Author: Dann Corbit
Date: 15:38:59 06/12/00
Go up one level in this thread
On June 12, 2000 at 18:22:16, Tom Kerrigan wrote: >On June 12, 2000 at 17:08:35, Ricardo Gibert wrote: >>On June 12, 2000 at 11:49:26, Tom Kerrigan wrote: >>>On June 11, 2000 at 22:18:41, Ricardo Gibert wrote: >>>>On June 11, 2000 at 21:38:27, Robert Hyatt wrote: >>>>>On June 11, 2000 at 17:46:32, Ricardo Gibert wrote: >>>>>>In the January/February issue of CiSE, there was published a list of 10 >>>>>>algorithms having "the greatest influence on the development and practice of >>>>>>science and engineering in the 20th century" . You can see this here: >>>>>>http://www.cecm.sfu.ca/personal/jborwein/algorithms.html >>>>>>Which I found this to be quite interesting, so naturally I wondered what the >>>>>>membership of CCC thought were the Top 10 Computer Chess Algorithms & Techniques >>>>>>having the greatest influence on the development and practice of Computer Chess. >>>>>>I'll get things started with my Top 10 List: >>>>>> >>>>>>1. Alpha-Beta search Algotihm >>>>>> >>>>>>2. Iterative Deepening >>>>>> >>>>>>3. Transposition Tables >>>>>> >>>>>>4. Null Move Pruning >>>>>> >>>>>>5. Chess Game Databases (Chessbase) >>>>>> >>>>>>6. Ken Thompson's Endgame Tablebases >>>>>> >>>>>>7. Judea Pearl's Scout Algorithm >>>>>> >>>>>>8. Bitboards >>>>>> >>>>>>9. Tim Mann's Winboard >>>>>> >>>>>>10. Robert Hyatt's source listing of Crafty >>>>>> >>>>>>BTW, I've tried to place items in the list in order of importance. The first >>>>>>four were easy, but I would expect a lot of disagreement in the next 6. In fact, >>>>>>I disagree with myself here. It's not easy. Deep Blue ought to fit in there >>>>>>somewhere. Also, I made a half-hearted attempt to include attribution, so any >>>>>>additional information or corrections will be appreciated. >>>>>> >>>>>>I can't help but notice the absence of commercial programmers from my list, but >>>>>>I think this is due to their keeping their methods "secret". History may >>>>>>remember their programs, but credit them with few innovations. >>>>> >>>>> >>>>> >>>>>I would delete 5, 6, 7, 8 and 10. I don't think the source for Crafty has been >>>>>an "important event". Gnuchess source has been out far longer, as has the >>>>>source for other programs like Sargon, Cray Blitz, chess 4.x, who knows what >>>>>else. >>>> >>>>Perhaps, but don't you I think your source has been more influential? Being >>>>first isn't everything. How influential a work is must be factored in too. >>> >>>What do you mean by "influential"? Do you mean that people have copied code and >>>algorithms from Crafty, i.e., cheated? I don't think this is such a great >>>quality. >> >>I think what I mean by influential is deliberately ambiguous. I don't have a >>strong opinion about items 5 - 10 in any case. So criticizing any of those does >>not bother me. I really have no interest in the list _I_ generated. I just put >>it out there as an example to help people come up with their own list. What I am >>_really_ interested in is the list _other_ people come up with. I want to know >>how _you_ think the word "influential" should be interpreted. >> >>So what would be your list? > >I think MacHack and CHESS deserve to be on the list... From what I understand, >both programs shaped chess programming philosophy. As far as programs on the list, and shaping the programs of others, I think the following programs deserve special mention (in no particular order): 1. SCP (Lots of programs are clones of this -- probably more than any other) 2. GnuChess (Similar to above) 3. Crafty (lots of programs have techniques clearly learned from crafty) 4. TSCP (without this, probably half of the programmers would not even try. "Le Premier Pas") 5. ExChess (another highly cloned program and very well done) Here are some programs which (as far as I can tell) are not used for techniques as much as the others but definitely *ought* to be: 1. Amy (the NegaScout search is beautifully simple and among the best) 2. Phalanx (search + evaluation in Phalanx is better than anyone else's. It's not even close. If Phalanx searched at 150-200K NPS it would absolutely crush everyone else and grind them into sawdust.) 3. ColChess (That's right, ColChess! His move generator is excellent) Here is another thing that would (might?) be a good idea: Study Heiner Marxen's CHEST program for mate discovery ideas. It might be especially good for panic mode searching (you're in trouble so hunt for any branch in which you might possibly mate by an opponent's mistake or perhaps even find a forced mate)
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.