hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I don't think so. A "pass" means you can win ANY piece I have pretty much, by making two consecutive moves. In lots of cases, you can mate me with two consecutive moves. If I try to find the worst move, sometimes even that is better than doing nothing...
When I first read about null-move in Beal's paper, one of my chess friends and I sat down to play some games, with a new rule. At one point in the game, one of us could say "two moves", and then make two consecutive moves. Only once in a game, and only one of us could do that in a game (pre-chosen before we started). The "two move" side won every last game. The other player has to be beyond careful. you can't move your queen to any square where it can be attacked, because the opponent will say "two moves", take it and then move the piece back. Once you do that a bit, it becomes clear about "the null-move observation". (we used to play lots of other odd games including "may I" but this was the first time we played games to become acquainted with a computer chess idea. )
Here's an example, for fun. 1. e4 e5 2. Bc4. If you play a natural-enough looking Bc5 you instantly lose. "two moves". Qf3 Qxf7#" Hard to get to thinking like that...
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I don't think so. A "pass" means you can win ANY piece I have pretty much, by making two consecutive moves. In lots of cases, you can mate me with two consecutive moves. If I try to find the worst move, sometimes even that is better than doing nothing...
When I first read about null-move in Beal's paper, one of my chess friends and I sat down to play some games, with a new rule. At one point in the game, one of us could say "two moves", and then make two consecutive moves. Only once in a game, and only one of us could do that in a game (pre-chosen before we started). The "two move" side won every last game. The other player has to be beyond careful. you can't move your queen to any square where it can be attacked, because the opponent will say "two moves", take it and then move the piece back. Once you do that a bit, it becomes clear about "the null-move observation". (we used to play lots of other odd games including "may I" but this was the first time we played games to become acquainted with a computer chess idea. )
Here's an example, for fun. 1. e4 e5 2. Bc4. If you play a natural-enough looking Bc5 you instantly lose. "two moves". Qf3 Qxf7#" Hard to get to thinking like that...
Here is an example, d4 d5, now give me your 2 moves.
If you play Bf4, and black still passes what white has got?, but if black will not pass and instead play the bad Qd6 - giving it to bishop at f4, then the pass is not the worst.
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I do not think that it is a good idea because the worst move is too weak.
The idea of the null move is to prune lines and if you try the worst move then usually this move is not going to produce a cutoff and you spent nodes for nothing.
If you want to replace the null move by another move then it is better to try some unconnected move that has a similiar effect
for example d4 d5 a3 when a2 or a3 is not connected to the black pieces(no black piece threats a2 or a3)
[quote="bob"] All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
Bob, how much have you experimented with Multi-Cut, either as a replacement for null move or in addition to it? We concluded that it was slightly inferior to null move as a replacement, and not helpful as an addition. But perhaps in pawn endings where null move doesn't work it would be viable, or perhaps even in endings with one minor piece per side. If HGM is correct that Zugzwang is not the main reason null move fails in pawn endings, then this won't help, but if Zugzwang is the main reason then multi-cut should work fine in pawn endings. Ever tried this? What do you think?
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I do not think that it is a good idea because the worst move is too weak.
The idea of the null move is to prune lines and if you try the worst move then usually this move is not going to produce a cutoff and you spent nodes for nothing.
If you want to replace the null move by another move then it is better to try some unconnected move that has a similiar effect
for example d4 d5 a3 when a2 or a3 is not connected to the black pieces(no black piece threats a2 or a3)
I do not think that it is a good idea because the worst move is too weak.
I can agree with this, I just want to point out that null move is not the worst move. But of course using the idea without some conditions and tunning applied is not recommended .
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I don't think so. A "pass" means you can win ANY piece I have pretty much, by making two consecutive moves. In lots of cases, you can mate me with two consecutive moves. If I try to find the worst move, sometimes even that is better than doing nothing...
When I first read about null-move in Beal's paper, one of my chess friends and I sat down to play some games, with a new rule. At one point in the game, one of us could say "two moves", and then make two consecutive moves. Only once in a game, and only one of us could do that in a game (pre-chosen before we started). The "two move" side won every last game. The other player has to be beyond careful. you can't move your queen to any square where it can be attacked, because the opponent will say "two moves", take it and then move the piece back. Once you do that a bit, it becomes clear about "the null-move observation". (we used to play lots of other odd games including "may I" but this was the first time we played games to become acquainted with a computer chess idea. )
Here's an example, for fun. 1. e4 e5 2. Bc4. If you play a natural-enough looking Bc5 you instantly lose. "two moves". Qf3 Qxf7#" Hard to get to thinking like that...
Here is an example, d4 d5, now give me your 2 moves.
