Author: Robert Hyatt
Date: 18:05:45 03/23/03
Go up one level in this thread
On March 23, 2003 at 11:22:47, Chris Carson wrote: >On March 23, 2003 at 10:55:09, Robert Hyatt wrote: > >>On March 23, 2003 at 09:32:04, Chris Carson wrote: >> >>>On March 23, 2003 at 00:03:46, Robert Hyatt wrote: >>> >>>>On March 22, 2003 at 05:29:41, Chris Carson wrote: >>>> >>>>>On March 21, 2003 at 22:47:29, Robert Hyatt wrote: >>>>> >>>>>>On March 21, 2003 at 16:46:42, Uri Blass wrote: >>>>>> >>>>>>>On March 21, 2003 at 16:12:46, Robert Hyatt wrote: >>>>>>> >>>>>>>>On March 21, 2003 at 15:36:43, Uri Blass wrote: >>>>>>>> >>>>>>>>>On March 21, 2003 at 14:26:53, Robert Hyatt wrote: >>>>>>>>> >>>>>>>>>>On March 21, 2003 at 11:13:39, Chris Carson wrote: >>>>>>>>>> >>>>>>>>>>>On March 21, 2003 at 10:20:53, Robert Hyatt wrote: >>>>>>>>>>> >>>>>>>>>>>>On March 21, 2003 at 08:17:32, Chris Carson wrote: >>>>>>>>>>>> >>>>>>>>>>>>>On March 20, 2003 at 23:32:19, Robert Hyatt wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>>On March 20, 2003 at 19:19:44, Uri Blass wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>>On March 20, 2003 at 18:57:55, Russell Reagan wrote: >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>On March 20, 2003 at 17:07:11, Robert Hyatt wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>It's always interesting to read your short snippets about the history of >>>>>>>>>>>>>>>>computer chess. So when are you going to do us all a favor and write a book? :) >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>It was named "scratchy". It had the best win/lose record of anything that ever >>>>>>>>>>>>>>>>>played on >>>>>>>>>>>>>>>>>ICC. Something like 130 wins, 1 loss or some such. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>Even with all of the rating addicts who no-play other computers, no computer has >>>>>>>>>>>>>>>>surpassed this mark? >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>I think that it is easy to surpass that mark. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>You only need to have friends that you can always beat and set a formula to play >>>>>>>>>>>>>>>only against your friends. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>Uri >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>>Not if almost all the opponents are GM players. :) >>>>>>>>>>>>> >>>>>>>>>>>>>Hmm, wonder how DT would do against todays "inflated" GM's whe have better >>>>>>>>>>>>>anti-computer experience and knowledg? >>>>>>>>>>>> >>>>>>>>>>>>I'd suspect it would do the same as it did back then. The people that played it >>>>>>>>>>>>a lot already >>>>>>>>>>>>knew a lot about anti-computer play and they knew how dangerous the machine was. >>>>>>>>>>>> >>>>>>>>>>>>> Also wonder how DT would do against >>>>>>>>>>>>>players who use 1Ghz or faster comps/programs to help if they played DT today? >>>>>>>>>>>> >>>>>>>>>>>>No idea. Most of the DT games on the chess server were 2 12 type games, so >>>>>>>>>>>>using >>>>>>>>>>>>a computer to help the human would be doable... >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>>I am sure that DT would not be 130 points higher than anyone or thing on ICC >>>>>>>>>>>>>today. DT was ancient history and so was DB, good in their day, but that day is >>>>>>>>>>>>>past. You can see DBII at the Smithsonian here in DC. It is gone and in a >>>>>>>>>>>>>museum with all the other old relics. ;) >>>>>>>>>>>> >>>>>>>>>>>>The atomic bombs are "relics of the past" as well. >>>>>>>>>>>> >>>>>>>>>>>>But don't screw around with those relics. There's nothing else close to them, >>>>>>>>>>>>60 years after >>>>>>>>>>>>they were created. :) >>>>>>>>>>>> >>>>>>>>>>>>Old != obsolete. >>>>>>>>>>> >>>>>>>>>>>I do not think DT would do as well today, not even close. The Nukes of 60 years >>>>>>>>>>>ago were very dangerous, however, they are not even close to the strength of >>>>>>>>>>>todays bombs or even bombs from the mid 1950's. Technology moves foward. DT >>>>>>>>>>>was great in it's day, but that day is gone. You can see it for free here in DC >>>>>>>>>>>and go across the street and see the airplanes/bombs from the 1940's (still >>>>>>>>>>>dangerous) to the 1990's (very dangerous, no comparison, all obsolete). Same >>>>>>>>>>>for the Dino's (dangerous, but obsolete). :) >>>>>>>>>>> >>>>>>>>>>>old technology != current technology strength/destruction/performance >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>That's a serious mistake to make. Back in my active Karate days, when we had an >>>>>>>>>>annual state event, there was one "old geezer" that was always there competing >>>>>>>>>>in >>>>>>>>>>the 3-4 degree black belt group. And several used to comment about "jeez, hope >>>>>>>>>>I don't >>>>>>>>>>draw him for a match, but I'll try to take it easy on him if I do..." >>>>>>>>>> >>>>>>>>>>That "old geezer" put more black belts flat on their backs than any other single >>>>>>>>>>competitor at the events. :) >>>>>>>>>> >>>>>>>>>>_never_ underestimate something just because it is "old". Deep Thought is >>>>>>>>>>_still_ >>>>>>>>>>faster than any PC program running today, although the PCs are getting closer >>>>>>>>>>every >>>>>>>>>>six months. >>>>>>>>> >>>>>>>>>I do not see how can you compare speed. >>>>>>>>>Nodes per second mean nothing and I know that Deep thought has some problems >>>>>>>>>with repetition detection so I cannot compare their nodes with nodes of other >>>>>>>>>programs. >>>>>>>> >>>>>>>>Deep Thought had a max search speed of 14,000,000 nodes per second. I >>>>>>>>can certainly compare that to machines of today at 1ghz, and conclude >>>>>>>>that I'd rather have deep thought. It obviously wasn't weak, producing a >>>>>>>>2650 result against 25 consecutive GM players. >>>>>>>> >>>>>>>>I dind't try to conclude _exactly_ how much better or worse Deep Thought >>>>>>>>might be. I just concluded that it would certainly not be a _lot_ weaker than >>>>>>>>today's programs, and probably a bit bit stronger in fact. