Author: Robert Hyatt
Date: 13:08:30 09/20/05
Go up one level in this thread
On September 20, 2005 at 15:31:46, Bruce Moreland wrote: >On September 20, 2005 at 13:08:04, Robert Hyatt wrote: > >>On September 20, 2005 at 12:14:46, Vincent Diepeveen wrote: >> >>>On September 20, 2005 at 11:52:19, Robert Hyatt wrote: >>> >>>You can analyze the games of deep blue. Those 6 it played in 1997, >>>as before that it was clearly beated by fritz3 in 1995. >> >>Deep Blue was never beaten by Fritz. Deep Thought was. I suppose I'll have to >>keep reminding you of this until one of us dies??? > >The thing entered the competition as "Deep Blue Prototype". It was listed that >way in the official program, if I recall correctly. The name was changed to >"Deep Thought II" coincident with the start of the tournament. > >I believe that Hsu has stated that it was running very very markedly slower than >what eventually beat Kasparov. It sounds like something that you could call an >earlier version though. > >bruce They played in the ACM event that same year, and used this in the tournament bulletin provided to all players/observers: "Deep Blue Prototype". Deep Blue software, running on the deep thought 2 hardware. What that says to me is that the software search stuff (C code) was what they expected to use against kasparov in 1996, but the hardware chips were not ready until just prior to the Kasparov match due to some inductive noise problems, so they used the deep thought hardware instead. They claimed to be hitting 2M nodes per second, roughly, in that ACM event. I believe that the 1996 DB chip didn't add a lot of features over the last deep thought chips. But the 1997 DB chips added significant new features (in terms of hardware evaluation). So many in fact that they were unable to use them all for the 1997 match since the chips were again "hot off the presses" about a month before the event started... Technically, from chiptest in 1986 to deep blue 2 in 1997, all were "related" just as all versions of Crafty are related. But claiming that "deep blue" lost to fritz is a way-big stretch... Of course one has to consider the source of that statement as well, so maybe no big deal. :) > >> >> >>> >>>The games are very bad. It plays ugly moves. See ICCA report journal june 1997, >>>with all the ugly moves it plays. On average 5 bad moves a game. >> >>Again, its predecessor produced a 2650+ FIDE performance rating, playing _only_ >>grandmasters, _only_ in 40/2hr games (no blitz or anything else was counted). >> >>So how can it be _that_ weak, unless all the grandmasters on the planet suck, >>including probably the best player of the last 50-60 years, Kasparov??? >> >> >> >>> >>>Additional it's an utmost passive program. >>> >>>None of todays programs is passive. Further may i remind you it just searched >>>10-12 ply. See the logfiles. Important moves just out of book it reached 10 >>>plies. >> >> >> >>So. How deep does chessmaster claim to search? how does your program do >>against chessmaster/the king even though you are out-searching it by 6-8 plies? >> >>There are plies, and there are "plies". Not all plies are created equally. >> >>One day you will figure out that NPS doesn't mean much when comparing different >>programs. One day you will figure out that ply depth doesn't mean much when >>comparing different programs. And maybe one day you will figure out that move >>generation speed doesn't mean much period. It is the overall result that is >>important in chess, and DT/DB produced good _results_. Again, show me your >>programs 2650 performance over 25 consecutive games against GM players only at >>40/2hr time controls only... >> >> >> >> >>> >>>Now it might have gotten 12.x on average of course, but that's nowhere near >>>todays 'average' search depth. >>> >>>My average is about 18 ply, thanks to the big depths in endgame. >> >> >>So? I averaged 15 at the WCCC. How did your 18 plies do against my 15 plies? >>As I said, plies are meaningless when comparing different programs, just the >>results have any meaning. >> >> >> >> >>> >>>Vincent >>> >>>>On September 20, 2005 at 11:16:16, Vincent Diepeveen wrote: >>>> >>>>>On September 19, 2005 at 10:06:16, K. Burcham wrote: >>>>> >>>>>From a chessplayer viewpoint the judgement is very easy. Deep Blue played like a >>>>>big crap. Todays software on other hand plays real strong. >>>> >>>> >>>>I'm sure there are a bunch of "big crap" grandmasters as well, since the lowly >>>>deep thought played so many of them and produced a 2650+ performance rating at >>>>40/2hr time controls. >>>> >>>>Yep, deep blue was "big crap". And since deep thought was 100x slower, it must >>>>have been "big crap * 100". What does that say about all the human players on >>>>the planet? "bigger crap"? >>>> >>>>Only "crap" I see here lies inside your inane comment... >>>> >>>> >>>> >>>> >>>>> >>>>>> >>>>>> >>>>>>Pretty remarkable, and it shows that they were extremely strong compared to >>>>>>everyone else during that period. >>>>>>DB was just "a lot faster, and a lot smarter" than deep thought. It was (and >>>>>>would still be) competitive... >>>>>> >>>>>>Robert Hyatt >>>>>> >>>>>>Robert, you have always had "faith" in Deep Blue playing in a tournament against >>>>>>todays programs. What do you base this on? Mostly just a gut feeling? Is there a >>>>>>game that you were impressed with Deep Blue knowledge? Maybe just the fact that >>>>>>Deep Blue held its own against Kasparov? >>>>>>I read your point that you thought Deep Blue was strong for several years, but >>>>>>its competitors may not do so well against todays programs. >>>>>> >>>>>>Maybe you are saying that with improvements between 1997 and 2005, Deep Blue >>>>>>would be very strong today. Are you saying that Deep Blue, exactly the way it >>>>>>was in 1997 would be competitive today, with its 1997 search depth and 1997 >>>>>>knowledge? >>>>>> >>>>>>I am a fan of Deep Blue, its hardware and what they accomplished. I have spent >>>>>>days trying to find a line or move that todays programs will not play. I cannot >>>>>>find this, in fact we know that Deep Blue could not see 44.Kh2 wins in game 2. >>>>>>Instead Deep Blue played the draw move 44.Kf1. The knowledge and depth was not >>>>>>there to avoid this move. Todays programs also will not play this with winning >>>>>>eval. Some will play, but like you said once, not for right reason. >>>>>> >>>>>> [D] R7/1r3kp1/1qQb1p1p/1p1PpP2/1Pp1B3/2P4P/6P1/6K1 w - - 0 44 >>>>>> >>>>>> >>>>>>2006 Unlimited World Open >>>>>>Fruit 2.2 >>>>>>Zappa 2.0 >>>>>>Deep Fritz 9 >>>>>>Crafty 21.4 >>>>>>Deep Blue 3.1x >>>>>>Shredder 10 >>>>>>Hiarcs 10.2 >>>>>>Deep Junior 10 >>>>>> >>>>>>kburcham
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.