If you play Bf4, and black still passes what white has got?, but if black will not pass and instead play the bad Qd6 - giving it to bishop at f4, then the pass is not the worst.
Pick a position that is more than 1 move into the game. You will see how hard this gets to defend if your opponent can play two consecutive moves just one time... for example, even after d4 d5 e3 e6 I might go for Bb5 and Bxe8 and you have no king.
You can never let me check you as I will play the second consecutive move, rip your king and that's that...
bob wrote: All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
Bob, how much have you experimented with Multi-Cut, either as a replacement for null move or in addition to it? We concluded that it was slightly inferior to null move as a replacement, and not helpful as an addition. But perhaps in pawn endings where null move doesn't work it would be viable, or perhaps even in endings with one minor piece per side. If HGM is correct that Zugzwang is not the main reason null move fails in pawn endings, then this won't help, but if Zugzwang is the main reason then multi-cut should work fine in pawn endings. Ever tried this? What do you think?
Larry
I didn't have much luck with it. I honestly do not recall whether or not I used the same "disable" rules (disable if no friendly pieces)... Right now I am still experimenting with ways to better limit reductions. Nothing significant so far. Found several things that works, but they work no better than what I am currently doing, and I favor simplicity unless I can improve on it...
Seems intuitive to me that this can be improved. Very elusive to figure out how, exactly. I don't want a bunch of static crap to make the choice, that leads to bloated code and the potential for long-lasting bugs. I'd prefer a dynamic approach that uses information gained during the search, somewhat like history counters are used currently by many... Seems like several like the idea I started in Crafty, where if a move fails high, but is not the first move, then all the other moves tried first have their history counters penalized, in addition to giving a bonus to the actual fail-high move. But something tells me there is something better...
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I don't think so. A "pass" means you can win ANY piece I have pretty much, by making two consecutive moves. In lots of cases, you can mate me with two consecutive moves. If I try to find the worst move, sometimes even that is better than doing nothing...
When I first read about null-move in Beal's paper, one of my chess friends and I sat down to play some games, with a new rule. At one point in the game, one of us could say "two moves", and then make two consecutive moves. Only once in a game, and only one of us could do that in a game (pre-chosen before we started). The "two move" side won every last game. The other player has to be beyond careful. you can't move your queen to any square where it can be attacked, because the opponent will say "two moves", take it and then move the piece back. Once you do that a bit, it becomes clear about "the null-move observation". (we used to play lots of other odd games including "may I" but this was the first time we played games to become acquainted with a computer chess idea. )
Here's an example, for fun. 1. e4 e5 2. Bc4. If you play a natural-enough looking Bc5 you instantly lose. "two moves". Qf3 Qxf7#" Hard to get to thinking like that...
Here is an example, d4 d5, now give me your 2 moves.
If you play Bf4, and black still passes what white has got?, but if black will not pass and instead play the bad Qd6 - giving it to bishop at f4, then the pass is not the worst.
Pick a position that is more than 1 move into the game. You will see how hard this gets to defend if your opponent can play two consecutive moves just one time... for example, even after d4 d5 e3 e6 I might go for Bb5 and Bxe8 and you have no king.
You can never let me check you as I will play the second consecutive move, rip your king and that's that...
For chess programs it is not relevant because you never do null move when the side to move is in check.
letting the opponent to have 2 moves one time in a game is also clearly worse than making a null move because the opponent can choose the first move that make null move worst case and not only he can choose it he can choose the time he does it.
If you change the rules of chess to allow a null move then
usually null move is not the best move but it is also not the worst move in most cases.
hgm wrote:I don't think zugzwang is the major issue that makes null-move fail in pawn endings. So this proposal could very well be a 'solution' to a non-existing problem.
Null-move doesn't perform in pawn edigs, because you cannot afford reductions in such endings. "When he can't hurt me when I pass my turn, he will likely not be able to hurtme when I make my best move" is of course a non-starter for explaining null-move pruning, because it doesn't describe what the latter is doing. More accurate would be: "When he can hurt me when I pass my turn, but I turn a blind eye to it (by reducing depth!), he won't be able to hurt me when I make my best move". And in Pawn endings that is usually not true.
The reason is that in Pawn endings it is usually not possible to find forcing moves that can push the threat over the horizon. (Like attacking his Queen, so that he has to withdra itfirst, or fail low, or to capture something, which he must recapture or fail low despite his threat.) The only way to be reasonably confident that you won't be slaughtered after your best move is to play an unreduced null move. Which kind of spoils most of the gains standard ull-move pruningbrings you.
How important zugzwags are for NMP can easily be tested on pawn endings in a chess variant that only differs from standard chess in that ull-moving is legal. Zugzwag is a non-existent concept in that variant. My prediction is that standard null-move pruning will also be counter-productive there.
