns
(4) the opponent has a lone bishop or a lone knight, and the claimant has any material
(5) the opponent has an a or h pawn and bishop on the "wrong" diagonal, providing that the lone king of the claimant is controlling the promotion square

The introduction note to the Article is:

5.2(f) The game may be drawn by the claim of a player if the material of the opponent is so reduced that a loss by the claimant is very unlikely (see Article 9.8).

The idea is that it should be claimable in all modes (Standard/Rapid/Blitz) and whether increment is used or not. As you can see, the claimant may have more pieces, but the opponent has only a single piece, or in (5) a bishop and a rook pawn. It is easy for an arbiter to decide a 9.8 claim. The only possible complication is if there is a forced win. The opponent would probably claim that he has a forced win, if that were the case.

It should be noted that the opponent or the arbiter can object if there is a forced win for the opponent. The claim should be rejected if anything looking like a forced win is present on the board, and let the players play it out until there is no appearance of a forced win, when the claimant may claim a draw via 9.8 again. If the claim is rejected, the following applies:

9.5 If a player claims a draw under Article 9.2 or 9.3 or 9.8, he or the arbiter shall stop the chessclock (see Article 6.12a or 6.12b). He is not allowed to withdraw his claim.
a. If the claim is found to be correct, the game is immediately drawn.
b. If the claim is found to be incorrect, the arbiter shall add two minutes to the opponent’s remaining thinking time. Then the game shall continue. If the claim was based on an intended move, this move must be made in accordance with Articles 3 and 4.

So the incorrect claim means two minutes is added to the opponent's clock (1 minute in Blitz).

There are 3 examples with a forced win that often occur with the mentioned material.

The following is an example of an incorrect draw claim in P vs. N:
White: Kf2, Ne7 Black: Kh2, h3

8/4N3/8/8/8/7p/5K1k/8 b - - 0 1

Black is on move. Black claims a draw under 9.8(4), however this would be incorrect, because white has a forced checkmate:

1...Kh1 2.Nf5! Kh2 (2...h2 3.Ng3# doesn't work) 3.Ne3 Kh1 4.Nf1 h2 (forced) 5. Ng3# (checkmate)

You may ask why a draw claim against a knight is not depending on if the opponent has pawn(s), while against 2 knights there may not be any pawns. The reason is that the only pawns that allow for a single knight to checkmate, is only a rook pawn, and only if the king is entrapped behind it. Thus it is better to just specify that there is no a forced win and not outlaw any pawn. In N+N vs. P there is always a possibility to try to mate based on a blocked pawn, so capturing the remaining pieces of that player will allow a forced win (even if it may be in 68 moves). These possibilities are simply eliminated from 9.8(3) by not allowing pawns, so the arbiter does not have to evaluate if a forced win is present.

Here is another example of an incorrect draw claim in Q vs. Q:
White: Kg4, Qb8 Black: Kg1, Qh1

1Q6/8/8/8/6K1/8/8/6kq w - - 0 1

Black has just played 1...h2-h1Q and hopes to claim a draw on the next move. White plays 2.Qb6+. Black now claims a draw under 9.8(1). The material is correct, but the draw claim is not because white has a forced checkmate:

2...Kg2 3.Qb2+ Kf1 4.Qc1+ Kg2 5.Qd2+ Kf1 6.Qd1+ (6.Kg3? would be a mistake as it allows 6...Qg1+ with a draw) Kg2 7.Qe2+ Kg1 (Kf1 is no longer possible) 8.Kg3 Qf3+ (Black is hoping for 9.Qxf3?? stalemate. Other moves are equally hopeless) 9.Kxf3 and mate in next move with 10.Qg2#

The Saavedra study is an example of an incorrect draw claim in R vs. R:

White: Kc2, Rc8 Black: Ka1, Rd4

2R5/8/8/8/3r4/8/2K5/k7 b - - 0 1

Black on move may claim a draw based on 9.8(2), but it would be incorrect. White is threatening Ra8+ with mate, and 1...Ra4 2.Kb3 assures White a forced win because of the threats Kxa4 and Rc1#

I believe a draw is a fair result in the 5 types of positions in the Article 9.8, if there is no forced win, whether it is a correspondence game or a bullet game and anything in between.

Jesper Norgaard
18-10-2014, 03:01 PM
Geurt Gijssen's answer in the October 2014 column:

Answer It is very difficult to answer this question. For the information of the readers I would like to mention that I received other letters from Mr. Norgaard and I have to admit that they are very interesting and will probably be very useful to discuss when there is again a possibility to revise some Articles of the Laws of Chess.

Regarding the game Socko-Foisor I would like to make a remark, although it probably has nothing to do with the matter discussed by Mr. Norgaard. This game was, as mentioned by you an Armageddon game. This a game in which White has five minutes and Black four minutes. In case of a draw Black is considered to have won the game. FIDE will delete this tiebreak because players are playing under different conditions and this can be considered as unfair.

Let me discuss the issue of your proposal. In principal it looks to me quite sympathetic and logical. Nevertheless, I have a problem and it is demonstrated by the queen ending example. I am wondering what the procedure is if Black claims a draw. I understand that with the best moves (notice the plural) White will win. Does the opponent have to show the winning series of legal and best moves? Will the arbiter start to investigate the position, and will Black be very grateful for the analysis? I do not know. I still like the idea, but there is a difference with the draw claims of Article 9.2 and Article 9.8. In 9.2 the completed moves and intended move will be considered, but in Article 9.8 the possible next moves have to be checked.
Article 9.7 is probably sufficient:
The game is drawn when a position is reached from which a checkmate cannot occur by any possible series of legal moves.
Also, the new Article is, in my opinion, of assistance in avoiding very long games:
If one or both of the following occur(s) then the game is drawn:
a. the same position has appeared, as in 9.2b, for at least five consecutive alternate moves by each player.
b. any consecutive series of 75 moves have been completed by each player without the movement of any pawn and without any capture. If the last move resulted in checkmate, that shall take precedence.

