Not quite, again. The "null-move observation" which is what this is based on says that any move must be better than passing (not moving at all). And it is such a powerful option (to give your opponent two moves in a row) that if you play a null-move, you can safely use a reduced-depth search to see if your opponent can punish your "pass". If he can't, your position is extremely strong and you can take a beta-cutoff option rather than doing a normal (and much more expensive) search.adieguez wrote:hi, you have mentioned this in the past too several times. But this is not very accurate. Nullmove is saying there must be a move better than pass and nothing else. The prunning is at the middle of the two movesbob wrote:
This misses the _main_ point of null-move. If a position is very good for me, so that even if I allow you to make two moves in a row and the position is _still_ good for me, then this position really is good for me and bad for you. Except when there is zugzwang. I'd be willing to play Kasparov if he would let me make two moves in a row a couple of times. Probably even once is enough. He would suddenly have to completely change the way he evaluates position since at any instant, I can choose to make two moves in a row. So he can leave no pieces open to a two-move threat, which is a real challenge. If I can attack any of your pieces in one move, I can win it if I so choose.
you mention. The position which is really good for you that you mention may not even exist. Explainning nullmove that way looks not correct, it's just something curious while thinking about it.. imo
Take any position you can think of and give your opponent two moves in a row and see what happens. If you are a queen up, it will be _very_ difficult for him to recover that even with two consecutive moves. You therefore give up quickly on such positions. If you are no material ahead, and your opponent still can't punish you with two consecutive moves, your positional advantage is so significant you can again bail out quickly.
I'm not quite sure what your argument is with this logic.. If null-move only says "there must be a move better than pass" then it would _always_ be true, except when zugzwang is an issue. That's not what it is about at all. It is about the advantage that two consecutive moves produces, and whether that is enough to bust the current advantage the null-move side has. 99% of null-move fail highs are in positions where one side is ahead in material and the other side has no way to recover it, even with two consecutive moves.