Do you have to compete promotion to win ?

[Jesper Norgaard]

To make it crystal clear that an illegal move must always be handled by reinstating the position before
the illegal move (except when making an illegal move actually loses the game), this definition should
be unequivocal:

========= Suggestion ==========================================================
7.5.a
If during a game it is found that an illegal move has been completed, the position immediately before
the irregularity shall be reinstated.

1. 1. If the player has moved a pawn to the furthest distant rank, pressed the clock, but not
   replaced the pawn with a new piece, the move is illegal. If a pawn move with that pawn is legal,
   the player must promote the pawn to a queen of the same colour from the reinstated position.

If the position immediately before the irregularity cannot be determined, the game shall continue
from the last identifiable position prior to the irregularity. Articles 4.3 and 4.7 apply to the move
replacing the illegal move. The game shall then continue from this reinstated position.
========== End Suggestion =====================================================

The old definition involving a pawn push without substituting promotion piece, is currently:

"If the player has moved a pawn to the furthest distant rank, pressed the clock, but not replaced the pawn with a new piece, the move is illegal. The pawn shall be replaced by a queen of the same colour as the pawn."

It is not clear from this definition whether or not the position before the illegal move must be reinstated. It sounds like
the arbiter can just take a queen of the same colour and replace the pawn with it, and the game can continue from that
position. That would be an error.

Additionally the old article insists on that the pawn must be replaced even if that particular move were illegal. The position of E. Michael
White very appropriately demonstrates that:

[Date "2016.04.22"]<br> [White "7.5"]<br> [Black "?"]<br> [Result "*"]<br> [SetUp "1"]<br> [FEN "b1n5/kP2K3/1R6/2P3n1/q7/8/8/3r1r2 w - - 0 1"]<br> [EventDate "2016.??.??"]

All the moves b7xa8, b7-b8 and b7xc8 may be possible, but it may be that not all are legal, and it may be that
articles 4.3 and 4.7 will restrict the choice further.

The formulation "If a pawn move with that pawn is legal, the player must promote the pawn to a queen of the same colour from the reinstated position."

The theoretically possible moves with a white pawn on b7 is (from any position)
b7xa8=Q
b7xa8=R
b7xa8=N
b7xa8=B
b7-b8=Q
b7-b8=R
b7-b8=N
b7-b8=B
b7xc8=Q
b7xc8=R
b7xc8=N+
b7xc8=B

... but since it is a requirement the player must promote the pawn to a queen
of the same colour, only these are allowed
b7xa8=Q
b7-b8=Q
b7xc8=Q

... and since only capturing the black knight will avoid that the king is in check
only this move is allowed
b7xc8=Q

One reason the position before the illegal move must be reinstated is that only by
letting the player perform the exchange in his own time, will the time keeping of
this crucial moment be correct, ensuring that the player does not benefit
neither timewise nor otherwise from the illegal pawn push.

I am fairly certain it was not the intention that the article should be misinterpreted
to that the pawn can just be replaced with a queen and the game continue from there.
However, the above redefinition should make it obvious that the article does not allow that.

[Stewart Reuben]

Jesper >All the moves b7xa8, b7-b8 and b7xc8 may be possible, but it may be that not all are legal<

The move b7-h8 is just as possible as b7-b8=Q mate. Neither are legal chess moves. The only legal ones are b7xc8 with one of the 4 promotions.

Why do you want to shuffle the pawn back and forth? Why not make 7,5a para 2 more sensible?

[Stewart Reuben]

Brian Towers >Doing a bit of delving I suspect that if there was a specific law then it would have been in the 2005 edition.<

This is about the Law concerning capturing the opponent's king. It is in the current Laws, but I simply forgot.

1.2. ... Leaving one's own king under attack, exposing one's own king to attack and also 'capturing' the opponent's king are not allowed.
This was in the 2005 Laws, but not the 2001 Laws.
Rapidplay or blitz game no arbiter observing the game.
White Ke1. Qe6 Rg2. Black Ke8 Re7.
1 Qg6+ illegal.
1...Kf8. Black's move has legitimised White's illegal move.
2 Qg8+ mate, although his king is in check. So Black can claim a win, but not if he removes the white king from the board.

Let us make it standardplay.
1 Qg6+ The illegality is not noticed
1...Kf8 is not that illegal?
2 Qg8 mate. White has made two illegal moves. He loses. Or is it only one?

[Jesper Norgaard]

Jesper >All the moves b7xa8, b7-b8 and b7xc8 may be possible, but it may be that not all are legal<

The move b7-h8 is just as possible as b7-b8=Q mate. Neither are legal chess moves. The only legal ones are b7xc8 with one of the 4 promotions.

