Author: Hristo
Date: 17:47:09 04/05/99
Go up one level in this thread
On April 05, 1999 at 20:02:41, Dann Corbit wrote: >On April 05, 1999 at 14:18:41, Hristo wrote: > >>On April 04, 1999 at 14:40:40, Mike Hood wrote: >> >>>On April 04, 1999 at 13:56:40, Dave Gomboc wrote: >>> >>>>On April 04, 1999 at 13:32:30, Mike Hood wrote: >>>> >>>>>On April 04, 1999 at 11:21:55, Dave Gomboc wrote: >>>>> >>>>>>On April 04, 1999 at 10:02:37, Adnan wrote: >>>>>> >>>>>>> >>>>>>>I think Deep Blue at best is not better than 2650 to 2700 Fide. People are >>>>>>>overestimating it's strength just because on few games. The twelve games that it >>>>>>>played in 96 and 97 weren't really impressive. Kasparov *lost* the 97 match by >>>>>>>playing weak. But as far as the quality of games is concerened, Deep Blue made >>>>>>>silly blunders, even tactical blunders, something which computers are supposed >>>>>>>to be at best. For example, Deep Blue's tactical blunder in game 2 that would >>>>>>>have forced a draw, or tactical blunder in game 6, 1996, where it could not even >>>>>>>calculate a simple combination accurately and allowed 22. Bxh7+. >>>>>>> >>>>>>>If took real pity, I would rate it at 2650 to 2700 Fide -- AT BEST. >>>>>> >>>>>>You're certainly entitled to your opinion! You won't convince me without some >>>>>>concrete variations, though! Maybe provide the FEN position after Black's 21st >>>>>>move, a winning PV for White, including replies to important deviations for >>>>>>Black? Or did I miss a previously demonstrated win posted here? >>>>>> >>>>>>Dave Gomboc >>>>> >>>>>Adnan's point of view is valid. Deep Blue has played too few games to be >>>>>accurately judged. Any ELO ratings are rough estimations at best. Dave and Adnan >>>>>have different opinions that have to be left unjudged, unless IBM decides to >>>>>leave Deep Blue active for a long enough period of time to be examined. And >>>>>don't forget... IBM's newest supercomputers are bigger and faster than the one >>>>>that played against Kasparov, so the next incarnation of Deep Blue will play >>>>>even better. >>>>> >>>>>All the same, I consider myself a sceptic, as far as the strength of Deep Blue >>>>>is concerned. Is the program's strength what you'd expect from a computer so >>>>>powerful? To ask it another way, how strong would Fritz play on a computer with >>>>>that power? That's a naive question, because it's impossible to compare a single >>>>>processor PC with a multi-processor supercomputer, but I think you understand my >>>>>point. PCs have been relatively weak for years, so PC chess programmers have >>>>>worked hard on developing smart algorithms to make the best of the available >>>>>power. I remember playing against Psion on an 8 Mhz 286. That was a brilliant >>>>>program in its day. I fear (although I hope I'm wrong) that the Deep Blue >>>>>programmers rely too heavily on the power of the computers they have at their >>>>>fingertips, and not on the optimization of every last Assembler instruction. >>>>> >>>>>Dancer >>>> >>>>Adnan made a specific claim: that it made a tactical blunder in game six. I'm >>>>asking him to back it up, anyone can make a claim, but <shrug>. I didn't even >>>>bother with the game 2 "blunder", the search depth to see that Kf1 was only >>>>drawing was enormous and had many delaying quiet moves. >>>> >>>>The Deep Blue developers state in "Search Control Methods in Deep Blue" (March >>>>1999 -- and I haven't forgotten about you, Ernst!) that their policy was to >>>>harness the tremendous search power to provide as selective a search as >>>>possible. That certainly doesn't leave the impression in my mind that they're >>>>trying to brute-force everyone, any hype about its nodes per second count >>>>notwithstanding. >>>> >>>>Assember optimization only buys you a constant-time improvement. Algorithmic >>>>improvements can be worth much more. Sure, maybe you can squeeze 25 elo out of >>>>hacking 80x86 assembly, but today's PCs are of a speed that I don't think it's >>>>necessary to do this anymore. Indeed, there are very successful programs that >>>>don't: I recall a discussion on CCC some time ago (involving at least Amir Ban >>>>and Bob Hyatt). Amir stated that Junior (one of the best programs today) was >>>>written in C++ and that the speed loss relative to C was negligable. >>>>(Disclaimer: I hope I haven't misremembered.) >>>> >>>>Dave Gomboc >>> >>>I can't go into discussions on "tactical blunders". My chess skills aren't >>>enough for me to comprehend blunders made by players in the 2500+ range without >>>lengthy explanations by those wiser than me. >Unless there was a programming error of some kind, Deep Blue was *incapable* of >a tactical blunder. Anything perceived as such is almost surely a failure to >see deeply enough on the part of the one making such a statement. At 200M NPS, >with an average of 3 minutes per move, there are just not going to be any such >things as tactical blunders. Unless you consider something from a congested >position 8 full moves out which does not appear at 7 to be a "blunder". > Dann, you seem to know much more than I do about DeepBlue! But a simple math would show you that even at this great speed (200 M Nps) DeepBlue can not get very *deep* (using brute force) unless it uses some *tricks*.