Author: Thorsten Czub
Date: 13:23:33 05/18/98
Go up one level in this thread
On May 18, 1998 at 11:42:41, Moritz Berger wrote: >>Moritz is more interested in having a go at me via my program, but never >>mind :) >>Yes, I did, Moritz. I did CHECK it :)))))))))))))))))))) >I am still interested in having a go at your program - motivated by your >statements about "paradigms" and your opinion about Bob's performance I >quoted above: The fact remains that CSTal sucks *big time* in this >position. I don't see your point. The whole POSITION is tactics. It is not only tactics, it is very deep, isn't it ? The fast searcher are faster. CSTal is slower, but relating to the fact that it does much less NPS than Crafty does, it finds the move only 6 times slower ? IMO the plot of the position is too deep. But I want to see Chris exact explanation about this too. But where is your point ? Fast searchers are good in tactics. They are good in finding key-moves in unbalanced positions. This is an unbalanced position. White HAS to play this or that otherwise this or that happens. THIS IS FORCED ! And fast-searchers have much capture-extensions to find out about forced material unbalanced stuff. Thats their main job. Chris approach is another. In a position that is balanced, where you can play many moves and not any move is dead or win, there cstal plans and sacs for a plan to kill the king. Thats different ! CSTal is able to lead itself into the positions it can sac and sometimes win. Don't you see the difference between finders (years ago e.g. The King was the best finder, in our days nullmove-searchers are better) planners ? So - this is a position like all the colditz and bt-test and sort of this positions. It leads to nothing. It only measures how fast some programs reach a certain depth, or ? And others not ! Ok - it is about king-safety. Thats true. That is your point ? Really ? You want to say: here the king safety is FOUND by search earlier than CSTAL sees it by knowledge. You are right. In this position. I am sure you can find even more positions where CSTal is also slower. But is this your only point ? The important question is: when there is NO forced stuff, is a program able to judge this sort of dynamic stuff accurate. When there is NO forced move , suddenly tals plays a strategic move, that leads into a weakness. Other programs can do similar things. Mchess, Hiarcs, Rebel. Why is it important to win the race with the fast searchers in a field that is their advabtage by definition. They are pretty good in finding. We all know this. But this is - like a famous banned programmer would say - bean counting ! Virtual 2 e.g. sometimes plays really good positional planning moves. Although there is nothing FORCED. And nobody can tell us exactly if there is REALLY something to find. Difficult to find out about. But it plays those moves. WHY ? And - you will never see other search-programs FIND the key-move. Because they will only find the move after they understood the sense. When there is NOTHING to measure, than the strength of the speculative programs appears. in 80 % of all chess positions where there is NOTHING to force out. Like in life Moritz. >Here's the logfile for Crafty 15.7 on the correct position and the same >machine on which CSTal needed 80 minutes on to find a move that doesn't >lead to it being checkmated: WCrafty 15.7 is still 6.4 times faster than WOW - you even measured it THAT exactly ! Siehst du nicht dass du damit genau in den Napf trittst den Chris seit Wochen hier aufgestellt hat ? That this is exactly the bean-counting he makes fun about for a couple of days. 6.4 are you sure this result is not rounded ? Maybe we can say it even more precise ?? >Chess System Potemkin (named after field marshal Grigory Aleksandrovich) >in this position, even if we assume that Crafty doesn't play 26.b6 after >6 seconds (but after 12:36) ... Between 6 seconds and 12'36" is a huge difference, factor 126 ! Factor 6.4 or factor 126 is exactly 19.6875 times difference Moritz. I hope I was enough bean-counting (precise, facts, evidence, prove ) that you understood me from the language side :-))) In opposite to you I put a :-))) behind my message that you can see I just making fun. I don't laugh about you, but with you. You have not made :-))) >Crafty v15.7 > 11 12:36 ++ 1. b6!! > 11 16:29 3.16 1. b6 Rb8 2. b7 Ng3+ 3. hxg3 Rxf1+ > 4. Qxf1 hxg3 5. Nf4 Be6 6. Ncxe6 Qh4+ > 7. Nh3 Qxe4 8. Qf7+ <HT> > 11 32:55 ++ 1. g4!! > 11 36:19 3.88 1. g4 Ra7 2. gxf5 Bxf5 3. Bxf5 Rxf5 > 4. Rxf5 gxf5 5. Qf3 Qa8 6. Qxa8+ Rxa8 > 7. b6 a3 8. Kg2 Piff - Paff - Puff ! Bean counters, from counting themselves they come back and count the less important BENCHMARKS of their machines. I am sure Moritz your Porsche is faster than my porsche.
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.