Completing the Move, Touch-Move, and Stalemate

Running Chess Tournaments
1

The following sequence of events occurred at a recent tournament:

White moved one of his pieces, putting Black in check. White’s move was determined upon letting go of his piece. White never pressed his clock.

Black presumably did not notice that White never pressed his clock.

Black then touched both one of his own pieces and the White piece that gave check with the deliberate intention of capturing the White piece. Under the touch-move rule, this would produce only one permissible move–Black capturing White’s piece. This move produces stalemate.

Seeing this, White then jumps up and shouts “draw!” and walks away from the board. In the midst of White’s distracting reaction, Black releases his capturing piece.

Several moments later, Black realizes that White never pressed his clock. As a result, Black claims that White never completed his previous move (the check) such that Black need not make the capture.

How do you rule?

2

Bullfeathers. Black’s jumping the gun (by starting his move before white pressed his clock) does not excuse black from touch-move and determined-move obligations. The stalemate stands.

Had black’s move not been stalemate, white’s failure to press the clock promptly would have simply meant that white could still press his clock even after black moved. Then black would have to notice, and press his own clock again, so that the correct clock would be running. In this case, though, the clock was irrelevant since the game had already ended in stalemate.

It might have been advisable, though, for white to have postponed the jumping up and shouting “draw” until the situation had settled down a bit. If black then tried to take back the stalemating move, white could summon the arbiter, point out the touch-move violation, and get the ruling he needed.

Bill Smythe

3

Did Black actually complete the capture? It sounds like White left the board with his clock running which means that White could still flag if the position on the board isn’t stalemate. (Being forced by touch-move rules to make a move that is stalemate isn’t the same thing as making the move.)

4

From the 4th edition:

9B. In the case of a capture, the move is determined … when the captured piece has been removed from the chessboard and the player’s hand has released the capturing piece …

Somewhere between the 4th and 5th edition (I remember seeing it in a rating supplement), and continuing in the 5th and 6th editions, the rule was changed to:

9B. In the case of a capture, the move is determined … when the player has deliberately touched both his or her own piece and the opponent’s piece …

I’ve never liked this change, but it seems to me the difference could affect whether white could still flag in the situation you describe above.

Bill Smythe

5

I’m pretty sure I know why the wording was changed, since a similar thing came up back in the days of the 3rd edition rulebook at a tournament in which a chess team I was coaching was playing.

One of my players had begun capturing an opponent’s piece with his queen, but after picking up the opponent’s piece and placing his queen on the square, but before he had released the queen, he noticed that his opponent had a pawn protecting the square and that he would lose his queen if he completed the move. The question was whether, at that point, he could choose to move his queen somewhere else, instead. The wording used in the 3rd (and 4th) editions seemed to imply that he could, since he hadn’t yet “released the capturing piece”, but a counterargument was made that since he had touched both his queen and the opponent’s piece, the touch-move rule required not only that he move his queen, but also that he capture the opponent’s piece. But then we made the argument: Why, then, would Rule 9B (it had a different number in the 3rd edition), which is specifically about capturing, say that the move is determined when “the player’s hand has released the capturing piece”? Ultimately, it was decided that the touch-move rule prevailed, but we all agreed that the ruling of Rule 9B was confusing and misleading.

Being compelled to make the move caused our player to lose the game (which until then he had been winning), which caused us to draw the match instead of winning it, and that ultimately bumped our team from 1st place down to 3rd place in the final standings.

Bob

6

Eek. Yes, with that, it appears that if QxR is checkmate, and you touch the Queen and touch the Rook, under USCF rules you can’t flag because you have “determined” checkmate even if you’re still thinking about your move.

7

Your team had no valid complaint. Your player ran afoul of the touch-move rule, rather than (either version of) the determined-move rule. He had touched both his own piece and his opponent’s piece, so he was required to take his opponent’s touched piece with his own touched piece. The touch-move rule has been unchanged since the beginning of time. No change in the determined-move rule would have made any difference.

As a result of this rule change, the determined-move rule is now more confusing than it was before.

Also see my next post, below.

Bill Smythe

8

IMHO the change in the determined-move rule between editions 4 and 5 came about through fuzzy thinking, by falling into two common fallacies that often plague rule-writers.

The first fallacy is the “OMG, somebody misread a rule, we’d better try to clarify it” syndrome. In this case, the correct ruling was already clear from a different rule (the touch-move rule), which has remained unchanged for centuries. The unfortunate change in determined-move simply muddied the waters elsewhere.