I am ready to continue the discussion, although I have to mention that a change or addition to the Laws of Chess can only come into effect from 1 July 2017.

Jesper Norgaard
18-10-2014, 03:09 PM
I do agree that it needs to be clear who determines there may be a forced win in 9.8 and why. In my opinion an arbiter of 1200 rating should be able to recognize most patterns that are necessary to recognize when a 9.8 draw claim should be rejected. It should not be necessary to go into extensive analysis, only to determine that the defender is in danger of getting mated, or losing a queen or a rook. He must be able to determine with precision if the material on board matches the requirement, in other words if there are too many or too few pieces. He is not required to analyze the position in any depth - just recognize patterns. The arbiter may be helped by a counter-claim from the opponent that he has a forced win, but it should not be necessary.

Here are the examples that an arbiter should be able to recognize. All positions have Black on move, claiming a draw with 9.8.

Diagram 1
8/8/1Q6/8/6K1/8/8/6kq b - - 0 1

The queen in the corner, Qg3+ or Qg3# defended by Kg4 is what makes this recognizable

Diagram 2
8/8/1Q6/8/5K2/8/8/6kq b - - 0 1

Although this is in fact a draw, the arbiter should be able to recognize the same pattern, and reject the draw claim

Diagram 3
8/8/1Q6/4K3/8/8/8/6kq b - - 0 1

Although this resembles the two previous diagrams, Ke5 cannot support Qg3# and for that reason it is a draw - claim should be accepted.

Diagram 4
8/8/8/8/8/3K4/Q7/2qk4 b - - 0 1

A pattern to recognize, there is no defense. Claim should be rejected

Diagram 5
Q7/8/8/4K3/8/5k2/8/7q b - - 0 1

Black is already in a skewer which will lose the queen. Claim should be rejected

Diagram 6
2R5/8/8/8/3r4/8/2K5/k7 b - - 0 1

Black is threatened with mate (Ra8+) and 1...Ra4 2.Kb3 leaves no defense. A variant has mate threatened and rook threatened by the opponent's king - both threats cannot be defended. Claim should be rejected

Diagram 7
4r3/8/8/8/5k1K/8/8/5R2 b - - 0 1

A skewer Re1+ will win the rook. This can also happen with queens instead of rooks. Claim should be rejected

Diagram 8
k7/8/K1N5/3N4/8/7b/8/8 b - - 0 1

An immediate mate threat is often a clue. Claim should be rejected

Diagram 9
8/4N3/8/8/8/7p/5K1k/8 b - - 0 1

Stamma's mate is easy to recognize, king is trapped behind his a- or h-pawn will allow mating possibilities. The arbiter need not verify the exact moves. If white had a knight already on f5, he would have the wrong tempo, and 1..Kh1 instead draws, but Black's draw claim should still be rejected - it looks like a plausible mate.

Diagram 10
8/B1k5/P7/8/8/2K5/8/8 b - - 0 1

This is a common theme in "rook pawn with wrong bishop". Black needs to occupy the promotion square or the surrounding 3 squares in order to prevent this. Claim should be rejected.

In my opinion there are no more examples that are significant for an arbiter to recognize to be able to accept or reject a 9.8 claim.

Of course, I'm interested in any arguments against that. Especially if you have a position where the 9.8 claim should be rejected. If you don't think a 1200 rated arbiter could handle this claim, I would also be interested in your comments.

Kevin Bonham
18-10-2014, 03:33 PM
In the mean time it has become a member supported portal, meaning you have to pay a fee to view the material. In other words it is copyrighted material.

Even without paying a fee it always was copyrighted material under Australian law, which CC operates under. But something being behind a paywall is just the same as it being in a book. You are allowed to quote for the purposes of critical discussion so long as you don't quote more than is necessary for that purpose. Quoting a whole column and then giving a one-line comment would generally not be allowed.

We have a post discussing this site's copyright policies in the non-chess section here:

http://www.chesschat.org/showthread.php?8973-Copyright-and-use-of-the-work-of-others-all-posters-in-this-section-please-read

Kevin Bonham
18-10-2014, 11:03 PM
I see my previous comments to this thread predate the awesome experience of witnessing (now IM) Ari Dale delivering Stamma's Mate on the board in a junior game (http://www.chesschat.org/showthread.php?11442-King-and-knight-mates-king-and-pawn-in-actual-game).

In that game, black blundered with 8...a3?? allowing white to force mate. So if black to move on any of moves 4 to 8 claimed a draw by the proposed 9.8 then presumably the draw would be granted, if the rule is taken literally. Looking at the position on move 4 it seems unlikely that black would actually be checkmated - but he ended up resigning with forced mate on the board.

At least the wording of the concept needs work. The current wording is:

"9.8 The game is drawn, upon a correct claim by a player having the move, if the opponent does not have a forced win, and [..]"

That allows the following sequence in a case like the Aus Junior example:

1. Black to play move 4 claims a draw under 9.8.
2. The arbiter is unconvinced and says play on.
3. White checkmates Black following the sequence given in the game.
4. Black appeals and argues that at the time he made the claim, the game was not a forced win and hence his claim was wrongly rejected.

The fact is there are positions with the material balances mentioned in which mate might occur.

If the arbiter is going to have the right to reject the claims if it looks plausible that the position could be a forced mate, I'd ask why not also allow the arbiter to reject them if the arbiter thinks mate is a likely or realistic outcome, even if it isn't forced?

I just don't see why a player who might still realistically lose on the board should be able to get a draw because they are running out of time.