Of course the legality of a move will have to take all specifics of the position in consideration, but there is no position where b7-h8 would be legal. The move b7-b8=Q would be perfectly legal if the white king was not in check.

Why do you want to shuffle the pawn back and forth? Why not make 7,5a para 2 more sensible?

I would love to learn how to make it more sensible, so please forward your suggestion!

In fact I'm not trying to shuffle the pawn back and forth, and would like to hear how you imagine that could be expressed. Of course the elephant in the room is that the requirement of promoting to a queen needs to be taken note of before actually reinstating the position. And this requirement is only valid if the player pushed a pawn to the last rank, but didn't replace it with a new piece, and only if there is actually a legal move with that pawn. When all these constraints are fulfilled, the position now needs to be reinstated moving the pawn back, and the clock must be started for the player, but the arbiter must take note that the pawn can only be promoted to a queen. b7xc8=N checkmate would not be allowed even if it is a legal move.

[Jesper Norgaard]

Brian Towers >Doing a bit of delving I suspect that if there was a specific law then it would have been in the 2005 edition.<

This is about the Law concerning capturing the opponent's king. It is in the current Laws, but I simply forgot.

1.2. ... Leaving one's own king under attack, exposing one's own king to attack and also 'capturing' the opponent's king are not allowed.
This was in the 2005 Laws, but not the 2001 Laws.
Rapidplay or blitz game no arbiter observing the game.
White Ke1. Qe6 Rg2. Black Ke8 Re7.
1 Qg6+ illegal.
1...Kf8. Black's move has legitimised White's illegal move.
2 Qg8+ mate, although his king is in check. So Black can claim a win, but not if he removes the white king from the board.

Let us make it standardplay.
1 Qg6+ The illegality is not noticed
1...Kf8 is not that illegal?
2 Qg8 mate. White has made two illegal moves. He loses. Or is it only one?

Very clever. But the arbiter must note that a checkmate with own king in check is not a legal move, and the position must be reinstated to before the first illegal move. Both players may well have made two illegal moves since the incident started - they can't both lose because of too many illegal moves. That is the correct arbiter decision IMHO. Only one illegal move is counted, the first illegal move (1.Qg6+). What went in between is not counted.

Surprisingly this is different in Rapid and Blitz. Here each new move exonerates the illegality of the previous move. Only the last illegal move counts. So yes white loses.

In fact it seems contradictory to me that an illegal move can exonerate an illegal move. It is mindboggling. I think FIDE should have avoided all that by making uniform laws for reverting series of illegal moves so that Standard, Rapid and Blitz are handled identically.

I can only see arbiter laziness as a bad excuse for justifying a quick result rather than a correct result.

[Brian Towers]

Brian Towers >Doing a bit of delving I suspect that if there was a specific law then it would have been in the 2005 edition.<

This is about the Law concerning capturing the opponent's king. It is in the current Laws, but I simply forgot.

1.2. ... Leaving one's own king under attack, exposing one's own king to attack and also 'capturing' the opponent's king are not allowed.
This was in the 2005 Laws, but not the 2001 Laws.

Deeply reluctant though I am, since I'm on the opposite side of the argument, I can't stop myself from helping you out here.

'Capturing' the opponent's king is not allowed but it is not 'illegal' according to 3.10.

3.10 a. A move is legal when all the relevant requirements of Articles 3.1 — 3.9 have been fulfilled.
b. A move is illegal when it fails to meet the relevant requirements of Articles 3.1 — 3.9
c. A position is illegal when it cannot have been reached by any series of legal moves.

Article 1.2 is clearly not one of articles 3.1 through 3.9
Obviously I would disagree with this, well 'interpretation' isn't quite the right word, perhaps 'claim' or 'argument'.

Let us make it standardplay.
1 Qg6+ The illegality is not noticed
1...Kf8 is not that illegal?
2 Qg8 mate. White has made two illegal moves. He loses. Or is it only one?

It is only one.

I had a superficially similar case just over a year ago. A 60+ year old, ~1600 player came to complain to me that his 12 year old ~1250 opponent had made 3 illegal moves, failing to move out of check, and he was claiming the game. More in hope than expectation I turned to the kiddie and asked him if this was an accurate description of what had happened. I was rather disgracefully hoping that the little lad would lie through his teeth and make my life easy. Needless to say, he confessed all, adding that his opponent had failed to say "Check". Oh dear!

