Game drawn by a bug

Sort:
Avatar of DarkGrisen

http://www.chess.com/livechess/game?id=1078357071

How is that draw? I'm pretty sure you can checkmate with a queen and a king...

Avatar of csalami

Your opponent run out of time but you have only a king, and you cannot mate with that therefore it is a draw.

Avatar of Eseles
csalami wrote:

Your opponent run out of time but you have only a king, and you cannot mate with that therefore it is a draw.

right, the insufficient material refers to the player who still had time on the clock

Avatar of ThrillerFan

Actually, there is a bug with the server here.  The server issues a draw in any scenario where the side with time has K, K+B, K+N, or K+NN.

There are 2 flaws to that.  Assume it's White that still has time in each scenario:

Scenario 1:  White has K and 2N, and Black has at least 1 pawn on the board.  This is not a draw as long as the pawn is on or behind the "line", which is a4-b6-c5-d4-e4-f5-g6-h4.  Black running out of time should not be a draw here.

Scenario 2:  Best done via illustration.  Basically the "correct" rule states that if mate can be FORCED, it overrides insufficient mating material, but here it doesn't.  See below for a case in point:

Avatar of ChessDoofus

Bugs are not very good at chess.

Avatar of Troll4ever22
dpnorman wrote:

Bugs are not very good at chess.

Truer words have never been spoken.

Avatar of DarkGrisen
FrontyardProfessor wrote:

You got a draw! Be happy.

-The Frontyard Professor

I don't care if i got a draw even if i would lose. I just don't like bugs.

So the rules are basically like this: If let's say white has no materials left except the king and your opponent runs out of time, it's draw?

I thought you always lose a game if you run out of time. You learn something new everyday :)

Now that i think about it. It doesn't sound weird that it will be draw like that

Avatar of bmarkandeyaprasad

yes,Iagree with you. A timed-out player is to be deemed a loser, no matter what material the opponent or he himself is left with. True, as a game is meant for enjoyment, a program is meant for perfection. that is all !  no matter it is w or l d 

Avatar of SwiftyReflex

*edited to reflect a change in an ending game mate situation where I was confused about which 2 3 point pieces the king can move to escape mate from*

*Old version, disregard, skip to new*

Yes, the program should be changed to only allow insufficient material draw if it is in fact only K vs K OR no pawns exist AND it's K+N vs K or K+B vs K.  It all comes down to lazy programming, a programmer would have to change the rules, even if they didn't change it to calculate all mate possibilites, they should still program it so that if there's a pawn left, it assumes the king will get behind it and trap itself into mate with just a 3 point piece (As ThrillerFan's diagram eloquently shows).   Another point, you can't FORCE mate with only K+N+B vs K, however, if the king moves incorrectly (stupid and almost absolutely would never happen) a king CAN be put in a checkmate situation with opposing player only having K+N+B vs K and therefore because of this, the program should also deliver a loss to the player that runs out of time, even if that's all the material the opposing player has.   Basically, if it's physically possible to have a mate situation, meaning anything except K, K+N, K+B vs K, (unless there's a pawn on the board) then the player that runs out of time should lose.

(New Version)

Yes, the program should be changed to only allow insufficient material draw if it is in fact only K vs K OR no pawns exist AND it's K+N vs K or K+B vs K.  It all comes down to lazy programming, a programmer would have to change the rules, even if they didn't change it to calculate all mate possibilites, they should still program it so that if there's a pawn left, it assumes the king will get behind it and trap itself into mate with just a 3 point piece (As ThrillerFan's diagram eloquently shows).   Another point, you can't FORCE mate with only K+N+N vs K, however, if the king moves incorrectly (stupid and almost absolutely would never happen) a king CAN be put in a checkmate situation with opposing player only having K+N+N vs K and therefore because of this, the program should also deliver a loss to the player that runs out of time, even if that's all the material the opposing player has.   Basically, if it's physically possible to have a mate situation, meaning anything except K, K+N, or K+B vs K, (unless there's a pawn on the board) then the player that runs out of time should lose.