Perhaps a better solution would be to just have a list of positions for which Appendix G (the former 10.2) applies even in a blitz game, with or without supervision. After all it is plainly stupid and unsporting that a standard KR vs KR should be a loss on time in blitz without increment, but it is not plainly stupid and unsporting that a player should lose a potential Stamma's Mate position on time, given that a player did actually lose such a position in a much longer game with a large increment! Indeed, in my view it would be totally unfair to declare such a game drawn just because there is no forced mate.

Concerning your examples I think many players above 1200 strength would not know of some of the examples you give.

Also an issue with 9.8 (5) - the claimant gets a draw if they have lone king, but not if they have other material in the same situation?

Jesper Norgaard
20-10-2014, 04:43 PM
I see my previous comments to this thread predate the awesome experience of witnessing (now IM) Ari Dale delivering Stamma's Mate on the board in a junior game (http://www.chesschat.org/showthread.php?11442-King-and-knight-mates-king-and-pawn-in-actual-game).

In that game, black blundered with 8...a3?? allowing white to force mate. So if black to move on any of moves 4 to 8 claimed a draw by the proposed 9.8 then presumably the draw would be granted, if the rule is taken literally. Looking at the position on move 4 it seems unlikely that black would actually be checkmated - but he ended up resigning with forced mate on the board.


This is a very relevant example. We need to distinguish between having a forced win, having something similar to a forced win, and simply playing a position where the opponent can blunder. Any position that has ever been settled in a draw with 10.2, the player who was granted a draw could possibly have blundered in the continuation, had the draw not been granted.

In my opinion after move 4 the position is such that White can lose if he advances the pawns incorrectly (blunder). If he doesn't the draw is easy (5...Ka4 6.Kc4 Ka3 7.Kc3 Ka4 8.Kc4 will repeat for the third time the position first seen after 3...Kxa4 4.Kc4, and Black can claim a draw). After 5...a4?! and 7...h4? we are entering a position similar to a forced win. Only 8...a3?? clinches that the position becomes a forced win for White.

Let's call the positions that can be claimed a draw via 9.8 for P. For the moment FIDE is willing to accept that all P positions cannot be claimed a draw except with 9.2 or 9.3 or 10.2 (the latter in fixed time Standard/Rapid games). In all of these situations, we can cut down the aimless shuffling of pieces by a 9.8 claim, whether there is a guillotine end to the game or not. In addition, that will stop guillotine wins that would have been probable draws with an increment.



At least the wording of the concept needs work. The current wording is:

"9.8 The game is drawn, upon a correct claim by a player having the move, if the opponent does not have a forced win, and [..]"

That allows the following sequence in a case like the Aus Junior example:

1. Black to play move 4 claims a draw under 9.8.
2. The arbiter is unconvinced and says play on.
3. White checkmates Black following the sequence given in the game.
4. Black appeals and argues that at the time he made the claim, the game was not a forced win and hence his claim was wrongly rejected.

The fact is there are positions with the material balances mentioned in which mate might occur.


In this case the appeals committee can indeed determine that there was no forced win in the original 9.8 claim, and that the draw should be given. It is understandable that an appeals committee with several members and sufficient time to discuss, can come to a more definitive and correct determination if there was a forced win, than an overworked arbiter that is required for a quick decision.



If the arbiter is going to have the right to reject the claims if it looks plausible that the position could be a forced mate, I'd ask why not also allow the arbiter to reject them if the arbiter thinks mate is a likely or realistic outcome, even if it isn't forced?

I just don't see why a player who might still realistically lose on the board should be able to get a draw because they are running out of time.

That is in fact what I'm suggesting, although I disagree that would be the situation after 4.Kc4, but I agree after 5...a4?! and especially after 7...h4?

I do think the P positions will probably have something like 98% that would be without realistic winning chances, if you count R vs. R and Q vs. Q without special circumstances in this category. In these 98% positions IMHO letting the players shuffle around to a time win is a miscarriage of justice. Only in the 2% we will find positions like Ari Dale's where continued play seem justified. I would rather stop 98 shuffle wins on time with a draw, even if prematurely deeming a draw in 2 where continued play for a win would be logical.

Perhaps 9.8 could be amended with a final arbiter instruction:

If the arbiter is not sure whether there is a forced win or not, he should reject the claim and apply Article 9.5.



Perhaps a better solution would be to just have a list of positions for which Appendix G (the former 10.2) applies even in a blitz game, with or without supervision.

Sorry for splitting your argument, but I think your idea would be better off as a list of positions directly in Appendix B. After all, allowing 10.2 in Blitz was rejected by vote in the Rules Commision, and I completely agree, having to watch many games with a 10.2 claim in Blitz would be chaotic.

However, for me the question is why having fixed 30 seconds left in a Blitz is any different to having fixed 30 seconds left in a Standard or a Rapid game (where App. G can't be claimed) and where Appendix B also does not apply. I don't see any difference. I think it is fundamental mistake claiming that Blitz is somehow a different beast than Standard or Rapid. It is just chess, with a different start time and possibly different increment (if used).



After all it is plainly stupid and unsporting that a standard KR vs KR should be a loss on time in blitz without increment, but it is not plainly stupid and unsporting that a player should lose a potential Stamma's Mate position on time, given that a player did actually lose such a position in a much longer game with a large increment! Indeed, in my view it would be totally unfair to declare such a game drawn just because there is no forced mate.


