Author: Robert Hyatt
Date: 13:16:58 10/30/99
Go up one level in this thread
On October 30, 1999 at 14:21:22, Christophe Theron wrote: >On October 30, 1999 at 10:08:51, Robert Hyatt wrote: > >>Go back to a 1990 ACM computer chess bulletin and note how many programs used >>null-move. It was already very common. I'll post a list on Monday as I have >>all of the ACM event publicity bulletins in a file in my office. In 1990 it >>had become a non-novelty (tournament applications at one point even asked >>null-move (Y/N)) when asking for details about an entrant. By 1995 it wasn't >>being asked. In fact, 1995 was the _last_ ACM event ever held. And the list >>of programs using null-move was already signficant before the JICCA article was >>published. > > >I would like to point out that Donninger's article was published in SEPTEMBER >1993 in the ICCA Journal. I have this issue in my hands. > >Let me quote the first words of the article: > >"At the 1991 Dutch championship, my program NIMZO played against QUEST by Frans >Morsch. QUEST and his twin brother FRITZ pull no punches in their weight class. >FRITZ, on a 486-PC, often searches to depths that can normally only be reached >by supercomputers (see, e.g., the game FRITZ-2 v. ZUGZWANG, Madrid 1992 (Friedel >and Valvo 1992)). FRITZ is a very compact program whose main ideas are inherited >from single-chip programs, so its evaluation function must be relatively simple. >Frans Morsch told me that he uses a null-move technique recursively (but, being >a professional programmer, he did not of course reveal the details)." > >In fact Frans revealed almost everything to Christian with these little words >"null-move technique recursively"! > > >I would also like to add that I personnaly read the first paper on null-move in >the Jean Christophe WEILL Phd thesis (published in 1995 and available since then >on the Internet). In this thesis, WEILL makes an explicit reference to >Donninger's article, as well as Beal, Goetsch and Campbell papers. > >May I point out that WEILL and BAUDOT are the authors of Virtual Chess? > >Null move papers may have been published before, but access to these papers was >not easy. Actually many amateur chess programmers know about null-move thanks to >the following people: > >* MORSCH (telling Donninger) >* DONNINGER (publishing a detailed paper in the ICCAJ) >* WEILL (publishing his Phd thesis on the internet) >* HYATT (publishing Crafty's source code) This leaves several gaps. Don Beal at the ACM events giving presentations on what he was doing, and describing null-move search in detail. Murray Campbell doing the same. All prior to 1990. The Chapter in the book "Computers, Chess and Cognition" about Cray Blitz was old. The book was written in 1989. It covered (I think it said that) Cray Blitz as of the 1989 WCCC in Alberta, Canada. Burton Wendroff (LaChex at Los Alamos weapons lab) sent me a prelim copy of Murray's paper at least in 1986 if not sooner and said "You ought to try this". The mistake is that "morsch" was the first. That's not close. Beal was first by Far, Hitech and Cray Blitz were done in the 1986 time-frame (null-move). I seem to recall others. Not to mention Donskoy using it in 1974 in Kaissa. > >These people made the null-move technique popular. That was the initial question >I think. It was not "who talked about the idea first" but "who made the >null-move so popular and widespread"? Depends on your definition of 'widespread'. By 1990 everyone I knew had tried it and either kept it or discarded it for reasons of their own choosing (IE Deep Thought never used it, although I think it was implemented in the software for Murray's testing. IE you could say that Crafty brought a great resurgance in bitboard programs. I would disagree. Slate/Atkin did the groundwork. Or Donskoy, et al, as Kaissa also used bitmaps. In scientific circles, the credit generally goes to the person that discovers/uses the idea first. Not to someone that tries it years later and finds that yes, it works well. The difficult part is coming up with the idea. Not in using it later, although that is not always easy. And there are always wrinkles, like the idea of a variable R factor that a few are using. > >There is no doubt in my mind about it. Every time we talk about null move we >refer to Donninger's article in the ICCAJ. > > > > Christophe Every time _I_ talk about it my reference is _always_ to Beal, and to Campbell. Perhaps my 'advantage' is that I was doing this stuff when they reported about null-move, and I clearly remember who coined the phrase and defined the algorithm.
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.