Author: Vincent Diepeveen
Date: 11:22:15 05/12/98
Go up one level in this thread
On May 11, 1998 at 22:35:03, Robert Hyatt wrote: >On May 11, 1998 at 17:33:44, Vincent Diepeveen wrote: > >> >>On May 11, 1998 at 14:06:34, Robert Hyatt wrote: >> >>>On May 11, 1998 at 02:27:47, Howard Exner wrote: >>> >>>>On May 10, 1998 at 20:49:17, Robert Hyatt wrote: >>>> >>>>> round 1 was a lucky win by Kasparov... one or two tempi >>>>>and things turn totally around... >>>> >>>>Could you elaborate on this assertion? The way I'm reading >>>>it know is, "You were lucky to mate me because I was just about >>>>to mate you!" >>> >>> >>>Here's what a GM told me in looking over this. Basically, after >>>the fireworks were over, Kasparov had two passed pawns, while DB >>>had material. in most positions, the passed pawns (only two) would >>>not be enough to win, they would be blockaded and won. But in this >>>game, everything "just came together" to make this not happen. But >>>if you move one of the pawns back a rank, or up a rank, or make some >>>other modification to the board, the outcome could be quite different. >>> >>>Now the question is, did Kasparov "see" the ending, and *know* that he >>>could win? Hard to say. *if* he did, then there was no "luck". But >>>I don't believe he calculated that deeply, and he relied on "intuition" >>>that just happened to work in that game. In another game his "winning >>>intuition" was wrong and he drew what most thought was a won endgame >>>for him. >>> >>>I've seen Crafty hit such positions, once in the GM match last week in >>>the first round against Roman. bishops of opposite color. What >>>appeared >>>to be dead drawn to all watching. But everything was just set up so >>>that >>>it wasn't... was it luck or skill? Hard to say... but luck certainly >>>smiled on silicon in that game, IMHO. I'd bet that against a GM, Crafty >>>would draw 8 of every 10 such endgames. I'd bet that against DB, Garry >>>would lose 8 of every 10 such positions... >>> >>>But that's only an opinion, because I don't know what he really "saw". >>>I just suspect it didn't include what actually happened... >> >>Crafty is not doing checks in q-search, which makes you play these >>type of endings horrible, no matter how well your evaluation of it is, >>if you cannot look it up in EGTB. >> >>Just test it for a while: checks in q-search in the rook >>endgame, and you'll see that besides searching less deeply, your >>moves will improve. >> >>For some reason think crafty is not the standard chessprogram to see how >>to play rook endings are played by pc programs. >> >>Better look at how The King or Rebel plays it. >> >>Greetings, >>Vincent. > > >maybe we just disagree here. Crafty is currently playing a rook ending >vs Rebel in the KKup, and apparently winning it. Checks in the qsearch >are not absolutely essential, if regular extensions cover it well >enough. > >Again, don't tell me my approach is wrong... prove it to me... You play >enough games vs Crafty on ICC to be able to report how it plays..??? >I can give your "lifetime" win/lose/draw score vs Crafty if you'd >like... Start with the last game played of the new Diep version running under DoctorWho at a slow P200+ vs 4 processor crafty, 20 moves and crafty was busted. DoctorWho elorating 2414 P200+, crafty 4 processor Pentium Pro rating 2500. Not bad for 6 time slower hardware, and some bugs in Diep which causes it to get hung (so losing on time) or offer a draw at move 1, which most users directly accept. It'll be interesting to discuss crafty at ICC: If's weird that certain crafty accounts manage to lose a lot of games, and that other crafty accounts at nearly the same hardware nearly win everything. I have extensively studied this and come to a conclusion which describing it will take some a4 format papers and i'll wait some time describing it, if ever, but it would be interesting to discuss this and hear the thoughts of OTHERS too what THEY think causes this. So the question is: How can certain crafty accounts win with big scores, and how can at nearly the same hardware other crafty accounts lose with big scores? Of course i'm very interesting what your opinion about it is too, Bob. Greetings, Vincent
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.