|
|
|
Topic: Should "stalemate" be considered a draw?
| |
|
|
Author
| Message |
|---|
|
|
Well "Evjen", I agree with you (theoretically). As a comp chess AI programmer I can count stalemate any way I want. In fact, when training AIs I don't have a concept of mate, checkmate, stalemate, or not moving into check. To keep the programming easy I only enforce the rule that whoever loses their king loses.
Of course, when I want to compete these AIs against the outsite world I need to follow the agreed upon current rules of the game.
|
|
|
its a draw bcoz stalemate is a big mistake of player on winning edge but dont know how to win. or perhaps not looking at the board.
when pawn promotion against lone king, if you afraid for stalemate - promote it with rook.
|
|
|
"It's a draw because stalemate is a big mistake of the player on the winning edge but doesn't know how to win or perhaps is not looking at the board."
The stalemater is still in a better position than the stalemated opponent.
Consider the following analogy:
Their are two kingdoms fighting one another for the right to rule a particular area.
They agree that as soon as one of the kings is captured by the enemy, the remaining men will all be ruled by the uncaptured ruler.
They will do whatever is necessary to protect their kings even if it means losing every last soldier.
Checkmate is analogous to soldiers surrounding the enemy's king to where he cannot avoid capture.
A True Draw is analogous to the opposing armies chasing each other around in circles for a while and finally giving up and signing a peace treaty.
Stalemate, however is different than either of the above situations. It is analogous to the following:
The stalemated side's soldiers are all captured or trapped from moving.
The only one left on that side who can still do anything is the king.
The king is in a small fort surrounded on all sides from the enemy.
If he leaves the fort, he will be captured, so he waves a white flag and surrenders.
I agree that stalemate should not be counted as a full win, since the stalematers often have made a mistake when they could have checkmated their opponent.
However, they are still clearly in a superior position.
There are also some cases where stalemate is possible, but not checkmate, such as king and bishop or king and knight verses king.
In the first case, I believe it is always possible to force a stalemate, but is not always possible with a king and knight.
Why not adopt rules that would reduce the number of draws?
http://www.michess.org/webzine_199907/okeefe.shtml
|
|
|
"Well "Evjen", I agree with you (theoretically)."
I am glad someone finally has openly agreed with me.
I suspect there may be others as well, but are afraid to speak out for fear of being ridiculed.
"Of course, when I want to compete these AIs against the outside world I need to follow the agreed upon current rules of the game."
You could include an option in the program preferences so the users could chose which stalemate rules to use
and also explain arguments for the alternative view in the instructions or rule section of your program.
I am pretty sure that the present stalemate rule will take quite some time to change, if it changes at all,
but things like that will increase the likelihood of the majority adopting more logical stalemate rules sooner.
|
|
|
| I suspect there may be others as well, but are afraid to speak out for fear of being ridiculed. |
1) I don't suspect so;
2) Could you *really* blame them?
|
| Previous 1 ... 4 5 6 7 8 9 10 11 Next |
|
|
|
