Stalemate in Shogi?

Unless I'm missing it, neither Wikipedia nor this site, http://www.ricoh.co.jp/SHOGI/rules/erules.html#DRA, discuss what happens if:

\* It is the players' turn and they have no legal move

I suppose this is fairly rare and would mean the player has no pieces in-hand to drop and none of their remaining on-board pieces can legally move.

In western Chess, such a situation is called "stalemate" and results in a draw. In Chinese Chess (Xiangqi) this is a loss. What's the rule for Shogi?

Thanks,

Alain

I have always assumed that since you have no legal move, you are forced to resign or lose on time. I doubt in would ever happen in practice though.

I wondered about this too when I was writing my mini Shogi bot, and I came to the conclusion that if you are put in stalemate, you lose. If you search for shogi stalemate on Google, you get a few pages that say this, e.g. the GNU Shogi manual: "Stalemate is a virtual impossibility, and is a loss for the stalematee."

Thanks JeanHerbert. I also assume the same thing, but assumptions can be wrong. There are situations where you are overwhelmingly ahead on pieces and allowing stalemate (if its ends in draw) is avoidable.

I'm experienced in both western Chess and Xiangqi and know that this particular rule means that in western Chess you have to be quite cautious in some positions if you're ahead (and look out for clever stalemates if you're behind). But in Xiangqi, this rule is often your saviour if you're ahead or your doom if you're behind. The stalemate = loss rule is absolutely crucial in allowing many classes of endgames to be winnable in Xiangqi.

@Tommah: Thanks. That website states it clearly. I've seen different rules printed on different websites (for example, in Lines of Action, if the last move results in both positions being connected, some websites say its a win for the mover, others say its a draw).

If a mathematical purist tries to define chess rules that make more sense, an interesting variation arises -- that stalemate should be a loss for the player who can't move, rather than a draw. The argument is that stalemate is the ultimate form of zugzwang, and a hopeless lone King should not get a magical "get out of jail free" card.

The implications of that rule change for chess are massive. For example, every K+P vs K endgame is now a trivial win (unless the lone King can capture the pawn, of course). K+B+N vs K becomes very easy, instead of a challenging bar bet (i've won many a beer that way:). And so on. The stalemate-is-a-loss rule would not only make more sense, it would greatly reduce the number of games ending in draws (thus reducing the threat of "draw death" in chess). The implications reach back to the middlegame and even to the opening. Of course, it couldn't really be called chess anymore -- the game is that different.

I helped decide the LoA simultaneous connection draw rule for use in the Computer Olympiad. In the end, we went with the most logical conclusion of the game, rather than the alleged wishes of Claude Soucie, the game's inventor. Since both sides accomplish the specified goal of the game at the exact same instant, a draw is the fairest result -- there is no need for an arbitrary choice of winner (i.e. the moving player wins). I found several example positions with Mona where the player with a large advantage would actually lose under that silly rule, which is why i was lobbying for a draw.

Anyone who sees the world in terms of clean logic has to be mystified and bewildered by many of the rules in popular board games. Especially ancient board games, like Go, that have suffered thousands of years of pollution from meddling humans... :-/

For example, "bent four in the corner is dead" -- ya, we really needed to have a special rule just for that, and we really want to _require_ professional 9-Dan players to serve as referees on top-level matches... Eesh. [But i digress... /rant:]

Old threat but still I want to add something controversal to Darse's post:

From a mathematical point of view stalemate in chess has to be a draw as else there would be absolutely no reason for a rule allowing promotion to bishop or rook. ;)

The analogous rule in shogi is the uchifuzumi-Rule that prohibits a mate delivered by a pawn-drop and thus makes unpromoted moves (where promotion is possible) useful also for Rook, Bishop and Pawn.

Stalemate in Shogi is a win. At least all engines known to me have it implemented like that.

Another interesting question to the mathematical purist would be what happens if a player has no available moves but not because the king would move into check (you could imagine the king being boxed in I think). In this case, I think a pass would be the appropriote action.

In LoA, if you have no moves, it is a forced pass. There are no optional passes (as in Go). I wouldn't say this is "purist" rule, just an arbitrary choice.

In combinatorial game theory, having no move is the end-of-game condition, and there are good mathematical reasons for that. All other games are "silly" in comparison. :-)

The question is still unanswered. What happens if a player with huge advantage manages to stalemate the opponent King?

How is it implemented in Golem?

I have a position in the 3x4 variant that I will soon find out (and I do hope I won't lose the game!)

If there is a stalemate, the player who can't make a moves loses.
I have no idea how it's implemented on littlegolem though :-)

Come on, there exists no concept of stalemate in shogi or the rules of shogi, though it exists in chess. As long as players engage in games with the mutual understanding that the goal of shogi is to checkmate the King, the state what you call "stalemate" will never practically happen in shogi thanks to the rule of allowing the reuse of captured pieces. If it happens, I believe the player who has no legal move will lose the game because shogi does not allow him to pass his next move and he has no option but to make an illegal move except for resigning the game.

I agree that in normal shogi it will certainly never happen, but as ypercube pointed out in 3x4 shogi it's not at all impossible for a player to loses his 3 pieces and be unable to make a move. It's even more relevant for a chess player who might see it as a way to force a draw, if he doesn't know the rules. So I don't see any harm in discussing it here :-)

The original rules of Dobutsu shogi Kitao Madoka has set have no concept of stalemate at all. It does not even have an concept of "illegal move of exposing or keeping your King in check" either. In Dobutsu Shogi, the move of exposing or keeping your King in check is completely legal. The game is not to end with checkmate but to end with the capture of the King, which is different from the rules of conventional shogi. The game has been intended for small children from the age of three and this is why the game is set to end with the capture of the King since the concept of checkmate is a bit too hard for them. If the 3x4 shogi of Little Golem follows the spirit of the original Dobutsu Shogi Rules, there should not be a stalemate for it.

I see.

I don't think littlegolem really applies these rules for dobutsu shogi though.

By the way, you say that there is no concept of putting the king in check in Dobutsu. Am I wrong to think that you can win the game if you reach the last rank, *without* putting your king in check?

LittleGolem has implemented Shogi (at least the 3x4 variant) with automatic detection of checkmate and as evidence below shows automatic detection of stalemate:

[game;id:1502999;title:Stalemate]

> Am I wrong to think that you can win the game if you reach the last rank, without putting your king in check?

No, you are right. the one who lets the king(or Lion) reach the last rank without putting your King in check shall win in Dobutsu Shogi.

> LittleGolem has implemented Shogi (at least the 3x4 variant) with automatic detection of checkmate and as evidence below shows automatic detection of stalemate:

Maybe that is why the game is called "3x4 Shogi" in LG. I think it is fairly a different game from Dobutsu Shogi, at least for me.

@takodori: In any case, computer implementations of any game, whether they are standalone or web-based, need complete rules. Because users/players will do the dumbest things, sooner or later.

I real life, these situations can be afforded to be resolved by an arbiter or a knowledgeable player.

Because users/players will do the dumbest things, sooner or later.

But I do miss HOT_SYNC_KAREN...