The second fallacy is the failure to distinguish between two different concepts of “determination”, and the resulting unfortunate combining of the two.

The first concept of “determined” (I’ll use lower case) seeks to define the moment at which the player is no longer allowed (possibly because of other rules such as touch-move) to play a different move.

The second concept of “Determined” (upper case) lists all the actions that a player is supposed to complete before he presses his clock.

It’s like the difference between democratic and Democratic. Perhaps a different word may be in order. How about “Executed” for the latter? I’ll even use upper case.

Rule 9B (along with other rules around it) should concern itself with defining when a move has been Executed. Thus, in the case of a capture, the original definition is better than the revision. A capture should be regarded as Executed when the player’s piece has been released on its new square and the captured piece has been removed from the board. Any move (capture or otherwise) must be Executed before the player presses his clock.

Now that we have two words, is the concept of “determined” even useful? YMMV, but I’d suggest, for example, that “determining” a game-ending move (checkmate, stalemate, dead position) should be sufficient to avoid losing on time, and that completely “Executing” the move should not be necessary.

In most other situations, “Executed” seems more relevant than “determined”.

Bill Smythe

9

Why? That makes no sense. To avoid loss on time, you should have finished “executing” your move. Period. If you’re moving to a vacant square, that’s what you have to do. If you’re promoting, that’s what you have to do. But if you’re castling, you don’t have to even move (or even touch??!!) the rook (the piece that’s actually going to deliver the check) or if you’re capturing you just have to deliberately touch both the capturing piece and the target piece? That’s absurd.

10

That’s fine. That way, we don’t need both concepts (determining and Executing). We can just use the latter, in all contexts, all the time. We could even revert to using “determined” instead of “Executed” as the official name of the (single) concept.

What bugs me is that the 5th edition version uses the “determined” concept for captures, but the “Executed” concept for other move types. This inconsistency causes problems. The 4th edition version used the “Executed” concept for all move types.

When I wondered out loud whether both concepts were even necessary, I tried to come up with a situation where the “determined” concept just might be considered preferable to the “Executed” concept. The only example I could even remotely consider was the game-ending situation – checkmate, stalemate, or dead position. And, as you pointed out, that doesn’t work well either.

So, let’s stick with the single concept, and revert the rule back to the 4th edition version.

Bill Smythe

11

OK, let’s try this one more time (he said, after a night’s sleep and a little re-thinking).

Instead of two words for two related concepts, let’s refer to the “purist definition” and the “practical definition” of a determined move.

The purist definition, obviously, is that a legal move is determined at the moment when, for any reason (including touch-move), all other moves become disallowed.

With this definition:

  • A (legal) simple move is determined when the player has released the piece on its new square.
  • A (legal) capture is determined when the player has touched both his own piece and the piece to be captured.
  • A (legal) promotion is determined when the player has caused a piece off the board to touch the promotion square.
  • A (legal) castling move is determined when the player has released the king on its new square.

And now for the practical definition:

  • A (legal) simple move is determined when the player has released the piece on its new square.
  • A (legal) capture is determined when the player has released his piece on its new square and removed the captured piece from the board.
  • A (legal) promotion is determined when the player has released the promotion piece on its new square and removed the pawn from its original square.
  • A (legal) castling move is determined when the player has released both the king and the rook on their new squares.

The question then becomes, which definition should be used:

  1. to define the moment at which it becomes legal for the player to press his clock, and
  2. to define the moment at which the player avoids a time forfeit in a game-ending situation (checkmate, stalemate, or dead position).

I submit that, in case A, the practical definition should be used. For example, a player should not be permitted to press his clock until all involved pieces are where they belong (whether on or off the board).

The problem is that, beginning with the 5th edition, the purist definition is used for captures while the practical definition is used for other move types.

Of course, in case B, the purist definition must be used – because, in effect, that’s already the rule in these cases.

Bill Smythe

12

Sorry, Bill, but this is absolutely not true!

First, the “purist” and “practical” definitions are the same for simple moves, so let’s ignore that.

And let’s also ignore, for the moment, the case of promotion, because none of the editions exactly conforms to either your “purist” definition or your “practical” definition.

So the real situation is that in the 4th edition, the “purist” definition was used for castling, but the “practical” definition was used for captures (which, like it or not, was confusing). But in the 5th and subsequent editions, the “purist” definition is used for both.