Actually I have had 3 decisive in Rapid games with R vs. R. In the middle game I also had 3 pawns, and my opponent had 2 pawns, but I lost on time. That perhaps doesn't count, but the first R vs. R I lost on time because 10.2 didn't exist, after lots of rook shuffling. I protested during the game to the arbiter, but in vain, the rules did not allow for any draw claim then. In the last R vs. R game, I won just after eliminating his last pawn, and he walked right into a skewer with lone rook vs. lone rook and 5 seconds on the clock (I had 40). Just the same I have no problem with that any player can claim a draw against a lone rook if he has a rook and perhaps more pieces. So in the last game I would not have objected to him claiming the draw at that moment (before the skewer). I also have no objection to that 9.8 can be claimed even when an increment is in place. And I would gladly allow Ari Dale's opponent to claim a draw in such a position against me (before 8...a3??), for the eradication of shuffle wins in the P positions.

I have had a shuffle Blitz game with 2 knights (black and white) where FIDE's only solution is to encourage each player to play quicker, and let the clock decide who wins. I think it is ludicrous, since no player would ever put his king in the corner and put the knight on the next square so that a smothered mate can happen.



Concerning your examples I think many players above 1200 strength would not know of some of the examples you give.

Yes, as a matter of fact I would suspect that many GM players would not know diagram 4 or 10 by heart. The idea is however that these 10 positions are included in arbiter training material - so that any arbiter will have seen them before. This might not prepare them for any possible position with a forced win, but it covers 95% of the reasons why these positions are typically won with this reduced material.

Personally I think it is uglier that Socko can win a knight shuffle in Armageddon, than that Ari's opponent could have claimed a draw before walking into a forced loss. If you want to make an omelette, you have to accept some eggs will be broken. I don't think Ari had a right to win on time, just because his opponent's king had walked in front of his own rook pawn.

The other argument for allowing the P positions to be claimed a draw, is that in fact in all Standard/Rapid/Blitz games with or without increment, these positions are very often agreed a draw by the players themselves, testifying that they think that a draw is a fair result. Not many are played out, even if they could from being a guillotine game. Some people have a sense of fair play, that will overrule the argument "I can win on time".

In fact I had a Rapid game where my opponent had 50 seconds, I had 2½ minute, and I had a pawn more that would be eliminated within 2 moves, which would mean R vs. R, where I could just shuffle around. He would have had a chance to claim 10.2 though. I just offered the draw right away.



Also an issue with 9.8 (5) - the claimant gets a draw if they have lone king, but not if they have other material in the same situation?

Yes it is possible change the rule to allow the claimant to have more material. One possible solution could be

(5) the opponent has an a or h pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimant may have any material)

This way will allow the claim when the claimant has a lone king, but will still oblige him to control the promotion square with the king, even if he has more pieces that could also control the pawn. A small concession.

I think that it is always possible to reach a conclusion whether there is a forced win or not, with tablebases and other chess software. Thus an appeals committee should be able to come to a correct decision. We cannot expect an arbiter to walk around with this software, and IMHO we should accept an imperfect decision to 9.8 from him/her.

I might make a post with some statistics on how few "Ari Dale positions" there are compared to simpler positions without many realistic chances to win, taken from a 5 million database I have. For instance there are also several "Stamma's mate" that simply emerged as the logical conclusion, without the opponent making any obvious mistakes like Ari's opponent.

Jesper Norgaard
20-10-2014, 06:14 PM
A small example of what can happen, Alvarado - Fernandez 2010

n7/PK1k4/8/2Bn4/1P6/8/8/8 w - - 0 78

I have no problem that White can claim a draw with 9.8 in similar positions, in fact White is completely winning after 78.b5. Actually he cannot claim a draw because of 9.8(3) does not allow the claimant to have pawns. But see what happened:

White played 78.Kxa8??? Kc8 and a knight mate on b6 or c7 is unavoidable. Ouuch!! He constructed his own Stamma's mate!

Kevin Bonham
20-10-2014, 11:25 PM
This is a very relevant example. We need to distinguish between having a forced win, having something similar to a forced win, and simply playing a position where the opponent can blunder. Any position that has ever been settled in a draw with 10.2, the player who was granted a draw could possibly have blundered in the continuation, had the draw not been granted.

I don't agree. The purpose of 10.2 was to save players from losing on time in a position where they couldn't have possibly blundered to the point of getting mated. That is, mate was still mathematically possible but realistically would never have happened no matter how strong the opponent, how acute the time trouble (etc). It's a draw if the player would have to do something like helpmate, or throw pieces away, or make no attempt to stop opposing pawns running through, or walk into blatant one-move skewers and forks for no reason.

If the player could possibly (with realistic chances of it happening) have blundered then the position was winnable by normal means. Taking advantage of opposing blunders is one of the normal ways to win a game of chess.


In addition, that will stop guillotine wins that would have been probable draws with an increment.

I don't want to stop them, except for stupid stuff like KN v KN that would have been virtually certain draws with an increment. If a player has not secured themselves against any realistic chance of loss, and their flag falls, and the opponent is trying to win, then they should lose. They haven't done enough to ensure they cannot lose within the time available for the game.


That is in fact what I'm suggesting, although I disagree that would be the situation after 4.Kc4, but I agree after 5...a4?! and especially after 7...h4?

If a game is won by mate then by definition it was winnable by normal means in all positions that appeared in it. (Barring illegal moves for example.)


I would rather stop 98 shuffle wins on time with a draw, even if prematurely deeming a draw in 2 where continued play for a win would be logical.

I'm happy to police the really blatant shufflers but not at the cost of depriving someone who is trying to win a position where they still have some kind of winning chances of a win just because their opponent has run out of time. I would rather have a million blatant shufflers winning on time than have one game be called a draw in that circumstance.


Sorry for splitting your argument, but I think your idea would be better off as a list of positions directly in Appendix B. After all, allowing 10.2 in Blitz was rejected by vote in the Rules Commision, and I completely agree, having to watch many games with a 10.2 claim in Blitz would be chaotic.

