Author: leonid
Date: 18:35:08 06/24/00
Go up one level in this thread
On June 24, 2000 at 15:02:09, blass uri wrote: >On June 24, 2000 at 13:58:26, leonid wrote: > >>On June 24, 2000 at 02:00:49, blass uri wrote: >>>I am not sure if I understand you. >> >>Ura, think about initial position on the chess board, as the most simple >>example. Final result if too far away. So, if you will do your calculation 5 >>moves ahead, you will have 20 moves of the same value as final result. Second >>stage (when you don't use openings) will help you to make the chose between >>them. > >Alpha beta will not tell you that there are 20 moves with the same value. Ura, never mind what kind of search you do, response is all the time the same. Difference is only in the speed of your response. >Suppose Na3 is the first move you start to search. > >alpha beta will tell you that Na3 gives equality and when the program(what you >call logic) analyzes other moves it will only prove that they do not give white >more than equality so only material evaluation will not tell you if they have >the same value or worse value. I am afraid that here I see some funny game of misused words that create all this fuss. For sure, when you use your evaluation logic (your logic look what material advantage will give you certain move), at given depth, you will have 20 moves of equal value for the initial position of the chess game. I presume that you don't use any openings and you have nothing more that something like Deep Blue. Final response is just too far away to be unique and only one. >It is only going to give you information if there is another move that wins >material >> >>>Do you say that you first evaluate material and if the evaluation is equal you >>>start to check between the equal moves which move is best by positional >>>evaluation. >>> >>>If you do it you do not use the alpha-beta because the alpha-beta gives you no >>>idea about the second best move[my opinion is that using alpha beta in the first Everything that you say to me sound very strange. I hardly can see how something could be done differently. Even when you look for mate you can have few moves, but when you look for mate situation is different. In mate solving logic when mate was not found, you have not even one single move as response. Response will be that move don't existe. When you have response saying that mate existe, and when you have more that one move to make it, you can drop your second, or more that one move, as superfluous. But in positional logic you have all the time some response. This response is one move or more that this. And when you have those multiple moves in positions logic, those moves are not necessarily winning moves, like in mate solving logic. This is the reason why you can't neglect any one from those moves as superfluous. You still must make your choice between them. >>Alpha-beta can give you very often as much as 15 moves of equal value in the >>first half of usual game. I speak all the time about what I call "positional >>logic". > >Alpha beta never gives you more than one move with a value and other moves that >you know that they are not better and you have no idea if they are equal or >worse. Once again, everything sound to me very, very strange. Maybe usually in other program two step approach is done at once. I doubt very much that this is the case. This way it could be very inefficient because sometime this second stage is just not necessary. It happened when some big material advantage existe. >Minimax can give you 15 moves with the same value but not alpha beta. Minimax and alpha-beta, like I said before, represent only difference in speed. > As I mentioned it before, multiplicity of the response is the main >>problem when you want to know efficency of your positional logic. More judicial >>choice between those moves can give the impression that one logic is better that >>the next one. All difference could be only in second stage, if other program >>work in the same way. Surprises are all the time possible. I did all my >>programming alone. Very often when I spoke with other programmers I found that >>they can hardly understand what I am been talking about. > >The main reason is that you do not explain yourself clearly. >For example you say logic when you mean program and you do not change it even >after people correct you. Not that big difference. My program can be divided into few difference independent part and contain only one logic. Mate solving logic or positional logic. Before my program contained just first mate logic and only later I started doing the rest. >Nobody else except you is using the name logic for a chess program. Will try to avoid confusion in the future but feel me too lazy to change what I already wrote. > Before I expected that >>all chess programs end sooner or later by the same main structure. But it is not >>what I see in reality. >> >>>plies is not a good idea because it is important to have an idea about the exact >>>evaluation to know better which lines to analyze(I see cases when programs >>>analyze moves that lead to mate in 1 because they do not reject the moves that >>>they analyze for the right reason) but you should use alpha beta most of the >>>time]. Mate or not mate, positional logic see all of them in complet way. It is just because positional logic do see all mate and draw that I even left in it the same initial structure for moves chain, like I have in mate solving logic. Checking moves goes first. >>I don't speak about bugs in the logic. I speak how it must work in general >>situation. >> >>Leonid. > >I did not speak about bugs but about alpha beta but it seems that you do not >understand what is alpha beta if you say that it can give you 15 moves with the >same value. My alpha-beta I found when I searched for "quick positional logic". It is possible that I found something else that I expected, but this I see as highly improbable. Leonid. >Uri
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.