Avatar of Martin_Stahl
SwiftyReflex wrote:

...  Another point, you can't FORCE mate with only K+N+B vs K, ....

That is incorrect. That is a forced mate if you know how to do it. Something like 32 moves with perfect play from the worst initial configuration of the pieces.

Avatar of Brian-E

There are many theoretical examples, similar to the scenarios which ThrillerFan gives above, where a player can deliver mate with K+B, K+N or K+N+N without pawns. They are rare, but not inconceivable, certainly at the end of a forced sacrificial line. K+N+N vs K+P occurs now and then in practical play.

 

Does the server indeed declare a draw in all cases if the opponent runs out of time in that situation, or does it look more intelligently at the position first? And if it's the former, can the duped player who should have won appeal to administrators to reset the result in his/her favour?

Avatar of baddogno

No endgame expert here, SwiftyReflex, but are you sure K+N+B vs. K isn't a forced mate?  Seems I remember watching a tournament here where that situation came up and everyone was somewhat embarassed that one of the players didn't know how to put her opponent away. K+N+N can't be forced, but with the B I thought it could be.  I'm wrong a lot of course. Laughing

EDIT:  OOPS, a day late and a dollar short again.  Thanks Martin, for the confirmation.

Avatar of SwiftyReflex
Martin_Stahl wrote:
SwiftyReflex wrote:

...  Another point, you can't FORCE mate with only K+N+B vs K, ....

That is incorrect. That is a forced mate if you know how to do it. Something like 32 moves with perfect play from the worst initial configuration of the pieces.

     No, it isn't incorrect.  A king can always escape if it's K+N+B vs K,  you can mate with only 2 bishops, but not with one bishop and one knight, do an analyze board and try it yourself.  King can always escape

Avatar of Brian-E
baddogno schreef:

 Seems I remember watching a tournament here where that situation came up and everyone was somewhat embarassed that one of the players didn't know how to put her opponent away.

Yes, former Woman's world champion Anna Ushenina failed to win K+B+N vs K against Olga Girya in Geneva in 2013. Very embarrassing.

Avatar of Brian-E
SwiftyReflex schreef:

     No, it isn't incorrect.  A king can always escape if it's K+N+B vs K,  you can mate with only 2 bishops, and only 2 knights, but not with one of each, do an analyze board and try it yourself.  King can always escape

You need to look at the theory books again. The others are correct.

Avatar of nobodyreally

KNB vs. K = win

KBB vs. K = win

KNN vs. K = draw

KNN vs. Kp = win or draw, depending on where the pawn is.

Avatar of nobodyreally
CensoredReality wrote:
Yeah K + N + B is a forced mate but is a difficult pattern to remember and execute correctly without reaching the 50-move draw rule. Its so difficult that engines (at least the ones I have used) cant figure it out without endgame table bases. I know this because I wanted to practice this ending with chess.com's computer but it wasn't strong enough to play the correct moves to force mate (i was playing the losing side to study the movements).

 Are you saying the engine couldn't do it by brute force even though it's only 4 pieces? That's weird.

Avatar of SwiftyReflex
CensoredReality wrote:
Swifty, that is incorrect. I have watched very deep explanations on this ending and it is forced, just very difficult. Find information about it on YouTube or whatever if you don't believe me.

Then maybe it's the two knights that the king can always escape, it's been awhile since I tried it myself on a board, but I know there's 2 3 point pieces that the king can always find a square to escape

Avatar of K_Brown

KBN vs K is indeed forced mate. I remember it being called by some "the w method" because of the path the knight takes. It is actually considered an elementary checkmate as far as I know.

It only takes a simple Google search to prove this and your argument makes no sense to begin with.

Avatar of Brian-E
SwiftyReflex schreef:

Then maybe it's the two knights that the king can always escape, it's been awhile since I tried it myself on a board, but I know there's 2 3 point pieces that the king can always find a square to escape

Yes, you're undoubtedly thinking of K+N+N vs K which is, as you say, a draw provided that the defending king avoids going to the wrong square by the time the attacking side has forced it to the edge or corner.