But the arbiters would not have to watch many games with 10.2 claims, since only certain limited positions would qualify. These would tend to be reached near the end of the round when most games had finished and most would be blatant stuff that could be ruled on instantly. If it's too hard then maybe allow a more limited version for blitz. The player can stop their clock and claim a draw based on the position on the board only, if the claim is not accepted then the claimant loses.


However, for me the question is why having fixed 30 seconds left in a Blitz is any different to having fixed 30 seconds left in a Standard or a Rapid game (where App. G can't be claimed) and where Appendix B also does not apply. I don't see any difference.

But you are creating a difference in the opposite direction by allowing players to claim draws in blitz in positions in which a competent arbiter would not award an App G draw in Blitz or Standard.

In any case there are valid arguments for less easy access to claimed draws in blitz. The first is practicality for the arbiters. The second is that by choosing to play blitz without increments a player has said they're not so interested in being a chess purist and more interested in speed as well as good moves. So there is no need to go to trouble applying purist standards to blitz-without-increment since it is not a purist form of the game by definition.


I think it is ludicrous, since no player would ever put his king in the corner and put the knight on the next square so that a smothered mate can happen.

And I agree there. But a position where no player would actually lose is quite different to a position which a player might lose but very probably wouldn't.


Personally I think it is uglier that Socko can win a knight shuffle in Armageddon, than that Ari's opponent could have claimed a draw before walking into a forced loss.

I'd rather not allow either of those situations.


I don't think Ari had a right to win on time, just because his opponent's king had walked in front of his own rook pawn.

The funny thing is they had oceans of time, there was an increment of a minute. So if he has the right to win on the board with an increment of a minute, why does he not have the right to win when the same position is on the board and the opponent has flagged (possibly by managing time badly)?


The other argument for allowing the P positions to be claimed a draw, is that in fact in all Standard/Rapid/Blitz games with or without increment, these positions are very often agreed a draw by the players themselves, testifying that they think that a draw is a fair result. Not many are played out, even if they could from being a guillotine game. Some people have a sense of fair play, that will overrule the argument "I can win on time".

This is a bit like arguing that because giving people presents is nice we should make everyone do it. If you legislate for virtue you actually deprive players of an opportunity to display it.


In fact I had a Rapid game where my opponent had 50 seconds, I had 2½ minute, and I had a pawn more that would be eliminated within 2 moves, which would mean R vs. R, where I could just shuffle around. He would have had a chance to claim 10.2 though. I just offered the draw right away.

I would do the same, at least once R v R was reached. But on the other hand, I once flagged an opponent who had just reached KR vs KRP where the P I had was a rook's pawn and the position was objectively an easy draw (it was G60 flat). I was material up and ahead on the clock, so figured if the opponent didn't remember to claim 10.2 correctly that was his problem. (He offered a draw but I declined.)


Yes it is possible change the rule to allow the claimant to have more material. One possible solution could be

(5) the opponent has an a or h pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimant may have any material)

Actually we need to be a little bit careful here because there are cases involving the claimant having only a knight's pawn on the same side that are wins or potential wins. The K and B stalemate the king forcing the pawn to move to where the rook's pawn can take it.

Jesper Norgaard
21-10-2014, 11:53 PM
Yes it is possible change the rule to allow the claimant to have more material. One possible solution could be

(5) the opponent has an a or h pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimant may have any material)


Actually we need to be a little bit careful here because there are cases involving the claimant having only a knight's pawn on the same side that are wins or potential wins. The K and B stalemate the king forcing the pawn to move to where the rook's pawn can take it.

Interestingly I could not reconstruct that effect.

For instance

k7/1pK5/8/P7/3B4/8/8/8 b - - 0 1

Here Black can play 1...b5 and 2.axb6 will simply will reinforce the stalemate. After some poking around I decided to make a full search in the 5 Million games database. I had searched "rook pawn with wrong bishop" before and found 66 games with this material, however additionally I had put the constraint that the king is already controlling the promotion square, so the positions found could in fact directly be claimed a draw via 9.8. All were drawn!

Instead I added the constraint that the opponent has a knight pawn. In this case I got 219 positions, where surprisingly 26 were wins. However, 5 of those happened to be fake versions of this ending, where White plays g5, Black plays hxg5 and in the moment where White will play hxg5 we in fact have a rook pawn against a knight pawn, but the h-pawn will disappear with the next move hxg5 so it doesn't really count. That leaves us with 21 genuine wins. When I began to focus on those games, it turns out that all of them had to do with quite silly mistakes. Here are a few examples

8/6k1/6p1/6K1/4B2P/8/8/8 b - - 0 1

Any idea for Black's strongest move? Well it certainly wasn't the game continuation 1...Kh7?? 2.h5! Kg7 3.hxg6+-

k7/2K5/3B4/1p6/8/P7/8/8 b - - 0 1

Here Black apparently thought that he was in stalemate, but the bishop is on d6 not on c5, so 1...b4?? 2.axb4 1-0

8/1k6/3K4/1pB5/8/P7/8/8 b - - 0 1

Here Black should just play Ka8, but played 1...Ka6? 2.Kc7!+- The reason Black is losing is because he can no longer return to the promotion square, not so much whether White can convert the a-pawn to a b-pawn or not.

So it seems that the knight pawn disturbs the defenders quite a lot, enough to perhaps rule them out in 9.8(5):

(5) the opponent has an a- or h-pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimaint may have any material except a neighbouring b- or g-pawn)

This way the above 21 positions cannot be claimed a draw in 9.8.

