Btw, stalemate as win could make the rules simpler by allowing the king to be taken. The games would last one move longer and no checkmate rule would be required. Zugzwang would apply to stalemate positions, so they would be lost.
After all, chess is a model of war. Hitler couldn't demand a draw from his bunker, right?
Not necessarily. What would you do about this?
Stalemate just means that u bamboozled ur opponent; you didn't win, but you (somewhat) outplayed your opponent (you didn't lose). Is just draw.
Go go
Btw, stalemate as win could make the rules simpler by allowing the king to be taken. The games would last one move longer and no checkmate rule would be required. Zugzwang would apply to stalemate positions, so they would be lost.
After all, chess is a model of war. Hitler couldn't demand a draw from his bunker, right?
Not necessarily. What would you do about this?
That would be hilarious
One of the main implications: ¡it would be boring!
Many endgames with material advantage would become trivial wins.
As it is now, there are many games which, in practice still involve a lot of fighting: one side fights to win, the other one to draw, and it is exciting and interesting. Say goodbye to many of them if stalemates become wins.
One thing I like about chess (as compared to Shogi or Xiangqi): essentially, you only win by checkmate (*). If you have the advantage and can not deliver checkmate, either because it is impossible or because you do not know how, then it is a draw.
(*) Ok, your opponent can resign; can run out of time; or can be disqualified. But those are not parts of the formal mechanics of the game.
In Shogi, you can win by having more material (not something frequent, but possible). In Xiangqi, if you can win if the opponent insists on delivering a perpetual check. I do not like either of those: winning without being able to force a checkmate. ¡I prefer chess! In both of them, you can win by stalemate.
@ThrillerFan
You clearly didn't read OP's post #13.
Nobody reads an entire thread. I responded to the original post.
Either way, doesn't matter. The rules are the rules. When it comes to variants, it needs to be a variant. Not just some stupid single rule change. 3-check chess. That's a variation. Suicide Chess. That's a variant.
Stalemate = win or Stalemate = Loss and everything else the same is not a variant.
So again I will say, those looking to remove the Stalemate rule simply don't know how to play chess and had some horrible experience of a moronic stalemate, or else are advocates of those morons.
um 3 check is "chess but you also win from checking thrice" so the same thing should apply
Many thanks to tygxc, FaultyBagnose, and especially MARattigan for the interesting responses to this question. MARattigan, that first two knights position is funny, and you're right, of course, that one knight is a goner. And even more amusing is the last one, where White lets the h2 knight get captured, and then Black is forced to self-stalemate in the corner.
But ultimately I'm more interested in what the effect would be on the most common endgames because that would determine whether such a rule change would make a serious difference to how we think about positions, piece values, etc, or whether it would just make an incidental difference here and there. (Which makes me less worried about positions like the one in post 22, not only because it's unlikely to occur in a game, but also because black's "stalemate-loss" could only occur because they made a terrible error in the moves before.)
If it turns out that the difference is only an incidental one, then I'd actually agree with Thrillerfan that such a variant would be of little value (though it does seem peculiar that he can know that with such certainty without, seemingly, reading even so far as the first sentence of the thread -- but, then again, Thrillerfan is not someone often cursed with self doubt...)
And maybe it is. Going back to rook endings (this was a good occasion for a refresher), I'm reminded that K&H and side checks don't rely on scenarios where the pawn ending is a draw, and that they'd save the same endings a Philidor position would (albeit with more chances to make a mistake). Also, I'm realizing I'd forgotten that there's even one sort of KP v K ending that would remain a draw (but only if the pawn is on the 6th rank).
So I guess the answer is that the status of a few endgames would change but most would remain the same as normal chess?
#28
Indeed the conclusion of Kramnik "it seems to almost always be possible to defend without relying on stalemate as a drawing resource" seems right. Also his conclusion that it does not change either opening or middle game play is noteworthy. This also means that the scoring system proposed by Lasker 10 points for checkmate, less for stalemate does not help towards the goal of avoiding draws. Neither Kramnik nor Lasker are/were idiots.
You can trust Kramnik on this.
#28 - I know you like to put people like me under the bus, but get real here. It would alter a lot of Rook Endings!
