Author: Vincent Diepeveen
Date: 05:42:36 12/12/02
Go up one level in this thread
On December 11, 2002 at 17:39:12, Maurizio De Leo wrote: >On December 11, 2002 at 14:04:19, Vincent Diepeveen wrote: > >>>It's interesting that you would rank Western Chess as more complex than Go. >>>Western chess programs are GM-strength. Are Go programs equally capable? >> >>If similar effort would have been put in go like it has been in chess, >>my answer would be yes. But i could imagine that after my chessprogram >>has been finished (perhaps never) that when i find someone who is good in >>GO that i improve my GO program. I see most of the strongest go programs >>make mistakes which some simple evaluation knowledge can fix easily in >>combination with search. > >You just seem not to understand how complex is Go to program. I'm a club player >of chess,checkers and go and I won't even dare to say which game is more >interesting or more complex from a human player point of view. > >But for the computer-side there is no question. I'm not a professional >programmer, but for sure I can write a chess-program that can beat all the >"non-club" player. I think a "brute-force" 5 or 6 ply should be enough. >While best Go program, with commercial interests and years of development (like >Many Faces) are still weaker than a relative beginner (let's say someone who is >10/12 kyu or 1500/1600 Elo). > >>Despite huge evaluation functions some still do not know much from evaluation >>programming i get the impression. They focus upon things humans find difficult >>instead of fixing some major strategical bugs which causes me to win from >>all GO programs. > >The problem is that fixing what you call "major strategical bugs" is not easy at >all. The strategic evaluation is really difficult to implement because of very >generic and not algoritmic things like "thickness", "shape" and so on. > >>It's an amateur league really. I lack positional knowledge to create a good >>GO program. Like how you attack a weak group. It's all standard for strong >>GO players. Not for me. > >This doesn't me that if you had this positional knoweledge you can easily make a >good go program. I'm sure that if you can make a 1/2 dan (candidate master) go >program, you will make a lot more money than making the strongest chess program >: why you don't try, if you are so sure ? When you have a chessprogram in the world top, the rest is complete pathetic compared to that. It is clear that also for chessprograms you need loads of chess knowledge, just like for Go. The difference is that a single person will be capable of writing a go program within a year or 2 that can easily challenge the world top. In my case if i would be busy a full year fulltime, i would surely, even without help of a strong go player, be capable of challenging the go-top. However it is impossible to write a chessprogram within 1 year that can challenge the world top computerchess. *impossible*. Even the best go programmer will be capable of telling you why. They are just busy figuring out knowledge from chessprogrammers and which is known in the computerchess world for years, in order to improve their own go program. They should figure out things like *how* do some of the oldy chessprograms do their selective search? >>But consider this. Nullmove works a lot better in GO than it does in chess. >>The quotes branching factors are bloody nonsense. Even for a 6 ply search >>the b.f. it is already way under 10.0 in GO. > >Ok. So let's analyze. For 6 ply 10^6 moves in Go and 4^6 = 4096 in Diep (or do >you have a branching factor over 4 ?). It seems that you have to squeeze some >three order of magnitude from optimization. Every beginner can get a better b.f. than the top go programs, but you won't be capable of writing a correctly searching chess program that can challenge the strong chess programs b.f. Getting my point slowly? >>Some also do not know how to >>program. Some evaluation functions scan the board in a very idiotic way. >>Branches all over which can be replaced. A simple lossless rewrite making it >>10 times faster... > >As I pointed before, you should get something more than that. > >>Add to that, that there is no GM title in GO. > >So ? What this mean ? >There is no "professional" title in Chess. There is no "Davis Cup" in soccer. >And so on with nonsense statements. > >>However important is to realize that if i pay a GO player money to play >>my go program, without paying him to win (like it happens in the computerchess >>world) and make all the conditions equal to what happens in the >>computerchess world, that the GO player will have a real hard time. > >If you pay the Go player to win he will have a real hard time ?? >So if professional isn't paid he can kick the butt of all programs giving 12 >stone advantage (= queen advantage) but he will start having problems if he is >paid ? >What kind of things are you saying ? > > >>Do not forget that the first 'beating' of GMs in the computerchess world >>was at 5 0 level. >>I have to see the first professional go player manage to play 150 moves >>within 5 minutes instead of a full day. > >Give the professional 15 mins for 150 moves (instead of 5 mins for 50 moves in >chess) and it will beat three program simultaneously. > >>Suppose you let a small FM play against a strong chessprogram at a level >>of a full day for the entire game. >> >>Of course the FM will win. > >What ? >Usually go players have 6/8 hours for the game which isn't far to 2h/40+2h/40 >+30min of top chess. Anyway you are a strong Fide Master, almost international >master. Would you mind beating Deep Fritz 7 by 10-0 (i would accept also 9-1) ? >Maybe time can be 10 hour for player for the game ? Or 12 hour, as you want. > > >Maurizio
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.