Author: Dann Corbit
Date: 17:37:28 01/20/05
Go up one level in this thread
On January 20, 2005 at 19:31:47, Steve B wrote: >> >>agree, more efforst should be put in designing a chess program that is most >>human-like in style. it does not have to be super-strong, and it will still sell >>well. Why do you think that a program that plays like a human will sell well? Have you done a market study? >why there is not much effort in that direction beats me. I suspect two reasons. 1. We have no idea how humans play chess and therefore we are not able to emulate it. 2. The people who write chess programs do not think it interesting enough to attempt it. There are chess programs that will blunder if you want them to. There are chess programs that will supposedly play like a certain GM if you ask them to (of course, that is mostly fiction). >sadly ,the direction is motivated by money >being on top of the SSDF list and having the best elo(even if it is by 1 point) >sells many chess programs If you are selling chess programs, what will be your motivation? If you want to see a chess program that plays like a human, why not do write one? If you don't know how to program, you should not let that stop you. Uri did not know how to program (according to his estimation) and yet he has written a strong program in a short time (movei). If you think a different direction is a good idea, then no one can stop you from going in that direction. ChessMaster has never been on top of the SSDF list, and yet it outsells all the others by a tremendous margin. I think that the SSDF will have a microscopic impact on sales, except for the peripheral buyers like those that read this board. But I am only guessing about that.
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.