lichess.org
Donate

Timeout rules

Congratulations to the player(handicap) who knows how to play chess to use this rule
White material: king at a1
Black material: pawn at a2, king at c2

It's white's turn, and white times out.

Mephostophilis, if I understood you correctly, this would cause white to lose the game, because black has more than a king.
however, it's obviously a draw. (white king a1->a2)

I'd really like this site to implement the chess rules correctly!
Almost, but nope, because that's a draw by Insufficient Material. :)

Like I said, it has to be where exactly one side cannot checkmate, not "neither side can checkmate", which was the type of position you gave.

If neither can checkmate you are not allowed to draw by time-out because the game is officially a draw instantaneously.
I admit, I can't find the counterexample you requested...

so you say #22 should be draw before the timeout? that's difficult to detect, could require some recursive calculation, but it has to be quickly detectable, because the game might go on...

I'd prefer to let the game go on even if it's predetermined to be a draw, until it's a quickly detectable draw.
but if the game times out during this "let go on", then do the more complicated draw or loss check.

the problem, why complete draw detection is difficult, is that (I think) it needs to be recursive on future moves, so it can't be done while the game is running.
I made a possible answer in post #20 of this thread explaining what seems to be one counterexample.

The basic concept revolves around the inflexibility of having one single bishop in the endgame, limited to one color of square the entire course (32/64 squares), while the opposing rook (or even two rooks) seem to make checkmate by the enemy King + bishop impossible?

So that might be one possibility, not sure.

But if it was King + 2 rooks against King and Knight, then no, King and Knight could potentially checkmate the King + 2 Rooks because rooks cannot block knight checks. against bishop though, I have yet to find a way

Yes, the FIDE rule is that "Insufficient Mating Material" requires a game to forcefully end in a draw the instant that checkmate is no longer achievable by either side.
Hence, the title of the rule is a bit of a misnomer.
There is an infinite array of possibilities of where that rule could apply, yet all 16 pawns could still be on the board! And still checkmate is physically impossible for either side to fuck up and reach.

I agree, it is perhaps the most challenging of all chess rules for thibault to attempt to implement on this website. The only solution I can think of would be to create a chess engine off of the Stockfish source which uses "Helpmate" mode to see if either side can commit suicide and be checkmated and externally request the answer from an engine after each player's move.
So, example setup.

Black King on a4
White King on d1
Black bishop on c4
White rook on b1.

We all know that White could checkmate if Black threw away his bishop senselessly, but can we position the pieces where all 4 of these men are on the board, such that Black has checkmated White?

I haven't found a way.
Stupid single-square-color bishop. That was the reason I almost drew my last game against AI Level 5 T_T, I hate those endgames.
>>but can we position the pieces where all 4 of these men are on the board, such that Black has checkmated White?<<

White material: king at a1, rook at a2
Black material: king at c1, bishop at c3

>>The only solution I can think of would be to create a chess engine off of the Stockfish source which uses "Helpmate" mode to see if either side can commit suicide and be checkmated and externally request the answer from an engine after each player's move.<<

me too,

BUT: I fear that this will slow down the game, because the engine takes too long,

my suggestion: ignore that rule saying there should be immediate draw, instead, wait until the draw is quickly detectable (e.g. insufficient material), or until timeout; in case of timeout, do the lengthy external engine request.
"White material: king at a1, rook at a2
Black material: king at c1, bishop at c3"

sorry, my fault, it should read:

White material: king at h1, rook at h2
Black material: king at f1, bishop at f3
About the position you gave, that didn't work because White plays Rh2:g2.

Black then wins the rook for free with Bxg2+, which brings us to a dead position immediately drawn by Insufficient Material.

- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - b - -
- - - - - - - R
- - - - - k - K

- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - - -
- - - - - - b -
- - - - - k - K

Now, if we put an imaginary square to extend the standard 8x8 grid of chess squares, we could include a theoretical, alternate direction the bishop could check from, and then it would be checkmate.

But, with the standard 8x8 chessboard boundary, the rook seems always there to get in the way of checkmate. :( Because, with the enemy King in the corner of the board, there is exactly only one possible direction/angle your bishop can deliver check from, which is insufficient because the rook always seems to adjust.

So I suppose it's not possible.
But hey, we can add it to the list!

on (TimeOut)
{
if (enemy has nothing but a king) draw
else if (enemy has exactly only a King and at least one rook/queen) && (I have only King + bishop) draw
else {I lose}
}

something like that
Other part of your reply,

"me too,

BUT: I fear that this will slow down the game, because the engine takes too long,"

I suppose, but it is still better than loads of people getting away with draws on timeout when their opponent could still checkmate them. That is the current system.

Doing the lengthy HelpMate engine request might be slow and unable to keep up with rapid chess play, but I suppose the ulterior issue is that it would almost certainly overload the server. :(

stupid network stuff T_T

But, the standard cases for Insufficient Material are King vs. King (duh, draw, no pieces), King vs. King + Knight, and King vs. King + Bishop.

Those are the fundamental reasons why people care to have the rule.

Past that, it may be FIDE worldwide chess regulations to instantly force a draw once neither side can checkmate, but local tournaments within the USA or whatever are starting to overthrow it due to its difficult-to-define nature in terms of algorithm of the chess position.

This topic has been archived and can no longer be replied to.