Author: Uri Blass
Date: 14:57:17 07/09/02
Go up one level in this thread
On July 09, 2002 at 17:46:02, Robert Hyatt wrote: >On July 09, 2002 at 17:19:40, Chris Carson wrote: > >>On July 09, 2002 at 16:35:07, Robert Hyatt wrote: >> >>>On July 09, 2002 at 16:10:46, Chris Carson wrote: >>> >>>>On July 09, 2002 at 15:26:58, Robert Hyatt wrote: >>>> >>>>>On July 09, 2002 at 13:38:03, Chris Carson wrote: >>>>> >>>>>>On July 09, 2002 at 13:27:31, Chris Carson wrote: >>>>>> >>>>>>>On July 09, 2002 at 12:51:35, Robert Hyatt wrote: >>>>>>> >>>>>>>>On July 09, 2002 at 07:35:55, Chris Carson wrote: >>>>>>>> >>>>>>>>>On July 08, 2002 at 23:18:00, Robert Hyatt wrote: >>>>>>>>> >>>>>>>>>>On July 08, 2002 at 14:49:22, Chris Carson wrote: >>>>>>>>>> >>>>>>>>>>>On July 08, 2002 at 14:26:22, Christophe Theron wrote: >>>>>>>>>>> >>>>>>>>>>>>On July 08, 2002 at 12:36:01, Robert Hyatt wrote: >>>>>>>>>>>> >>>>>>>>>>>>>On July 08, 2002 at 12:15:06, Uri Blass wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>>On July 08, 2002 at 11:32:38, Robert Hyatt wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>>On July 08, 2002 at 00:32:42, Christophe Theron wrote: >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>On July 06, 2002 at 20:15:06, stuart taylor wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>I suspect that search may see that the right move help to push the opponent king >>>>>>>>>>>>>>>>>>closer to the corner relative to the wrong moves and it may be enough. >>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>Uri >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>Yes, that looks like the best thing to try and work on, doesn't it? >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>If not, can I ask two questions?: >>>>>>>>>>>>>>>>>1)What should be done during the near future to push computer elo forward as >>>>>>>>>>>>>>>>>much as possible? >>>>>>>>>>>>>>>>>2)If Deeper blue was really much stronger than todays tops, what was that due >>>>>>>>>>>>>>>>>to? Better long-term planning? Seeing deeper? >>>>>>>>>>>>>>>>>S.Taylor >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Huge speed. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>It was doing most things worse than the best micro programs, but it was doing it >>>>>>>>>>>>>>>>so fast that it was eventually stronger. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Hum... Let me rephrase for the sensitive people out there. There was nothing >>>>>>>>>>>>>>>>Deep Blue did better than the best micro programs. But it was so fast that it >>>>>>>>>>>>>>>>allowed it to hide its defficiencies. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Shit. That's not very diplomatic either. Let's try again: Deep Blue was build >>>>>>>>>>>>>>>>around a concept outdated by 2 decades but fortunately it was so fast that >>>>>>>>>>>>>>>>nobody noticed until their creators published their paper. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Oops... OK, once again: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Bob likes Deep Blue a lot, and that should be a reason good enough to convince >>>>>>>>>>>>>>>>you that it was well designed. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Christophe ;-) >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>Er... excepting one game by Fritz in 1995, when was the last time you saw >>>>>>>>>>>>>>>any micro beat any predecessor of deep blue? When was the last time _your_ >>>>>>>>>>>>>>>program beat or drew them? Etc... >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>Results speak far louder than prejudice... >>>>>>>>>>>>>> >>>>>>>>>>>>>>Results can only prove that they were better than their opponents but this is >>>>>>>>>>>>>>not the question. >>>>>>>>>>>>>> >>>>>>>>>>>>>>Uri >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>>That is the problem. That was _the_ question. But since the answer is >>>>>>>>>>>>>clearly known, everybody wants to change the question to something that would >>>>>>>>>>>>>try to make deep blue look "less" than what it really was. But it was >>>>>>>>>>>>>unbeatable, considering that it lost to one micro in almost 10 years of >>>>>>>>>>>>>competition. Nobody _else_ has ever come close to that kind of dominance. >>>>>>>>>>>>> >>>>>>>>>>>>>I think it funny that _now_ the question becomes "was their search optimal"? >>>>>>>>>>>>>Implying that current micros _are_. Which is a joke. Both have enough holes >>>>>>>>>>>>>to supply a swiss cheese factory for years. The concept of "optimal" is a >>>>>>>>>>>>>joke. The concept of "results" is the only scientific way to measure the >>>>>>>>>>>>>programs against each other. The rest is only subjective opinion. >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>>There has been a big smoke fog spread around Deep Blue. >>>>>>>>>>>> >>>>>>>>>>>>At the time of the Kasparov match, we have been told that: >>>>>>>>>>>> >>>>>>>>>>>>1) it was extremely fast. >>>>>>>>>>>>2) it had much more knowledge than any other program around. >>>>>>>>>>>>3) it was using some revolutionnary search techniques. >>>>>>>>>>>> >>>>>>>>>>>>Now that we are able to see more clearly what it was, it turns out that: >>>>>>>>>>>> >>>>>>>>>>>>1) its superiority came from its speed. >>>>>>>>>>>>2) the rest was nothing new, and we are still trying to figure out what part was >>>>>>>>>>>>actually superior to what the best micro programs are doing. >>>>>>>>>>>> >>>>>>>>>>>>I don't think that noticing the above is against the interest of science. >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> Christophe >>>>>>>>>>> >>>>>>>>>>>I will be happy to publish the steps to pass muster for human (including GM's) >>>>>>>>>>>experiments. One quick note is that any "scientific" test to be valid must be >>>>>>>>>>>reliable/published so that it can be shown to be repeatable by an independant >>>>>>>>>>>scientist. >>>>>>>>>>> >>>>>>>>>>>The DB project was a secret thing, it was very nice " h/w technology", but I do >>>>>>>>>>>not consider much about DB to be related to science. I am not sure the DB >>>>>>>>>>>results are reliable, I would expect significantly different results if the >>>>>>>>>>>Human GM played a few more game (say 100 prep like the 2700 GM had against Rebel >>>>>>>>>>>recently). I expect DB 1996/97 would get beat by the PC's today in a "true" >>>>>>>>>>>double blind match/tournament. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>You were doing OK until that last sentence. Do you _really_ think you could >>>>>>>>>>take _any_ program from 1997, run it at 200M nodes per second, and that program >>>>>>>>>>would lose to today's micro programs at 1M nodes per second. I _hope_ you don't >>>>>>>>>>believe that. And yet we _know_ that DB 97 was certainly stronger than any >>>>>>>>>>1997 micro, because deep thought was stronger than any micro of its time and >>>>>>>>>>DB took a quantum leap 100X faster than Deep Thought. >>>>>>>>> >>>>>>>>>Read my last statement again. I said "PC's today", not programs from 97. Yes I >>>>>>>>>do believe that in a double blind match/tournament the top "PC's (single and >>>>>>>>>multi-processor chess programs" would beat DB 96/97. I would add that the >>>>>>>>>Programmers for Fritz, Junior, Tiger, Hiarcs, Shredder, Rebel would have to be >>>>>>>>>included and independant arbiter used. >>>>>>>>> >>>>>>>>>I also agree with Uri's reply: >>>>>>>>>http://www.talkchess.com/forums/1/message.html?239295 >>>>>>>> >>>>>>>> >>>>>>>>Reread what _I_ said. >>>>>>>> >>>>>>>>"if you take _any_ PC program from 1997, and magically find hardware fast enough >>>>>>>>to make it run at 200M nodes per second, then according to your above statement, >>>>>>>>you _must_ believe that today's micros would smash that PC in your 'double- >>>>>>>>blind' match". >>>>>>>> >>>>>>>>I don't believe that for a minute. And since DB 97 was stronger than any >>>>>>>>micro in 1997, you must believe that today's micros are far superior to 1997's >>>>>>>>micros, based solely on software. That is a crock. Today's programs are >>>>>>>>stronger. But not a _lot_ stronger, if you run 1997 vs today's programs on >>>>>>>>equal hardware. Hardware is a _lot_ of the strength gain. And DB had a _lot_ >>>>>>>>of strength. I don't believe today's programs could beat a 1992 micro program >>>>>>>>if it were running at 200M nodes per second. That is simply too large a time >>>>>>>>handicap and the tactics will rule the game. >>>>>>> >>>>>>>(1) What was the specific W L D record for Deep Blue 95 against the 1995 >>>>>>>programs/hw? It was 1 win 1 loss 1 Draw. (2) What was it for Deep Thought W L >>>>>>>D against the 1997 programs/hw? 0 wins 0 loss 0 draw. Deep Thought did not >>>>>>>play any of the 1997 pc programs. I do not see actual results to support your >>>>>>>statements. Please post the games and results for Deep Blue or Deep Thought >>>>>>>against the 1997 programs. >>>>>>> >>>>>>>Yes, the Programs today on today's hardware would smash the programs that Deep >>>>>>>Thought beat in 1989 on 1989 