Author: Tony Werten
Date: 06:57:35 04/08/01
Go up one level in this thread
On April 08, 2001 at 08:41:19, Uri Blass wrote: >On April 08, 2001 at 06:47:56, Aaron Tay wrote: > >>On April 08, 2001 at 06:14:44, José Carlos wrote: >> >>>On April 07, 2001 at 11:18:34, Urban Koistinen wrote: >>> >>>>I have written down a algorithm for computing endgame tablebases that should be >>>>about 10 times quicker than the Nalimov algorithm and requires much less ram. >>>>It is similar to the Arlazarov&Futer algorithm of 1979 but is more general as it >>>>does not require a pawn. >>>>It might be too technical for most here, >>> >>> Thank you. Too technical algorithms are not for a crowded-of-programmers forum like this. We wouldn't understand anything. >>> >>> José C. >> >>Hey , be nice. He's new here. >> >>He is just feeling a little fustrated after posting at r.g.c.c and getting only >>one response. Knowing the high level of general technicial expertise here, I >>suggested to him that he should post here. >> >>So what do the experts think? Is the algorithim as good as he claims? Layman >>like me want to know! > >1)I tried to explain the paper in more words in order to understand it and the >first part that I did not understand was the sentence: >"t4 depends on t3 and d" >It seems to me that t4 is dependent also on t1. > >Here is an example of position in t4 Why is this an example of t4 ? Might be t100, white has just captured a queen. Or t99, black has taken the other white queen. Counting the number of moves for the 50 move rule has nothing to do with distance to mate. ( Or I really don't get it ) BTW with pawns on the board, t100 and t99 will almost be the only tables with any information. cheers, Tony > >[D]2k5/8/2K1Q3/8/8/8/8/8 b - - 0 1 > >The position after Kb8 is in t3 but the position after Kd8 is in t1. > >I prefer not to try to understand the rest of the paper before I understand this >point. > > > >2)Here is my explanation of the part that I did understand: > > >I call capturing or pawn move in the same word conversion. > >tables ti are clear >They include all the positions when it is a win for white but white need i plies >to get the conversion. > >Table d include only legal positions that are not stalemate and black cannot >save itself by conversion. > >We want to find positions that are win for white. >other positions are not intereting because it is obvious that they are not win >for white. > >It is clear that t1 depends on t0 and previous tables that can be reached by a >conversion(a position is in t1 only if white wins by conversion or by checkmate >in 1 ply). > >t2 are positions when black is to move. >The position is legal and black cannot save itself by conversion and it means >that t2 is included in d. >If black try to play a move that is not conversion it gets to position in t1 so >it depends also on t1. > >t3 are positions when white is to move. >White cannot win by conversion otherwise the position is in t1. >White can get by a move that is not conversion to position in t2. > >This is the reason that t3 is dependent on t2 and t1. > >Uri
This page took 0.01 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.