Let’s examine the castling rule. In the 4th edition, it reads:

9C. Castling. In the case of castling, the move is determined with no possibility of change when the player’s hand has released the king, which has moved two squares toward a rook, and completed when that player, having transferred the rook to its new square, punches the clock.

In the 6th edition, this rule is almost identical:

9C. Castling. In the case of legal castling (8A2, 8C1, 10I1, 10I2), the move is determined with no possibility of change when the player’s hand has released the king, which has moved two squares toward a rook, and completed when that player, having legally transferred the rook to its new square, presses the clock (5H).

Note that the only differences in the wording of these two versions are (a) the inclusion of the words “legal” and “legally”, (b) the inclusion of references to other related rules, and (c) replacement of the word “punches” with “presses”.

In both versions, castling is “determined” at the moment when the player has released the king, but before the player has moved the rook, which corresponds to your “purist” version. So the 5th and 6th editions use the “purist” version for both castling and capturing, whereas the 4th edition used it for castling but not for capturing.

To determine whether the “purist” version or the “practical” version is best, we need to consider the purpose of the distinguishing between “determination” and “completion”. The term “completion” is intended to define the moment at which one player ceases to be “on move” and the other player begins to be “on move”. But it was felt that there needed to be a separate definition that did not depend on pressing (or “punching”) the clock button, but when the player was committed to a particular move (hence the wording, used in all 3 editions: “determined with no possibility of change”). The purpose was not to define the moment at which a move has been fully executed, but the moment at which the player is committed to a particular move.

And herein lies the complexity of the promotion rule, for there are really two parts to the move: the advancing of the pawn and the promotion. In the 4th edition, the rule reads:

9D. Pawn Promotion. In the case of the promotion of a pawn, the move is determined with no possibility of change when the pawn has been removed from the chessboard and the player’s hand has released the new piece on the promotion square, and completed when that player punches the clock. If the player has released the pawn on the last rank, the move is not yet determined, but the player no longer has the right to play the pawn to a different square.

In the 6th edition, it reads:

9D. Pawn Promotion. In the case of the legal promotion of a pawn, the move is determined with no possibility of change when the pawn has been removed from the chessboard and the player’s hand has released the new appropriate piece on the promotion square, and completed when that player presses the clock (5H). If the player has released the pawn on the last rank, the move is not yet determined, but the player no longer has the right to play the pawn to a different square. See also 8F6, Pawn promotion and 10H, Piece touched off the board.

Neither of these wordings corresponds exactly to either your “purist” wording or your “practical” wording. But the sense of both wordings is the same: The player is not committed to the entire move until the pawn has been removed from the board and the promotion piece has been released on the promotion square, but if he has released the pawn on the last rank, he has committed himself to that part of the move (i.e., moving the pawn straight forward or diagonally to capture a piece). And he is not committed to the choice of promotion piece until he has actually released the piece on the promotion square (i.e., simply touching a piece off the board, or even placing that piece on the promotion square without releasing it, does not commit the player to that choice of promotion piece).

What is very clear from the wording of 9D is that its purpose (and the purpose of all of the determination rules) is to define when the player is committed to a particular move, with no possibility of change. And this is precisely why the wording used in the 4th edition for Rule 9B is so unacceptable! There are, in general, two ways a player can execute a capture (capturing piece Y using piece X):

Scenario A:

  1. pick up piece X
  2. remove piece Y from the chessboard
  3. release piece X on the destination square

Scenario B:

  1. remove piece Y from the chessboard
  2. pick up piece X
  3. release piece X on the destination square

The 4th edition version of the Rule 9B states that “In the case of capture, the move is determined with no possibility of change when the captured piece has been removed from the chessboard and the player’s hand has released the capturing piece, and completed when that player punches the clock.” As worded, this implies that the moment at which the player is committed to his move is, in both scenarios, at the completion of step 3. But this contradicts Rule 10C (“Touching pieces of both colors”), which clearly indicates that, in both scenarios, the moment at which the player is committed to his move is at the completion of step 2. Your argument, as far as I can make out, is that this is perfectly okay because, since the rules contradict each other, it is clear that Rule 10C should prevail and Rule 9B should be ignored. My argument is that a more sensible approach was the one actually used: to modify the wording of Rule 9B so that it agrees with Rule 10C.

