Author: Miguel A. Ballicora
Date: 22:35:53 02/28/01
Go up one level in this thread
On February 28, 2001 at 16:52:05, Robert Hyatt wrote: >On February 28, 2001 at 14:49:19, Miguel A. Ballicora wrote: > >>On February 28, 2001 at 13:22:41, Robert Hyatt wrote: >> >>>On February 28, 2001 at 11:10:30, Miguel A. Ballicora wrote: >>> >>>>On February 28, 2001 at 05:56:36, Leen Ammeraal wrote: >>>> >>>>>I am not sure about when to avoid nullmoves. >>>>>I omit it: >>>>>a. when in check >>>>>b. when there are less than 5 pieces (including pawns) on the board >>>>>c. when the last move was a nullmove >>>>>d. at the root node >>>>>Should I also omit it in some other cases, >>>>>for example, when any hashmove (even with a low draft) was found, >>>>>or when beta = alpha + 1? >>>>>Thanks in advance for any help. >>>>>Leen >>>> >>>>Hi Leen, >>>> >>>>Regarding b, I do not know whether what I am doing now is correct but I think >>>>that works for me: >>>>When either black or white had no "long range" pieces (bishop, rook or queen) >>>>I disable null move. The rationale is that one side cannot waste >>>>a tempo in a given position having pawns, king and/or knights making the >>>>position prone to have a zugswang. >>>> >>>>Miguel >>> >>> >>>That seems dangerous. you are white, with a bishop on d5. I am black and I >>>have a pawn on a7 and g7. The bishop is zugged here. If your king can't move, >>>you lose even though you have a long-range slider on the board. And null move >>>will fail high here naturally as not moving is better than having to move and >>>lose. >> >>I think that you meant a3 and g3? > >Sorry. I am white, you are black trying to stop both of my pawns. > > >> If that's so I got your point >>and you're right. However, I disable nullmove when _either_ side lacks a >>long-range slider. In your example, it will be disabled because you do not >>have a bishop. If you do have a bishop, it won't be disabled (both sides got a >>slider) but at least I don't have "mutual" zugswangs which are the nastiest (I >>think). At least, with a slider per side the mutual zugswangs are more difficult >>(of course not impossible but I have to draw a line somewhere). > > >That only makes it worse. So I have a bishop and two pawns threatening to >promote. You have the bishop as above. You are _still_ zugged. I don't >see why you would limit null move based on _both_ sides. You should only >limit it if the side on move can be zugged. But in any case, it still fails >if we both have a bishop. The idea is to limit the nullmove when there is a zugswang that matters. In other words, there are zugswangs where it does not matter if I move or not. The example that you give could be one as pointed out by R. Gibert. If you move you lose because you are zugswang if you don't I win anyway advancing one of the pawns. Let me give an example trying to keep the spirit of your criticism so I can illustrate the idea: [D]2k5/1p6/1K5p/2p4P/2P5/1P2b3/8/4B3 w - - 0 1 White plays Bg3 and Black is zugswang. Black king can move without losing a pawn and Bishop can move defending c5 and at the same time avoiding Bf4. This is not a mutual zugswang, because if white is on the move still win wasting another tempo with Bh2 thanks to the slider. I think that is relevant for nullmove. For instance, nullmove won't fail high since "passing" gives no extra advantage to black, just keep the material.. so score won't be >= beta. Besides, in the nullmove line white plays again Bh2 and black still loses. What I think it is terrible is the mutual zugswangs because they always fail high. So, the idea is that if both sides have at least, say, a bishop, Q or R, mutual zugswangs are mode difficult (not impossible of course). That's where I draw the line. How do you limit your nullmove? you said a rook. One of the sides? Regards, Miguel > > >> >>Regards, >>Miguel
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.