Jesper Norgaard
22-10-2014, 12:32 AM
I have found 13 genuine wins with lone knight or lone bishop. In all 13 a rook pawn was involved! That is quite remarkable, since several positions were quite non-standard. There were 8 genuine Stamma mates (just like the Ari Dale game), but here are some other examples:


8/2P5/8/8/p7/6Q1/P4K1p/7k w - - 0 55

Menyhart-Donka somewhat a fake example. Instead of the simple 55.Qg2# perhaps White wanted to punish Black for not resigning and played 55.Qb3 axb3 56.c8=B bxa2 57.Bh3 a1=Q 58.Bg2# Without the h2-pawn, this wouldn't have been checkmate.

8/3B4/7p/4p1bk/8/6K1/8/8 b - - 0 53

Trueman-Conway 53...e4?? 54.Be8# Again we see the h6-pawn making a fundamental role of trapping Black's king.


8/8/6p1/5n2/4N2p/7k/8/6K1 b - - 0 74

Pak-Kolpakov Perhaps the dream of exhanging knights was sweet, but not so much after 74...Ng3?? 75.Nf2# The h4-pawn is vital for trapping Black's king.

So it seems that the rook pawn having a decisive effect in all 13 wins for the lone bishop or lone knight enough to perhaps rule them out in
9.8(4):

(4) the opponent has a lone bishop or a lone knight, and the claimant has any material but a- or h-pawns

This will also make it impossible for Ari Dale's opponent to claim a draw previously in the game e.g. after capturing the a-pawn.
It will effectively avoid the draw claim in Stamma's mate situations.

Jesper Norgaard
22-10-2014, 12:58 AM
However, for me the question is why having fixed 30 seconds left in a Blitz is any different to having fixed 30 seconds left in a Standard or a Rapid game (where App. G can't be claimed) and where Appendix B also does not apply. I don't see any difference.





But you are creating a difference in the opposite direction by allowing players to claim draws in blitz in positions in which a competent arbiter would not award an App G draw in Blitz or Standard.



I suppose you are talking about App. G.5 examples, not G.4, and I suppose you mean Standard/Rapid, not Standard/Blitz.

The question is if there are any more examples? Would you disagree for KQ vs. KQ or KR vs. KR positions, for instance? In other words, would you not grant a draw in G.5 (the old 10.2) for these positions?

Just to recap, here is the full 9.8 after several changes:


9.8 The game is drawn, upon a correct claim by a player having the move, if the opponent does not have a forced win, and
(1) the opponent has a lone queen, and the claimant has at least a queen
(2) the opponent has a lone rook, and the claimant has at least a rook
(3) the opponent has two knights only, and the claimant has any material but pawns
(4) the opponent has a lone bishop or a lone knight, and the claimant has any material but a- or h-pawns
(5) the opponent has an a- or h-pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimaint may have any material except a neighbouring b- or g-pawn)

If the arbiter is not sure whether there is a forced win for the opponent, he must reject the claim and apply 9.5.

Kevin Bonham
22-10-2014, 11:54 AM
I suppose you are talking about App. G.5 examples, not G.4, and I suppose you mean Standard/Rapid, not Standard/Blitz.

Correct; I meant to write Standard/Rapid.


The question is if there are any more examples? Would you disagree for KQ vs. KQ or KR vs. KR positions, for instance? In other words, would you not grant a draw in G.5 (the old 10.2) for these positions?

A typical KR v KR with no other material is exactly the sort of position the rule was created for. Likewise KQ vs KQ except in certain up-against-the-edge type situations where mate is either forced or else possible if an error is made. My concern would be that there may be positions of the sort stated where the presence of extra material for the claimant means that they have to be careful to avoid being mated. Maybe for instance a KQ (plus many pawns) vs KQ where mate is not forced but the player with the pawns has to exercise some skill to prevent the opposing KQ from delivering mate.


Interestingly I could not reconstruct that effect.

I thought I had seen a puzzle with that as a theme or in a game, but I cannot now reconstruct it either. However I can reconstruct it with two defending pawns:

5KBk/6p1/6p1/8/7P/8/8/8 w - - 0 1

Whoever is to move white forces mate. Although that is fine by your previous version as it is a forced mate, it introduces the possibility that black might have had to be careful to avoid this position.

For instance black may have just played ...Kh8 believing he had to keep his king in the corner instead of ...Kh6 which still draws (and rather obviously because he catches the pawn, but perhaps there are less trivial examples).

Kevin Bonham
22-10-2014, 02:18 PM
This sort of thing is also relevant here, suggesting that one or more rook's pawns would also be an exception in (5)

7k/7p/7K/7P/8/3B4/8/8 w - - 0 1

1.Bg6. Now black draws with ...Kg8, but if he blunders with ...hxg6 he loses.

7k/5K1p/7p/7P/8/3B4/8/8 w - - 0 2

1.Bg6 forces hxg6 and white wins.

Admittedly with these cases I am struggling to find a position that is not either a forced mate or loseable only by an obviously and immediately bad move.

Kevin Bonham
22-10-2014, 02:29 PM
5KB1/6pk/6p1/6p1/7p/7P/8/8 b - - 0 1

Black to move. :lol:

(This and above two posts are all constructed positions, not real games.)

Jesper Norgaard
22-10-2014, 06:42 PM
5KB1/6pk/6p1/6p1/7p/7P/8/8 b - - 0 1

Black to move. :lol:

(This and above two posts are all constructed positions, not real games.)

This massive mobilisation of g-pawns prompted me to one more little change (to pluralis, and using "adjacent" instead of "neighbouring"):

(5) the opponent has an a- or h-pawn and bishop on the "wrong" diagonal, providing that the king of the claimant is controlling the promotion square (claimaint may have any material except adjacent b- or g-pawns)

Under this rule the example of course is not claimable as a draw with 9.8 because g-pawns are not allowed when the opponent's pawn is an h-pawn.
Still it seems to me this study even doesn't quite work under the original rule (claimant allowed any material) since it seems to be a forced win both with Kh8 and Kh6, thus there is no "blunder" element.