It is worth noting that the problem was really created by the invention of the distinct terms “determined” and “completed” in the 4th edition rulebook. The 3rd edition used only the term “completed”, and the meaning of the term did not agree with either the term “determined” or the term “completed” as these terms are used in the 4th and subsequent editions. In the 3rd edition, the term “completed” essentially meant that the move had been fully executed - not that the clock button had been pressed. But the rule for castling included the concept of determination (as it is defined today), and its application to castling was the same as it is now:

I.8. The Completion of the Move

A move is complete–

…

c. In the case of castling, when the player’s hand has released the rook on the square crossed by the king. When the player has released the king from his hand, the move is not yet completed, but the player no longer has the right to make any move other than castling on that side, if this is legal;

In the tournament game I described earlier, our confusion arose because we were trying to use rule I.8.B (which covered capturing) to decide when the capture was “determined” (in the sense that this term is now used), but rule I.8 was simply defining when a move had been fully executed, and it was really rule I.9.B.2 (the equivalent of the modern-day Rule 10C) that defined when a capture had been determined.

And this, I suspect, was the cause of the erroneous wording of Rule 9B in the 4th edition rulebook: They correctly modified the wording of the castling and pawn promotion rules (I.8.C and I.8.D) to properly utilize the new term “determination”, but failed to realize that the proper wording of the rule for “determination” with regard to capturing was not found in rule I.8.B, but in I.9.B.2. But, happily, this error was recognized and corrected in the 5th and subsequent editions.

Bob

13

The concern I have is with 9E combined with 9B and 9C. As I read that (please correct me if I’m wrong), if O-O is check/stalemate and you play Ke1-g1, you do not have to move the Rook to avoid losing on time. (After all, Kg1 “determines” the move as castling, and 9E says that that is the same as “completion” for check/stalemate). Similarly, touch your piece, touch an opponent’s piece you don’t actually have to make the move if it’s check/stale mate. I suspect that this is more likely to be an issue with stalemates then with checkmates—how many times has someone picked up a piece then before moving it realized that any move was stalemate? If someone wants to go into “oops” mode and not move until they convince themselves that they may have just goofed (or perhaps more nefariously, ask themselves whether they can get away with reneging on the move), they shouldn’t be protected by the rules from flagging while they’re doing that.

To me, 9B and 9C seem to be a (not even complete) description of how touch-move applies. For instance, if you touch a piece that has one and only one move, you also have “determined with no possibility of change” or similarly if you touch an opponent’s piece which can be captured one and only one way. That sounds more like a TD tip than a rule.

14

I agree with your analysis of how the rules currently work. And while I was not a TD (or even thinking of becoming one) back in the days of the 3rd edition rulebook, it appears to me that this is a genuine difference between the way things worked then and how they worked under subsequent rulebooks. The 3rd edition rulebook, like the more recent rulebooks, did not require that the clock be “punched” or stopped before the player’s flag fell in the case of checkmate and stalemate. But if I were ruling under that rulebook today, I would conclude that the checkmate or stalemate did not occur until the move had been completed, as the rulebook then defined it (i.e., the move had been fully executed, which would mean releasing the rook on the square crossed by the king in the case of castling, or removing the captured piece from the board and releasing the capturing piece on its destination square in the case of capture, even though in both cases the player was committed to making no other move prior to this). But in the 4th edition, where the concept of determination was introduced, Rule 9E clearly stipulated that in the case of checkmate and stalemate, “the move is completed simultaneously with its determination.” And that meant, even in the 4th edition rulebook, that if a checkmate or stalemate were created by castling, a flag fall after release of the king but before the rook had been moved had no significance, since the game was already over.

I don’t see that either of these definitions is inherently superior to the other, as long as the rules are clear and consistent. In any case, the likelihood of a flag fall between determination of a move and its full execution is not likely for either castling or capture.

Bob

15

This looks like a potentially long debate for rare situations and I’d rather see a long and involved debate happen in the forums as opposed to at a delegates meeting dealing with ADMs (well, I’ll grant that the first time it would be referred to Rules but then it would still come up the following year).

16

The situation is more complex. White jumping up and screaming… Is that a claim? I order for the stalemate to take place, doesn’t the touch—move claim need to be made ??

17

The correct ruling is that the game ends in a draw by stalemate. By rule 9E, a legal move resulting in stalemate is completed at the very same time it is determined. Stalemate immediately ends the game (rule 14A); no claim is necessary. After the game has ended, it is too late for Black to make any claims.