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>>> >>>>>>>>>It is comparing apples with orange because the program of today cannot run on >>>>>>>>>deep thought hardware and deep thought could not run on today machines. >>>>>>>> >>>>>>>> >>>>>>>>Nobody suggested either to the best of my knowledge. Just comparing deep >>>>>>>>thought >>>>>>>>to a 1ghz processor/program today. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>>> >>>>>>>>>Speed is also not the point but level of chess and there was a big progress in >>>>>>>>>software from the time of deep thought(for commercial programs more than 200 >>>>>>>>>ssdf points on the same hardware). >>>>>>>>> >>>>>>>> >>>>>>>>what "big progress"? DT did pretty well against GM players. On ultra-fast >>>>>>>>hardware today's programs might well be better. But at just 1ghz I doubt it. >>>>>>> >>>>>>>The problem is that GM's of today may be prepared better against computers so >>>>>>>even if programs on 1000 Mhz can do only 2600 performance against humans it does >>>>>>>not prove that they are not better than deep thought. >>>>>>> >>>>>>>When Tiger14 played in argentina against humans it got a performance that is >>>>>>>close to 2800 and it used less than 1 ghz. >>>>>>> >>>>>>>I believe that in other tournaments like the israeli league part of the humans >>>>>>>were prepared better against computers and I know at least about one master who >>>>>>>trained at home against Fritz before he drew against it(this option was not >>>>>>>possible for the opponents of deep thought and looking at games is not the same >>>>>>>as playing). >>>>>>> >>>>>>>Uri >>>>>> >>>>>> >>>>>>What was the time control of the argentina tournament? DT played what is >>>>>>probably the best time control (for humans) of 40 moves in two hours or >>>>>>40 moves in 2.5 hours. NO faster games were used for the 25 game fredkin >>>>>>requirement, as the fredkin prize stipulated 40/2hrs or slower. >>>>> >>>>>40 moves in 2 hours, we have debated this over and over and over many times >>>>>before. DT is history, it can be exceeded with older commercial programs (Chess >>>>>Tiger for one) on slow single processor (800 Mhz)hardware. >>>> >>>> >>>>I've heard the same nonsense _many_ times. 1983: "Belle's time has passed, >>>>it has been surpassed by the faster hardware and newer software." >>>> >>>>For an interesting tidbit, look up who won the 1986 ACM computer chess >>>>tournament in Dallas Texas. >>>> >>>>Count 'em out if you want. But that doesn't make 'em obsolete. >>> >>>Obsolete: "1 a : no longer in use or no longer useful". Source Websters Online: >>>http://www.m-w.com/home.htm >>> >>>CT/DT/DB are no longer in use. They are still useful, but no longer in use. >>>Obsolete by definition. I wish it were not so, I would like to see a new >>>version with updated s/w, hw, ..., but it is not going to happen. >> >> >> >>Actually point 1 does not apply. Deep Thought and Deep Blue Jr are _still_ >>around. We've had DB Jr on ICC a few times giving analysis for GM games on >>occasion. >> >>"no longer useful" certainly would not apply to a machine that fast... >> >>Being "inactive" is a long way from being "non-competitive" as implied. Belle >>was a prime example of that in 1986. > >DT is not compteative. When was the last time it played a game? DBjr might be >competative, but would loose a match aginst any of the top 5 on 2Ghz or faster >hw. > >I challenge you to produce DT and play any of the top 5 in a 6 game match on >2Ghz HW. Same for DBjr. You said DT and DBjr are still around, so I call your >bluff, set up the match and talk when you can produce some evidence. Good luck. > >Fritz 3 beat DB prototype (DT on DB hardware) for the 1995 Computer World Chess >Championship. In Fact, the last time DT won a world Championship was in 1989 >(DB never won) and the last time Belle won the World Computer Chess Championship >was in 1980. The commercial programs have held the World Computer Chess >Championship since 1992. > No "big iron" competed in the 1992 WCCC event. Deep Blue prototype (deep blue on deep thought hardware, not deep thought on deep bue hardware) did compete and did lose a game to fritz. One game. If you conclude superiority from one game, more power to you. I look at the previous ten years and conclude exactly the opposite. As far as "calling mhy bluff" there's nothing to call. Deep Thought hardware _is_ still around according to the deep blue team. And DB Jr _has_ provided analysis on ICC within the past two years. So it _does_ exist. Whether it will every play a game against a public opponent is another matter beyond my control. But feel free to draw any conclusion you want... >See: http://www.mark-weeks.com/chess/wcc-comp.htm
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.