7k/8/6K1/6p1/6Bp/7P/8/8 w - - 0 1

White can win with 1.Be6 g4 2.hxg4 h3 3.Bd5 h2 4.Kh6 h1Q+ 5.Bxh1 Kg8 6.g5+-

I don't know if that was the intended effect of the h4-pawn.

Kevin Bonham
22-10-2014, 11:03 PM
Still it seems to me this study even doesn't quite work under the original rule (claimant allowed any material) since it seems to be a forced win both with Kh8 and Kh6, thus there is no "blunder" element.

It is certainly a forced mate with ...Kh8. I'm not sure about ...Kh6 and would be interested to see a forced winning line after 1...Kh6 2.Be6 g4 3.Bxg4 (or it's an easy draw) Kh7. Computer gives +8.66 but then shuffles pieces unable to find a plan!

The complexity of the exceptions needed to be developed to make your rule work is an issue. Arbiters won't remember them all and more can be generated.

Jesper Norgaard
23-10-2014, 09:31 PM
It is certainly a forced mate with ...Kh8. I'm not sure about ...Kh6 and would be interested to see a forced winning line after 1...Kh6 2.Be6 g4 3.Bxg4 (or it's an easy draw) Kh7. Computer gives +8.66 but then shuffles pieces unable to find a plan!

Since this is a 7-8 man ending, there are no tablebases available that can help us find the right moves from the start. However your line cuts off a pawn and in many cases can lead to a 6 man ending. This is one of my wins against slightly faulty engines like Rybka or Stockfish on 3 minutes for the whole game for the engine, so not very reliable but gives quick results. Komodo seems to be faultless in these types of endings, when I started playing against it, no more wins!

Here is Rybka's mistake:

5k2/6pK/8/6pB/7p/7P/8/8 b - - 0 1

Somethings gotta give, but 1...Ke7 seems to be fine, for instance 2.Kxg7 is a tablebase draw. 1...g4? 2.hxg4! This is a tablebase win. Even after 2...g5 3.Lg6 White is in time to stop the h-pawn, and will win the g5-pawn in the long run.

I had several other wins against the different engines, but Komodo convinced me it must be a draw. The problem is that even if White can drive the Black king very far towards the queenside, it still isn't possible to run back and win the pawns g5 and h4 without letting Black return the king to g7 or g8.

Since the claimant is now not allowed to have b-pawns (against a-pawn and "wrong" bishop) or g-pawns (against h-pawn and "wrong" bishop) this position is not claimable for a draw under 9.8. With a single double pawn you have showed there is a forced win, also seen in a game Kasparov won against Short. The forced win is not strictly a problem because they are also not claimable for a draw, but not allowing knightpawns removes an important set of "winning chances" so in my opinion makes sense. Not allowing rook pawns when claiming a draw against lone bishop or lone knight makes sense in the same manner, it removes a set of "blunder chances" from positions which are not yet a forced win.



The complexity of the exceptions needed to be developed to make your rule work is an issue. Arbiters won't remember them all and more can be generated.

I wonder which other exceptions you are thinking about, and which additional examples that can be generated? Aren't new examples all forced wins, and rather easy to recognize as such? Diagram 2 with a study-like draw in Q vs. Q with Kh2-g1-h2 at the right moment, is certainly an example of "almost forced win" so difficult to handle for arbiters.

With the clause that the arbiter should reject the draw claim if he is not sure if there is a forced win, I think a workable environment is in place for the arbiter and the two players. If the opponent has very convincing arguments why he thinks he is winning, the result will probably be that the draw claim is rejected, and the continuation will show if he was right, but as an advantage to 10.2, the arbiter is not obliged to stand around and watch, or to make a difficult decision after a flag fall (draw or loss). If there is a flag fall after a rejected 9.8 claim, then that is a loss and the end of the story (except for a possible appeal).

The USCF Blitz rules have some other sets of material that can be claimed a draw:

"TD TIP: Blitz tournaments allowing "Insufficient Losing Chances" (ILC) claims should be advertised and announced in advance. The TD should be aware that common practice has shown that in addition to the official Blitz rules that “Insufficient Losing Chances” claims are upheld only:
1.) If both players each have just one identical piece and if neither side can show a forced win.
2.) In K+bishop vs. K+bishop of opposite colors, with only 1 pawn on the board, provided there is no forced win.
3.) K+rook pawn vs. K can be claimed as a draw once the defender is on the rook file in front of the pawn. K+pawn vs. K can be claimed as a draw once the defender is immediately on the square directly in front of the pawn as long as it’s not on the 7th rank.
4.) K=rook+rook pawn vs. K+rook is a draw if the pawn is blockaded by the king and there is no immediate win.
5.) The claimant has a significant material and positional advantage."

These rules are not generally in use in Blitz tournaments in USCF, so only used when the TD has announced it specifically for the tournament. What is generally in vigor is that a flag fall of the opponent only wins if the player has mating material, which is defined as two minor pieces (so 2 knights vs. bare king is OK), a pawn, a rook or a queen. So a lone bishop or a lone knight is not sufficient mating material, and will be a draw if the opponent flags or makes an illegal move.

First of all, these rules are a small subset of "Insufficient Losing Chances" since the latter specifies that is not more likely than 10% that a Category C player will lose to a Category A player in the given position (without judging the actual players strength, by the way). So it is similar to a Appendix G.5 claim.

It is very similar to comparing G.5 and 9.8 since only a small subset of reduced material is claimable in 9.8, while G.5 handles all types of positions, where winning chances are limited.

I did consider each of the clauses whether it could be included in 9.8, and rejected most - see details below.

