In this position white wins no matter where the kings are!
Not quite true.
Obviously excluding these positions, the same way we wouldn't exactly consider this Queen vs 2 Knights position:
The tablebase is obviously going to take into account instant stalemate or wild plot twist losses from the "winning side", you don't have to give these examples in every single discussion lol
Good thing chess has a 50 move rule.
An excellent rule.
Like touch move in live play is also excellent.
And good thing that both rules are widely accepted and will likely continue to be the practice.
I understand touchmove, you should touch a piece until you are 100% decided om your move, but the 50 move rule is completely arbitrary, and decided on lazily based on its divisibility into our number system. At least 100 moves (or 101, might as well make it based on a prime number) would give some leeway to make a mistake in an endgame, but given that 50 is less than 1/10th of the moves required to win the longest discovered tablebase mates, it is illogical.
The 50 move rule is not 'arbitrary'.
It is immensely practical and was found to be best after a long process of experience over the centuries.
EE are you going to try to claim its wrong because you say so?
Oh wait - you're already doing that.
Not a 'big red telephone' situation.
More like 'center of the universe' position ...
Nope they just said "50 moves = 100 ply = round numbers good enough for me!". Even the Rook + Bishop vs Rook Endgame can take in the 60-65 move range to win. 2 bishops vs knight can take 78 moves with perfect play. So it's not even allowing the wins with perfect play let alone with a margin of error. Even realistic endgames such as 2 rooks + pawn vs queen can take 200 moves to mate where the first pawn move isn't until move 119. 100 moves would at least take into account the realistic endgames but nope lol
Nope they just said "50 moves = 100 ply = round numbers good enough for me!". Even the Rook + Bishop vs Rook Endgame can take in the 60-65 move range to win. 2 bishops vs knight can take 78 moves with perfect play. So it's not even allowing the wins with perfect play let alone with a margin of error. Even realistic endgames such as 2 rooks + pawn vs queen can take 200 moves to mate where the first pawn move isn't until move 119. 100 moves would at least take into account the realistic endgames but nope lol
Your arguments don't hold up EE.
Nor does your 'nope' either.
You can say 'water is wet' too.
The reasons for the rule go right over your head.
Like so many things do.
My guess would be Ruy Lopez decided that KBNK represented the most you'd ever need to go without progressing to a different phase, estimated it around the correct value of 33, then added 50% for breathing space.
My guess would be Ruy Lopez decided that KBNK represented the most you'd ever need to go without progressing to a different phase, estimated it around the correct value of 33, then added 50% for breathing space.
And after that - a lot of testing and experience told the chess community of western europe over a period of a hundred years that 50 was the most practical number.
100 moves - 200 ply is just too many.
They've got to make it 500 moves because of some weird obscure ending?
------------------------------
There's another 'guess' though.
They knew that normally - any run of 20 moves (40 ply) anywhere within a game with no captures nor pawn moves - doesn't happen a big percentage of the time.
But they know it still happens quite a bit.
20 would just be too low.
50 makes sense because its very rare you'd go that many moves (100 ply) with no captures/pawn moves while 'getting somewhere'
Its like the lowest round number where that would be so unlikely.
A lot of players would even agree to a draw well before that.
'50' to take care of unscrupulous types.
does this horsey fortress work ?
...or the love triangle fortress ? (where white plays for a win if black plays like a homer)
does this horsey fortress work ?
...or the love triangle fortress ? (where white plays for a win if black plays like a homer)
Is there any real point in posting tablebase-solved positions in a thread where people all know about tablebases and how to use them?
My guess would be Ruy Lopez decided that KBNK represented the most you'd ever need to go without progressing to a different phase, estimated it around the correct value of 33, then added 50% for breathing space.
I highly doubt they decided 50 moves based on taking the perfect play maximum for one endgame and multiplying it 150%. Probably something to do with the average time that amount of moves takes in common tournament scheduling/time controls...etc.
The 50 moves weren't finally accepted by 'formula'.
Its about practicality.
So is threefold repetition.
Repetition is a draw on the board, not a regulatory rule. There's no logical result for repetition other then a draw. 50 moves has nothing to do with the game itself only a regulatory add on to keep tournaments moving.
Repetition is a draw on the board, not a regulatory rule. There's no logical result for repetition other then a draw. 50 moves has nothing to do with the game itself only a regulatory add on to keep tournaments moving.
You're wrong again EE. On multiple counts.
Repetition is a rule. And rules are 'regulatory'.
You didn't know that?
50 moves has everything to do with the game.
You just don't like it.
You could say you don't like the rule or the number -
without making false assertions.
Apparently you do so to feel more comfortable with your dislikes.
Its the same with your denial of climate science.
Where you don't even consider the realities of high tide recordings all over the world the last 100 years.
does this horsey fortress work ?
...or the love triangle fortress ? (where white plays for a win if black plays like a homer)
Is there any real point in posting tablebase-solved positions in a thread where people all know about tablebases and how to use them?
Regarding R N B versus queen - that looks interesting.
I would guess that two rooks versus queen with no pawns on board would be drawn because the queen would keep checking.
But maybe with three pieces their King could effectively hide from checks.
Tried to look it up on the net but will try the analysis feature here.
Am thinking the queen would try to pin the rook.
Absolutely no point. White just plays Bd6+ and takes the queen.
I wasn't referring to that particular diagram.
But that Bd6ch there depends on it being white's move.
Anyway using https://www.chess.com/analysis?tab=analysis-----------------------
I created R B N versus Queen position and then clicked on its 'lines' and the positions continued to maintain as under 0.4 advantage for black with his 3 pieces.
In other words its apparently a draw.
2 rooks was worse - even smaller advantage for black.
Absolutely no point. White just plays Bd6+ and takes the queen.
I wasn't referring to that particular diagram.
But that Bd6ch there depends on it being white's move.
FEN says it is.
No point in trying it in analysis anyway - you can't believe the answers. SF would probably give you 100% or almost 100% draws against itself (and definitely against Syzygy) from the longest winning positions. It can't play long KQKNN wins either.
This is a winning (ply count 0) position. Try it in analysis. (It gives me 0.00 by the time it's produced 42 moves.)
The 50 move rule should be abolished due to those complex endgames that require hundreds of moves to win. I say replace it with a simple time limit, if a tournament game is seemingly going on endlessly without any progress, over 100 moves without captures/pawn moves/repetition..etc, and the time controls increment are preventing anyone from flagging, just declare the game drawn if they are still going at it for a half hour. Have FIDE implement a better version of USCFs "insufficient losing chances" rule. Then tablebases wouldn't even need that "Depth to Zeroing Move" nonsense, and could focus on pure calculation and game tree generation.