hardware. In 1989 DT beat Rebel X and Fidelity X >>>>>>>on 1989 harware, so what, big deal. I am sure any of the top programs on todays >>>>>>>hardware would have no problem winning. >>>>>>> >>>>>>>I do not see any "results" based evidence to support the statement that DB 96/97 >>>>>>>or Deep Thought (any year) was stonger than programs in 1997. I only hear that >>>>>>>Deep Thought beat two programs in 1989 and DB was 100 times faster. The >>>>>>>programs/hw in 1997 were close to DB96/97 and the programs today are better >>>>>>>than DB 96/97. >>>>>> >>>>>>I never said that the 1997 programs needed to be run at 200mnps. You said that. >>>>>> I think the 1997 programs were close to DB, not that far behind. 1997 version >>>>>>of Rebel on todays fastest single AMD would beat DB 96/97 in my opinion. DB >>>>>>96/97 needed the blazing speed, not the commercials. The HW/SW today would beat >>>>>>DB 96/97. >>>>> >>>>> >>>>>I didn't say you said _anything_ I clearly said that if you took a 1997 >>>>>program, and put it on "magic hardware" do you _really_ think that a program >>>>>/ machine from today would beat it, if this "magic hardware" ran the 1997 >>>>>program at 200M nodes per second? I don't think today's program would stand >>>>>even a small chance of winning any significant numbers of games at that time >>>>>handicap. >>>>> >>>>>And that time handicap is _exactly_ what 1997 DB would hold over _any_ program >>>>>of today on today's hardware... >>>> >>>>DB nps does not equal Rebel nps or Tiger nps or Fritz nps or ... You can not >>>>compare nps to nps. I look at results and there are no games (except human vs >>>>computers) for comparison. >>> >>> >>>You can't compare 1M nps to .5M nps to be sure. But you can _definitely_ >>>compare 1M nps to 200M nps and conclude something about the outcome. A factor >>>of 2-3 in NPS is possibly not significant. A factor of 200 is _always_ >>>significant. >> >>OK, I believe that top 5 comercial 97 programs at 200Mnps would beat DB 96/97 > >I don't. From experience. Going that much faster requires significant changes >to the search extensions and evaluation. Otherwise you go N plies deeper, your >extensions trigger far too much and the search explodes. My experience in few positions when I gave Movei to search for many hours is that the search did not explode. I also know that the search of genius3 did not explode after many hours but it happens with part of the other programs(Hiarcs7.32 and Fritz5). Today it is not a problem with most of the top programs and they seem to have a stable branching factor. Uri That is why it is not >easy to take a program and drop it into a supercomputer and have it play well. >It takes a _lot_ of tuning... > >DB had all the tuning. > > >>and I believe that 2002 programs Fritz, Junior, Tiger, Hiarcs, Rebel, Shredder, >>Nimzo running at 200Mnps would beat DB 96/97. > >Very possible. No way to debate that. But that is not what I said. DB at 200M >nodes per second is _far_ stronger than any program today at 1-2M nodes per >second. As a simple test I suggested taking a 1997 PC program, which was >certainly weaker than 1997 deep blue, and run that program at DB speeds. IE >give a 2002 program 1 minute per move, give the 1997 program 200 minutes per >move. There is little doubt which would win... which simply shows how >important that factor of 200 is in speed, and how unimportant the relatively >modest advances in software are. > > >> I also believe that the 97 >>progs/hw were close to DB 96/97 strength (not speed) and that the 2002 progs/hw >>are stronger than DB 96/97 in strength (not speed). > >Again, that is simply a crock. If the 1997 programs were close to DB, then >the 1993 programs were far stronger than deep thought since deep thought was >far slower, proportionally, than deep blue, compared to 1993 micros vs 1997 >micros. > >That argument simply can't be defended by any technical discussion of any kind >as 1993 deep thought was the _only_ program to produce a 2600+ performance >rating against GM players to claim the Fredkin prize. Many programs played in >rated tournaments back then too. But none were able to come even _close_ to >taking the Stage 2 Fredkin prize, until DT came along... > >Way too much evidence suggests that in 1997 _nothing_ was close to deep blue. >Yes, if the 1997 programs could have hit 200M nps they might have been as good >or even better. But they didn't, and they weren't.
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.