Author: Jeremiah Penery
Date: 11:50:36 01/23/00
Go up one level in this thread
On January 23, 2000 at 14:27:02, Tom Kerrigan wrote: >Here's my original question, which started this whole argument: if FHH knows how >to make such a terrific evaluation function, why doesn't he put one in a PC >program and make a lot of money from it? My question is "Why hasn't someone _else_ done it?". It sounds pretty much like DB just had a bunch more knowledge, and extended more/differently than PC programs. If it would be a good program this way, why hasn't someone else implemented all their knowledge and search techniques (as much as they can figure out) already? Probably because it would perform pathetically on the PC. Take CS-Tal, for example. Ignoring its speculative evaluation, it often gets tactically killed by the current 'top' micro programs. It gets at most 10k NPS on fast hardware(?), and I doubt it has even as much knowledge as DB. It also uses a selective search, which DB does not do. If you add all the extra DB knowledge, the program will be even slower. With no forward pruning/selective search, the depth will be even less. With all the extensions DB did, the depth would be still less. But you're asking why Hsu doesn't do it? Here's what I think would happen: It would probably sell a bunch. So what? It won't perform very well, and the anti-DB stuff will be even worse. A lot of people won't really understand the speed difference between this and the real DB; they will call DB weak and stupid. People who already think DB wasn't very good will have their opinion reinforced. It's pretty much a no-win situation for Hsu. Besides the money, which he probably doesn't need (or may not care about, if you consider the above), there is absolutely no motivation for him to do something like this.
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.