Since I wasn't going to default the lad, who was barely holding back the tears at this stage, and I didn't see much point in getting into all the "did you press the clock" business since the old boy knew the rules (albeit slightly out of date regarding the number of illegal moves allowed), I fell back on the preface and some dimly remembered stuff from the Open University philosophy degree I did in my late 20's and early 30's. There's a branch of philosophy called epistemology which has to do with how we know that external objects exist. Somebody called Bishop Berkeley arguing that "to be is to be perceived" asked the famous question "If a tree falls in a forest and nobody is around to hear or see it fall did it really fall?" Berkeley's argument, I think, was that God hears and sees it fall. Therefore it exists.

In this case I argued that for the purposes of punishment a corrected illegal move only exists as a punishable offence if I as the arbiter am told about it and take action before anything else untoward happens. You can't be punished for a second illegal move unless you've previously been punished for a first one because you've missed out on the warning of the consequences that a second illegal move will bring.

Your case, Stewart, is slightly different and much easier to deal with. The first illegal move was uncorrected. Therefore the correct action is to return to the position immediately before the illegal move, dole out warnings and extra time and have the player make a legal move. Therefore the second illegal move doesn't exist within the context of the game.

[Stewart Reuben]

Jesper, Brian
You both seem to be agreeing with me. The player completes an illegal move, the opponent responds with an illegal move. They have both transgressed and one should not be penalised rather than the other.

Glossary
illegal 3.10a (is where illegal first appears). A position or move that is impossible because of the Laws of Chess.
3.10c. A position is illegal when it cannot have been reached by any series of legal moves.
A position with only one king on the board is thus illegal. The act of taking the opponent's king is stated to be against the basic Laws of the game.

What about a player who takes off his own king? The word capture in the glossary refers to a player capturing an opponent's piece. There is no suggestion that he can capture one of his own. Just as well, otherwise there would be a whole series of new checkmates.

Jesper, I love your idea of relative illegalities. Thus b7-b8=Q is slightly illegal. b7-h8 is severely illegal. I wonder how many ways a move can be illegal all at the same time.

[Jesper Norgaard]

Stewart
I don't know if it is relevant or not, but I thought I could bring up the notion. I have been thinking a bit about how the rules are currently organised concerning a pawn push without choosing a promotion piece.

If that happens, the current rules should result in that only promoting to a queen is allowed. I think that is appropriate.

However, what happens if the player is really skilled in scheming? He might realize in the position constructed by E. Michael White that he is not sure which promotion piece to use for the square c8, but instead can utilise that the move b7-b8=Q+ is illegal

In the current rules:

If he plays 1.b7-b8=P? then he must play 1.b7xc8Q and will be mated with 1...Rf7+ checkmate.

If he plays 1.b7xc8=Q? then the move is not illegal so the move will stand, and he will still be checkmated with 1...Rf7+.

If he plays 1.b7-b8=Q+! then the move is illegal, but he still retains any option for promoting on c8, e.g. 1.b7xc8=N+ checkmate is still possible.

Perhaps then it would be a good idea to modify the rule so that any illegal move will lead to that only a new queen is allowed for a promotion move. That would thwart the scheming.

Stewart, if you have an idea how the rule could be formulated, I would appreciate it a lot. Perhaps you can make it more sensible. Others are of course also encouraged to comment or make suggestions.

[NickFaulks]

I have not had the patience to follow this thread in detail, but the cases discussed seem to have practical importance only when a player does something illegal when short of time. This leads me to a general complaint about the Laws, which I have argued for many years.

Suppose that a player who is very short of time commits an infraction. The clocks are stopped and an arbiter is summoned. The position is explained to him and he makes a ruling. He adjusts the position and the clocks as required and the games is restarted. Throughout the intervening period, the players have of course had at least half of their minds on the position on the board.

I'm sorry, this is no longer the same game. The game as it was before the infraction, with momentum favouring one player or other, is irrecoverably dead. In my opinion the player deemed to be at fault should lose two minutes of his time. If that means that he loses on time, then so be it. The current "penalty" that his opponent gets 26 minutes instead of 24 is meaningless.

[Ian Thompson]

However, what happens if the player is really skilled in scheming? He might realize in the position constructed by E. Michael White that he is not sure which promotion piece to use for the square c8, but instead can utilise that the move b7-b8=Q+ is illegal

In the current rules:

If he plays 1.b7-b8=P? then he must play 1.b7xc8Q and will be mated with 1...Rf7+ checkmate.

If he plays 1.b7xc8=Q? then the move is not illegal so the move will stand, and he will still be checkmated with 1...Rf7+.

If he plays 1.b7-b8=Q+! then the move is illegal, but he still retains any option for promoting on c8, e.g. 1.b7xc8=N+ checkmate is still possible.