The first clause with both players having one identical piece is in broad terms similar to the idea of 9.8. I note that B vs. N is not handled. I do wonder if P vs. P is also included, e.g. the meaning of the word "piece"? Probably not. The position must not be a forced win, for either side (I should think whether the claimant is winning, should not retract from his chances to claim a draw, so I don't like that detail). In any event here we see Q vs. Q, R vs. R, N vs. N, and B vs. B. This clause did not inspire any change to 9.8. A player that has any more pieces is out of luck, unless the fifth clause can actually be used.

In the second clause opposite-colored bishops with a single pawn on the board. This is not unreasonable, but doesn't mention how the pawn is restrained, only talks about "no forced win". There is also the consideration that many other opposite-colored bishops endings are just as worthy, because own pawns are protected by the bishop, and thus unassailable for the other bishop or the king. Therefore it would be a bit unfair that in some dead positions you could get the draw just because only 1 pawn is left, while in the rest there is no draw. So I did not include it.

In the third clause there is one subclause for a single rook pawn, and another for any other single pawn. The first is claimable simply from having the king anywhere in the file in front of the pawn, while the second can be claimed when having the king on the square in front of the pawn (thus demonstrating there are no opposition rules in effect) while the pawn obviously could not be on the 7.th. rank when it would be a forced win, at least if the claimant is on move and not stalemated. That the single rook pawn is claimable is not unreasonable, but can also be claimed a draw based on 3-times repetition much before the 50 move rule, unless the pawn is continually advanced, in which case it will still not last more than 25-35 moves. With the second rule I think there are many ways to err for the defender, for instance I found 3790 positions with this material, and 118 of those were won, so there is a distinct chance of a decisive result even though the position should be a tablebase draw. This means 3.11% of the games were won. I think that percentage is too high to allow this to be a claimable draw. The rook pawn has my sympathy, but it can be claimed a draw via 3-times repetition even quicker than any other pawn - why should it have special treatment then?

In the fourth clause there is R+P vs. R where the pawn is a rook pawn. While this is very drawish, there is clearly a greater possibility to win than in a pawnless R vs. R. I counted a total of 1512 positions with this material, although some of these were when one player was immediately losing his rook, so not a drawn situation at all. I did not count how many of those fake positions were included, instead I just counted the actual wins/losses how many were a genuine draw later misplayed and lost by the player with no pawn, and that was 33 games. We get a percentage of 33/1512 positions that were actually won from a drawn position, or 2.18% while 107 were reported wins, or 7.07%

I will just let an example show how easily this can be bungled. This is Alarcon - Melero 2012


71...Ra3 72.Rh4? Not really losing, but a bad idea - the rook should work from the side, so 72.Rb4! would be best. 72...Kg5 73.Rh3?! although
this is not losing, White must concede territory - always a bad idea 73...Ra2+ 74.Kh1 Kg4 75.Rh2?? the decisive error - now discoordination reigns, it is a tablebase win - here White actually resigned! After 75...Ra1+ 76.Kg2 h4 77.Kf2 h3 78.Ke3 Kg3 79.Rb2 h2 it is all over.

In the fifth clause it mentions the claimant has a significant material and positional advantage. This is of course a very subjective statement - essentially any number of pieces or pawns are possible. Quite hilariously there is even no requirement that the opponent has no forced win. What is a significant material advantage? Is a percentage relevant? An extra piece is good unless your opponent has a mating attack. But if he has a mating attack, is that a "positional advantage" for him? I just don't like this clause at all, it's a mess.

Jesper Norgaard
25-10-2014, 07:03 PM
I checked the 5 million games database how many positions could actually be claimed a draw via 9.8(1), when the claimant has more than the lone queen. There were 12,167 games that were Q vs. Q + material, and out of those 7,384 were won by the player with Q + material. As expected, the extra material made that player happily reach wins in a high number of games, 7,384/12,167 were won corresponding to 60.69%. All of these games could have been claimed a draw by the player with extra material if 9.8 existed. The point I am getting at is that in some games the player that had extra material actually lost against the lone queen. That happened in 44 games, so in 44/12,167 corresponding to 0.36%. There are some pretty spectacular blunders in this set, if anyone is interested I can provide the PGN file.

Quite a number of games were lost while trying to press home the full point, for instance trying to avoid perpetual check and walking into a mate. That would probably happen less frequently if the player with extra material is quite happy with a draw, and just shuffles the queen around.

Here is a typical example, Konshina - Stolnikova 1967


White is trying to utilize his 2-pawn advantage with big practical problems. 90.Qf6+ Kg8 91.Qd4 Qa3+! 92.Kd5 Qa8+ 93.Ke5 Qe8+ 94.Kf5 Qc8+ 95.Kg6 Qe8+! 96.Kh6 Qe6+! Black is using the whole board, and here White even has to think about saving the game and instead went astray with 97.Qf6?? Qh3+ 98.Kg6 Qh7#.


Instead also losing is 97.Kh5?? Qh3+ 98.Kg6 Qh7+ 99.Kf6 Qh8+ 100.Kf5 Qxd4 winning the queen.


The only move was 97.g6! Qh3+ 98.Kg5 Qg3+ 99.Kf5 Qh3+ 100.Ke5 Qh8+ 101.Kd5 Qh5+ 102.Qe5 Qxg6 103.Qe6+ this looks winning, but it's just a stalemate trap 103...Kh8! 104.f5 Qf7 in fact it is a tablebase draw.

I admit there are some winning chances for the player with the lone queen when the player with extra material goes all out for a win, but still the percentage of wins 0.36% is quite low. This percentage will not ever be 0% because of the human factor - blunders will happen. The question is only if we are prepared to allow a draw to be claimed despite this chance to win. My opinion is that it is justified.