Author: Robert Hyatt
Date: 08:38:24 03/24/03
Go up one level in this thread
On March 24, 2003 at 05:22:55, Chris Carson wrote: >On March 23, 2003 at 21:05:45, Robert Hyatt wrote: > >>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... > >We may have different opinions on this issue, however, I do value your insights >on the subject. > Just remember that my "insights" are based on: (a) programming actively in computer chess since 1968. 35 years and counting. (b) competing actively in computer chess events since 1976. 27 years and counting. If I had to play Belle today I would _not_ take it lightly. I'd expect to win, but I'd know that Belle was very capable of beating anything in one game. If I had to play Deep Thought today I would take that _seriously_ as it would be faster still than I am on the best hardware available (PC hardware of course). And I would _not_ expect to win easily. I'd be quite happy to just win at all. I had plenty of chances to see its tactical skills. While we might never _know_ how such a match would turn out, my experience says that it would _not_ be a foregone conclusion as to the winner, for either side. It would be a _match_ to say the least... > >> >> >> >> >>>See: http://www.mark-weeks.com/chess/wcc-comp.htm
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.