Note that Crafty was reported to not have any problems winning KRK with null-move switched on. While we all know zugzwang is completely essential for wining that end-game.
You are correct about KRK, which means I will have to think about that. As far as zugzwang in KP endings, you may wall be right there.. but that will take a ton of analysis that I am not sure is worth the effort. All I know, for sure, is that the deeper into the endgame I extend "null-move enabled" the higher the Elo, until I reach just king and pawns, where the Elo drops -6 from turning null-move on all the time, as opposed to turning it off when the last piece disappears.
I still think that the "null-move observation" is a critical component of this, that "if I let my opponent make two moves in a row and I am still >= beta, my position is good enough to stop searching it". Zugzwang is the thing that makes that fail, as doing "nothing" is sometimes better than doing "anything" which is a real problem. Certainly reduced depth is an issue. But I think it is less of an issue so long as zugzwang doesn't bite...
Bob I don't know if you have tried this, instead of doing a null move, try to find a worst move (say a move that gives material) and make it. Now I think this is much worst than passing.
I don't think so. A "pass" means you can win ANY piece I have pretty much, by making two consecutive moves. In lots of cases, you can mate me with two consecutive moves. If I try to find the worst move, sometimes even that is better than doing nothing...
When I first read about null-move in Beal's paper, one of my chess friends and I sat down to play some games, with a new rule. At one point in the game, one of us could say "two moves", and then make two consecutive moves. Only once in a game, and only one of us could do that in a game (pre-chosen before we started). The "two move" side won every last game. The other player has to be beyond careful. you can't move your queen to any square where it can be attacked, because the opponent will say "two moves", take it and then move the piece back. Once you do that a bit, it becomes clear about "the null-move observation". (we used to play lots of other odd games including "may I" but this was the first time we played games to become acquainted with a computer chess idea. )
Here's an example, for fun. 1. e4 e5 2. Bc4. If you play a natural-enough looking Bc5 you instantly lose. "two moves". Qf3 Qxf7#" Hard to get to thinking like that...
Here is an example, d4 d5, now give me your 2 moves.
If you play Bf4, and black still passes what white has got?, but if black will not pass and instead play the bad Qd6 - giving it to bishop at f4, then the pass is not the worst.
Pick a position that is more than 1 move into the game. You will see how hard this gets to defend if your opponent can play two consecutive moves just one time... for example, even after d4 d5 e3 e6 I might go for Bb5 and Bxe8 and you have no king.
You can never let me check you as I will play the second consecutive move, rip your king and that's that...
For chess programs it is not relevant because you never do null move when the side to move is in check.
letting the opponent to have 2 moves one time in a game is also clearly worse than making a null move because the opponent can choose the first move that make null move worst case and not only he can choose it he can choose the time he does it.
If you change the rules of chess to allow a null move then
usually null move is not the best move but it is also not the worst move in most cases.
I was talking about the "2-moves-in-a-row" games my friend and I played when we were trying to understand null-move. And while a null-move might not be worst, it is REALLY BAD in most cases. The main benefit of null move is to detect positions where you are hopelessly behind in material, and you can spend less effort searching in that part of the tree...
lkaufman wrote:So in both cases it is much safer than null move. The only question in my mind is whether it will save enough time to be worthwhile. I don't expect a big gain from this, maybe just one elo or so.
There are positions where null-move makes impossible to find the right solution regardless of spent time. LMR and "legal null move" does not suffer from it, so they are better at least in analysis mode.
You are comparing your algorithm to traditional null move. But most programs don't use null move in pawn endings. The proper comparison for pawn endings is between your algorithm and no null move. I think your idea should bring a noticeable speedup in pawn endings, at almost no cost. It might be worth 50 elo in pawn endings. Unfortunately that translates to only one or two elo in actual play. Similarly in analysis of pawn endings it may bring a noticeable benefit, but it is rare to encounter a pawn ending which a computer in analysis mode can't solve correctly.
I still believe that because of the repetition problem it is going to hurt, not help...
Hi Bob,
I'm jumping into this discussion late and have not read anything that comes after, but I have already tried it in Komodo based on Larry's suggestion. I don't really unmake the move, I reverse it and only if it's legal. So I cannot reverse pawn moves since it is not legal to move them backwards. I just don't prune these moves - admittedly in king and pawn endings that may nullify some of the benefit.
The repetition is no problem, I can't understand why you see that as a hurdle with your programming skill. In my case I just call the reversed move an irreversible move so that I do not check it for repetition or any moves prior to it. That is pretty simple, right?
Since you are reversing a move, it's like you never made it, so it's generally going to be similar to a null move in power. There may be bizarre exceptions but in general this should not normally be a useful move.
Having said all of that, the idea did not pan out in Komodo.
)