W - Ka1, Rb5, Ph4
B - Kg4, Rb2, Pb4, Pc3, Ph5
White to move. It was Black's job not to get into this predicament in the first place. It is his own problem that this draws (1.Rg5+ Kxh4 2.Rxh5+!! With the eternal Rook draw.)
Why should Black be credited with a win due to stupidity? Either allow White a move, or make sure he is in check. If he has no move, and he is not in check, it's a draw! Suck it up buttercup!
#28 - I know you like to put people like me under the bus, but get real here. It would alter a lot of Rook Endings!
W - Ka1, Rb5, Ph4
B - Kg4, Rb2, Pb4, Pc3, Ph5
White to move. It was Black's job not to get into this predicament in the first place. It is his own problem that this draws (1.Rg5+ Kxh4 2.Rxh5+!! With the eternal Rook draw.)
Why should Black be credited with a win due to stupidity? Either allow White a move, or make sure he is in check. If he has no move, and he is not in check, it's a draw! Suck it up buttercup!
Thanks -- I'll look at this when I can get to a chessboard.
Not sure why I'm the one throwing people under the bus. I'm not the one who walked up to the thread, didn't read it, trashed it, had it pointed out he hadn't read it, and then doubled down on trashing it. But this looks like a more helpful and interesting response. Thanks.
So I guess the answer is that the status of a few endgames would change but most would remain the same as normal chess?
Nope. The big majority of endgames would change completely.
So I guess the answer is that the status of a few endgames would change but most would remain the same as normal chess?
Nope. The big majority of endgames would change completely.
Thanks. Do you mean that the evaluation would change or just that the way they're drawn would change (with, perhaps more losses being the practical result, if the drawing technique is more difficult)?
#28 - I know you like to put people like me under the bus, but get real here. It would alter a lot of Rook Endings!
W - Ka1, Rb5, Ph4
B - Kg4, Rb2, Pb4, Pc3, Ph5
White to move. It was Black's job not to get into this predicament in the first place. It is his own problem that this draws (1.Rg5+ Kxh4 2.Rxh5+!! With the eternal Rook draw.)
Why should Black be credited with a win due to stupidity? Either allow White a move, or make sure he is in check. If he has no move, and he is not in check, it's a draw! Suck it up buttercup!
Here's that diagram:
I love eternal rook draws and have a memorable save to my credit as a result of one. So I confess to a strong emotional attachment to the sort of trickery that stalemate can allow. Like I've said, I'm not advocating for a rule change -- I'm really posing this as a question to try to understand the dynamics that endgames are premised on. (The question popped into my mind as I was trying to explain in another thread how Zugzwang isn't just something that sometimes happens to you in an endgame, but rather, the underlying idea toward wich many endgames are headed and around which they completely revolve.)
If we switch to the "should-it-exist-as-a-variant" quesiton (which I'm sort of resisting as a question and only secondarily interested in), I suppose one could argue that (1) by the same logic, if Black played poorly enough to get into this endgame, why do they deserve a chance to draw it? (I certainly can't say I'm proud of anything but the few couple of moves in the last game I saved with a stalemater trick); and (2) what's lost in nifty saves is gained in having an extra pawn count for more, meaning that people actually get credit for games where they get an advantage that, according to the standard rules, would only be enough for a draw. Me, I'm with you in my sympathies here, but I can't, in good faith, get so certain that the standard rules are the only right way to do things.
... MARattigan, that first two knights position is funny, and you're right, of course, that one knight is a goner.
Actually neither knight is necessarily a goner, it's just that keeping them is a full time occupation for White and he can't find time to bring his king in closer. I've added some analysis to the diagram in the original post.
With the king on e.g. f3 the story is different.
It's fairly obvious that Black can't allow the a8 knight to escape the corner if the other knight is safe. I've played out one variation against SF. (It's the first time I've tried the exercise, so the sequence is probably hugely incompetent, especially considering nobody has told SF what he's meant to be doing.)
I've assumed the dead position rule is also altered to include stalemates.
And even more amusing is the last one, where White lets the h2 knight get captured, and then Black is forced to self-stalemate in the corner.
Of course Black doesn't have to capture the knight and his stalemate may need a little coaxing. But if he doesn't capture the knight he falls into a loss anyway. I've added some analysis to that diagram too, generally stopping at the point where a win is possible under current rules.