Perhaps then it would be a good idea to modify the rule so that any illegal move will lead to that only a new queen is allowed for a promotion move. That would thwart the scheming.

It might thwart this particular scheme, but it would not be effective in stopping the unacceptable conduct, so it's pretty pointless. (If the rules were as you suggest, the player could just play any king move in E. Michael White's position, leaving him free to play any legal move he likes once the illegal move is retracted because the king has no legal moves.)

If a player is short of time and decides to deliberately play an illegal move to get the clocks stopped while it's corrected, so he gets extra thinking time, the only way to deal with that under the current rules is to have supervision by an arbiter and to have an arbiter who is prepared to apply an appropriate penalty for such behaviour. The difficulty, of course, is that it's almost impossible to determine that a player deliberately played an illegal move, so they're very likely to get away with it. The rules would be much better if they applied an automatic time penalty to the player who made the illegal move as well as giving extra time to their opponent.

[NickFaulks]

The rules would be much better if they applied an automatic time penalty to the player who made the illegal move as well as giving extra time to their opponent.

You have made no reference to my post immediately above, but I take it that you are in complete agreement. Does anyone wish to disagree?

[Ian Thompson]

The rules would be much better if they applied an automatic time penalty to the player who made the illegal move as well as giving extra time to their opponent.

You have made no reference to my post immediately above

That's because it wasn't there when I started writing my post.

I take it that you are in complete agreement.

Yes.

[Stewart Reuben]

The only easy way to solve the problem Nick mentioned is for the first illegal move to lose. This was Nigel Short's first suggestion to me, which resulted in the change from 3 illegal moves to lose to 2. This had the idea that in due course it might be shortened to the first illegal move losing.
Since then Nigel has changed his mind. I, too, made an illegal move against Chris Ward about 20 years ago. I felt very uncomfortable about continuing the game. Indeed I should have drawn had I not blundered.
Many people are unhappy about the first illegal move losing in rapidplay. The winner has complained to me about his game being spoilt on at least one occasion. He wanted to play, not win by forfeit.
It gets worse with pushing the pawn to the 8th rank and pressing the clock.

Up thread I asked how illegal a move could get. Here is one.
White Ke1 Ra1, Rc3 Re3, Ba4. Bf3, P d2 and also a7 through h7.
Black Kd3, Pf2
White plays 0-0-0 illegal.
He removes his own Pd2. illegal. But now mate.
But he has 8 pawns and 3 rooks. Illegal.
He has 2 White squared bishops illegal.
How did the black king get to d3?. illegal.
OK, remove all those pawns.
The arbiter coming to the position will initially see mate and nothing unusual. It is only after extensive analysis that he realises Black's previous move was illegal.

Jesper asks me to modify 7.5a para 2.
Change the word illegal to' incorrect'. Define incorrect in the glossary. Since pushing the pawn to the 8th rank and pressing the clock is no longer illegal, this removes the problem of mating with an illegal move and the game thus ending up as a draw.

7.5b. After the action first completed illegal or incorrect move by a player, the arbiter shall use his best judgement to determined the penalty awarded. Normally the opponent shall be awarded an extra two minutes. But he may decide this is too disturbing and that it is better simply to continue the game without altering the clock times.
This is similar to 7.1 and I expect they can be amalgamated.

An alternative might be to permit the opponent to refuse the penalty.

This is only a first conceptual draft. But it does seem to overcome Nicks' objection.

I have my doubts that I can get it through the Rules Commission. But David Welch and Alex Holowczak will be in Azerbajan.

[NickFaulks]

The only easy way to solve the problem Nick mentioned is for the first illegal move to lose.

That is the only easy way to solve it completely, but it is draconian and I'm only really troubled by cases where the player is very short of time, which is where real problems generally occur.

In the most common case, where the illegality consists of putting / leaving the king en prise, I have never understod why, if noticed by the opponent, this should not lead to the king being taken and the game lost. What possible reason is there for the penalty for leaving a queen being left en prise generally amounting to instant loss of the game, whereas in the case of the king it is randomly somewhere between loss of the game and negligible? I have neved heard any explanation other than "that's the way it has always been done".

Why was the rule ever made this way? I'm sure we have all heard the old story about regicide being anathema, but I don't even understand that. Killing kings could not be contemplated, but killing queens or bishops was to be encouraged? It's all nonsense.

[NickFaulks]

But it does seem to overcome Nicks' objection.

It barely even makes a start. The only element I like is that the players could have the right to say to the arbiter "please go away and let us get on with our game", an idea which might then spread more widely. As you say, it won't happen.