[Retros] Illegal moves by grandmasters

Noam Elkies elkies at math.harvard.edu
Wed May 11 11:15:52 EDT 2011

Francois Labelle <flab at wismuth.com> writes:

> I agree that my title is provocative, but I think that the logic behind

> it is sound, so why not? :)

Even if you're write as a matter strictly of logic, that's not the only
consideration here. I'm no lawyer, but from what I've read (mostly about
the application and interpretation of law in the US legal system), the
history of a law's application, legislative intent to the extent that
it can be determined, and even plain common sense, are also relevant.

You evidently recognize this too, by admitting that you're deliberately
staking out a "provocative" position and using ":)", even though you
exclude such ideas from your analysis. There would be nothing provocative
or ":)"-worthy about saying a player had played illegally if they had
castled a second time undetected (this has happened already, albeit not
at GM level). Conversely, it is common for a player in actual tournaments
to Queen a pawn by promoting it to an inverted Rook and saying "Queen"
if there's no actual Queen handy, especially if the promoted Queen will
surely be captured next move; I haven't checked recently but I imagine that
FIDE rules didn't always explicitly allow this even if they do now, and yet
no arbiter would disallow such a move (excluding Soviet shenanigans
and the like) because it's a nearly universal convention.

> Yefim Treger wrote:

>> IMO: an illegal move is a move, which breaks fundamental rules of

>> chess (piece movement, etc.)

> So according to you, some rules are "fundamental" and others are not,

> and breaking a fundamental rule would be illegal, but breaking a

> non-fundamental rule would be called something else (called what?).

> The FIDE rules make no such distinction. [...]

The rules might not explicitly make such a distinction but that doesn't mean
it doesn't exist. The distinction might also be implicit in the penalties
the rulebook prescribes for violating a rule. What consequence would
a player suffer for this kind of "illegal move", and what would anybody
gain from enforcing the penalty? As long as neither player is trying to
win on time in a dead position (which is the only reason for the rule's
existence) it would be pointless.

> Noam Elkies wrote:

>> This kind of "illegality" is a fun addition to the arsenal of a

>> problemist, but doesn't change the outcome of over-the-board games,

>> as long as "dead" draws are still not affected by the clock.

> It's true that A1.3/A5.2b/A9.6 don't change the outcome (win/draw/loss)

> of over-the-board games much, but that's irrelevant. The rules are there

> so Mamedyarov's 69.Kd4 is illegal. If FIDE had wanted 69.Kd4 to be

> legal, then those rules would not be there or they would have been

> written differently. [...]

Legislators don't always realize the full implications of the laws
they pass. For many years the laws said a King is in check when
attacked by one or two of the opponent's units. At some point sombody
noticed the flaw and constructed some joke problems (one was something
like <http://www.janko.at/Retros/d.php?ff=7k/r2n4/5KP1/4Q2b/8/8/8/8>:
1 Kf7+ Nxe5++ 2 g7+! Kh7 3 g8Q+ Kh6 4 Qg7#; another was the observation
that in positions like <www.janko.at/Retros/d.php?ff=rnbqkbnr/8/4K3/8/8/8/8/8>
White seems to mate on the move with Ke7, to which I might add that
is likewise a mate in 2). It was universally recognized as a
joke revealing a flaw in the drafting of the laws, and the next
edition of the FIDE laws corrected it in the natural way
("one or more" instead of "one or two").

A J Mestel <A.J.Mestel at damtp.cam.ac.uk> contributes:

> I don't think I've ever made an illegal move.

> I have, however, made hundreds of criminal moves.

> I deduce crime is not against the law.

Crime is against the law, but in chess as in life punishment is
contingent on being caught. Some of your opponents punished
your criminal moves; others didn't and were punished for their own
criminal negligence ;-)

> I have never for example, seen a problem/study which specified that

> "White has touched his rook and an opponents pawn simultaneously,

> and so must either move the rook or capture the pawn (or both.)"

> Or "White only has 3 seconds for his next 10 moves. Physically, this

> requires that the distance of the moving piece to the clock

> (the left-hand side of the board) must on average be less than

> 2 squares per move."

> Such problems would count as "fairy problems" to us,

I've seen similar things, I think, but indeed as fairy or joke problems.

> but arguably they are closer to over-the-board play than

> your average 2-mover...


But again we (and even OTB players, I hope) share the sense that
even the most abstruse #2 is closer to the essence of chess than
puzzles that hinge on such incidentals as clocks and the biomechanics
of wood-pushing.


More information about the Retros mailing list