( Say 8 plys at 35 moves per ply is 2.25*10^12 positions(!), for 3 min DeepBlue can evaluate only(!!!) 3.6 * 10^10 positions ). Hash tables can help a lot, but this is only the first step. If one is to go Deeper, say 15-16 plys it will take much more than just hash tables, and any one of these algorithms that help reduce the amount of positions being evaluated could cause a blunder. Even if one could evaluate EVERY position(8 plys ahead) there is still a possibility of a blunder caused by the finall evaluation(there is always one of those). It is very possible a *concept* and not the actual implementation to be the problem with DB blinders. >>>I'm not suggesting that Deep Blue uses brute force. I'm only suggesting that if >>>IBM's computers were slower the programmers would have spent more time trying >>>every possible trick to squeeze 5 extra ELO points out of their baby. >>> >>>What annoys me most is the IBM mentality. They claim "We beat the world >>>champion", and now they've withdrawn to prevent the danger of being beaten by a >>>lesser grandmaster or *gasp* another chess program. >> >>Well, for 99% of the people out-there IBMs super computer won! I do not beleive >>there is any question about that. These 99% are the people that IBM as >>organization was going after anyhow. So the corporate mentality got what they >>paid for. From *chess*-scientist stand point we might all feel cheated( I do), >>but not because IBMs mentality. The *corporate* mentality is what caused my >>frustration. If we want to get further into the depths of chess we have to do it >>our selfs, because most of us will not be effected by the revenue vs expense >>situation. Even if we are then it will not be in scale that IBM is. >We should be grateful that IBM spent the enormous pile of money that they did to >advance computer chess to the incredible heights that they did. I will bet that >you will see a permanent spike in computer chess software sales after that >point. Many modern innovations are from the Deep Blue team. > >They have surely invested *millions* of dollars in research, development (and >yes, hype) on the Deep Blue project. Further, the chess games required a >supercomputer of great magnitude. Time on such a machine is a precious >commondity. Go and see what it costs to buy an hour on such a machine. I know it costs alot of money! >Now, to >imagine that they should run this machine endlessly for the good of mankind >would certainly be a noble gesture. But who's going to pay for it? Honest to >goodness, somebody has to. That same machine can be figuring out octane >mixtures using the simplex algorithm to make some oil company billions of >dollars. With an asset that valuable, it is asking a lot to have it sit around >and play chess with you. How can you perform a simplex algorithm on octane mixtures to improve oil and other *lubricants* with a hardware that has instructions like "Nc3 Bf4 Qc5 ..."? ;-)))) I wouldn't mind to play, though I don't think the games will be of great benefit to others. Perhaps 100-200 more games from DB would be all that we need to understand better how strong it realy is. Right now all we have is 6 games and lots of speculations. There is nothing wrong with this fact!!! Having 100 or so games to analyze makes for an interesting evening! Don't you think? > >On the other hand, I too, see that machine as one of the wonders of the world. >It sure would be nice to see it in full swing again. > I don't see it as a wonder of the world just yet !!! :)) It is amuzing for sure! regrds. Hristo >>> I doubt Deep Blue will ever >>>return, unless a future world champion challenges it. (Kasparov won't ask for a >>>rematch). If Deep Blue really is the world's best chess player, which it MAY be, >>>it ought to be left online to be challenged both by human grandmasters and by >>>aspiring chess programmers. What would it take to get Deep Blue out of >>>retirement? Maybe Chessmaster 7000 (or Fritz 6, or Hiarcs 8) should write on the >>>box "This is the program that Deep Blue is afraid to face". Maybe, just maybe, >>>the taunts would force IBM to take up the challenge. >I'm sure that they would cringe in fear at a micro challenge. >"As soon as you beat Kasparov in a 7 game match at 40/2 we will be waiting." >;-)
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.