And yes, a legal capture is determined once the player has touched both his piece and the opponent’s piece. It is not necessary to have actually executed the capture on the board. Interestingly, the rules committee had this very situation come up in the consideration of an appeal a couple of years ago.

18

I’m curious and may have missed it in the discussion, but how is a player not on move obligated by the touch-move rule?

  • Enrique
19

It is common enough that a player will forget to press his clock after moving (“determining his move”). Mr. Huerta is absolutely correct that the opponent is not on the move until the player has completed his own move by pressing the clock. Nevertheless, players tend to be pragmatic about this. Some opponents will press the player’s clock on the player’s behalf (yuck!), move, and press their own clock. Most opponents will just move while the player’s clock is running once it is obvious that the player has forgotten. I can’t say I’ve ever had a player claim that the opponent violated rule 9 by moving before the player had pressed the clock. Most players just shrug it off, and the game continues.

(I personally had a game once against a rather elderly gentleman. He was utterly disastrous about remembering to press the clock. I was willing to give him three reminders about his clock. Then, I simply sat completely immobile while his clock continued to run. That ended up being a real mistake. Several times, he was utterly confused and tried to move again before I had moved. I finally gave up, ignored rule 9G [causing me considerable pain!], and simply moved while his clock was running. It was the only practical solution.)

So, should the opponent who is moving when the player has forgotten to press his clock be bound by rule 10 (The Touched Piece)? While it is true the opponent is, strictly speaking, not on the move, practical considerations and the “spirit of the rules” strongly suggests that the opponent should indeed be governed by “touch move.”

In the statement of the situation that prompted this thread, Mr. Swiszcz did not give any indication of how long Black waited before moving while White’s clock continued to run. However, White did not make a claim about the matter before Black determined the legal move that caused stalemate.

(This situation does not arise under the FIDE Laws of Chess because the Laws allow a sequence such as “White moves, Black moves, White presses the clock, Black presses the clock.” Unlike the Official Rules of Chess, the FIDE Laws strictly disallow any draw claims based on triple occurrence of position or the 50-move rule after a player has touched a piece. Under the FIDE Laws, a player wishing to make such a claim must write his move on the scoresheet [if the move is required to cause triple occurrence or the 50 move rule to be effective], not make the move on the board, and make his claim to the arbiter.)

20

Well, let’s see, what are the alternatives?

One is to remind the player yet again to press his clock, even though he has been forgetting, move after move. Double yuck!

Another is to wait indefinitely, not moving, no matter how long it takes. Triple yuck!

Or, you could just go ahead and make your move anyway, risking having a stone-headed TD impose a penalty on you for your opponent’s error. Quadruple yuck!

So it’s back to single yuck. But let’s make a slight modification:

Take as long as you want to think about your move. When you’re ready, go ahead and play your move, then immediately press first your opponent’s clock and then your own.

Hmm, I don’t see any yucks in the above sentence. Problem solved. On top of everything else, you’ve restored both players’ increment times (not to mention the move count, if the time control is not sudden death) and now the game can proceed normally.

Last time I suggested this idea, a few NTDs among you expressed aghastitude at the idea, but I think it’s the most logical approach. In fact, I seem to remember reading some FIDE rule (or FIDE advice) somewhere, maybe in the current rules or the last version, that suggested this very course of action.

Note, too, that if your opponent is the type who repeatedly forgets to press his clock, he is unlikely to notice your quick double-press. And even if he does, what can he do about it?

Absolutely – unless, of course, the forgetful player tries to take his move back, or something similar.

Here is a case where the FIDE rules are superior to their U.S. Chess counterparts. FIDE keeps clock considerations separate from discussions of touch-move and released pieces. In fact, clocks are not even mentioned until a later chapter. This approach delineates different topics and clarifies things marvelously!

IMHO the rules (both FIDE and U.S. Chess) ought to function as follows:

  • Player B is not obligated to wait for player A to press the clock before player B makes his own move.
  • However, player A is still entitled to a clock press.
  • Therefore, if player B makes his move (and dummy-presses his clock) before player A has pressed the clock, player A may then still press the clock also.
  • It is then player B’s responsibility to notice player A’s “extra” clock press, and to press the clock again.
  • Failure of player B to notice player A’s extra clock press could have disastrous consequences for player B, such as a time forfeit. This risk is the price player B must pay for jumping the gun in the first place.

Bill Smythe