But ultimately I'm more interested in what the effect would be on the most common endgames because that would determine whether such a rule change would make a serious difference to how we think about positions, piece values, etc, or whether it would just make an incidental difference here and there. (Which makes me less worried about positions like the one in post 22, not only because it's unlikely to occur in a game, but also because black's "stalemate-loss" could only occur because they made a terrible error in the moves before.)
I doubt if the position in #22 has occurred very often in practical games, but I do recall looking at a position in Averbakh, which I think was based on a practical game, where a variation in which one side became gridlocked was germane. I couldn't find it when I looked, but it has to be a very rare occurrence anyway.
As far as practical endgames go, my feeling is that most of the basic endgames I've looked at would become easier for the winning side if stalemate were counted as a win, but I haven't seriously thought about how many extra wins would be generated in general.
My grandson has just learned the moves, so I set him the task of mating me with a queen which initially resulted in several stalemates. I read a post from someone who teaches children chess and this is apparently the general result. I also remember my father setting me the task of mating with two bishops and I think I perpetrated a half dozen stalemates from the off.
Of course the distinction between endgames and other positions is essentially an arbitrary one based on limited look ahead ability by anything that plays without access to appropriate tablebases, so I would expect what applies to basic endgames to apply in general, but in practice to be less important because beginners, grand masters and Stockfishes without tablebase access all start to lose the plot in many positions with only five men on the board.
If it turns out that the difference is only an incidental one, then I'd actually agree with Thrillerfan that such a variant would be of little value (though it does seem peculiar that he can know that with such certainty without, seemingly, reading even so far as the first sentence of the thread -- but, then again, Thrillerfan is not someone often cursed with self doubt...)
And maybe it is. Going back to rook endings (this was a good occasion for a refresher), I'm reminded that K&H and side checks don't rely on scenarios where the pawn ending is a draw, and that they'd save the same endings a Philidor position would (albeit with more chances to make a mistake). Also, I'm realizing I'd forgotten that there's even one sort of KP v K ending that would remain a draw (but only if the pawn is on the 6th rank).
So I guess the answer is that the status of a few endgames would change but most would remain the same as normal chess?
I'd expect just the fact that KPK is radically changed to have major effects on any positions with pawns and these will swamp the pawnless positions as the number of pieces increases.
#35
See figure 2 (b) of the cited paper:
Stalemate = draw: 97.9% draw
Stalemate = win: 97.1% draw
It does not change the evaluation that much.
" it seems to almost always be possible to defend without relying on stalemate as a drawing resource" - Kramnik
Thanks, @MARattigan. Really interesting analysis of the knight endgames, in which I'd missed some ideas.
I think the penultimate paragraph in my previous post is relevant here.
The paper states earlier on
The absence of baselines makes it hard to formally assess the strength of each model, which is why it was important to couple the quantitative analysis and metrics observed at training and test time with a qualitative assessment in collaboration with Vladimir Kramnik, a renowned chess grandmaster and former world chess champion.
which I think translates as
We have no idea how AlphaZero's limited look ahead play compares with objectively accurate play but we rather hoped Kramnik might.
So how good is AlphaZero in absolute terms? I don't know, but I once downloaded Leela to see if it played basic endgames any better than the engines I had. The answer was a very definite no. And the reason I was looking for a better engine was because the engines I have can't play some of the endgames I'm interested in without an EGTB.
So I think your question needs to be more specific about whether we're talking about evaluations in terms of practical play or objectively accurate play.
Also which kind of chess. FIDE defines two main flavours of game, basic rules and competition rules. The presence of the 50 move rule in the latter has a very distinct effect on the objective evaluation of many basic endgames - you need only look at the "frustrated wins" (=draws) values on the syzygy-tables.info site.
Thanks to all who have taken the question seriously. When I have time tonight, I hope to follow up on some of your good points.
I answered your original question.
I said clearly that using stalemate as a win being a "Variant" is stupid.
Well, I'm glad that's settled!
Funny, though, I remember my original question being "How extesively would endgame theory would have to be rewritten -- and, by extension, how much would middlegame play and even opening theory be affected?" I've seen several people offer thoughtful answers to that one, but somehow can't find yours. Oh well.