Author: Robert Hyatt
Date: 18:20:34 08/20/05
Go up one level in this thread
On August 20, 2005 at 15:07:25, A. Cozzie wrote:
>I did not much like Zappa's play in this game, but I can't complain too much.
That was really a fun game.
First, congrats. Clearly this wasn't luck. You beat too many good programs,
too convincingly... So it is well-deserved.
As far as our game, my brief comments would be:
Good book line for us, then castling king-side started the fun. This has to be
a bug, because I actually have a king safety term to discourage castling
opposite to avoid these kinds of games that Crafty simply doesn't play very well
(yet). I'll look at that later tonight to see if I have a sign backward or
something since I saw it happen twice.
So we castled into an attack. Crafty promptly (and correctly) vaporized the
center and the attack pretty well dissipated. Except for the white kingside
pawns. But I think black was just fine after the center evaporated and
suddenly white was occupied on the queen-side defending its own king. And here
is where the current king safety was tried and found wanting. It played
_exactly_ like I would want it to play against any GM, by aggressively
prosecuting the attack on the white king. As a human, I would have chosen to
consolidate because there were still minor pending issues with open files and
the Bc2 diagonal. So rather than weigh its own king safety against attacking
white, and deciding (as I would likely have done) "I need to consolidate my
kingside, do something about those doubled rooks on the d file before they
become a problem." It chose to "lets go after the white king" and it almost
worked.
I can think of several scoring terms that could have turned this into a probable
draw. Opposing rooks on the open file. Nc6 and Rd8 was playable before it got
tangled up attacking the white king at a1. But it thought "OK, I have open
files, he has open files, that is equal" and it was a mistake. If I had the
king-safety asymmetry turned up higher, it would also have taken what I consider
the safer "consolidation" path rather than the aggressive path it followed that
exploded in its face. Unfortunately I turned the asymmetry down months ago, and
had ignored a couple of warning signs even in this tournament. I saw it castle
kingside, and then consider moves like h4 more than once, which sayd (to me)
"king safety is right on the edge of becoming unstable and overly aggressive."
But, I didn't give it enough thought...
Good game. White could certainly have lost with a mistake. It was a tactical
tight-rope that was fun.
We were attempting to compare search depths, since Vincent kept saying you would
out-search me by 2+ plies. Here is what I have in my log file for the first 10
moves out of book, search-depth wise:
14, 15, 15, 14. 14, 16. 15, 16, 15, 15
For every one of those searches, my target time limit was 2:10 +/- 9 seconds. A
couple went far longer when you did some very long searches. For example, you
took 12 minutes on move 12. So our search depths might not be exact comparisons
since I was primarily going for 2 minute searches around there.
Good luck tomorrow. Still working with the referees on the JICGA paper. Got a
revision to send to them next week I hope...
>
>Zappa now has at least 9.5 / 11 and is the 2005 World Champion.
>
>[Event "WCCC 2005"]
>[Site "Reykjavik"]
>[Date "2005.08.20"]
>[Round "10"]
>[White "Zappa 2.0b x64"]
>[Black "Crafty"]
>[Result "1-0"]
>[ECO "D45"]
>
>{800MB, zappa.ctg, VEL5933} 1. d4 {[%emt 0:00:00]} d5 {[%emt 0:00:15]} 2. Nf3 {
>[%emt 0:00:00]} Nf6 {[%emt 0:00:17]} 3. c4 {[%emt 0:00:00]} e6 {[%emt 0:00:43]}
>4. Nc3 {[%emt 0:00:00]} c6 {[%emt 0:00:11]} 5. e3 {[%emt 0:00:00]} Nbd7 {
>[%emt 0:00:10]} 6. Qc2 {[%emt 0:00:00]} Bd6 {[%emt 0:00:09]} 7. g4 {
>[%emt 0:00:00]} Bb4 {[%emt 0:00:14]} 8. Bd2 {[%emt 0:00:00]} Qe7 {
>[%emt 0:00:13]} 9. Rg1 {[%emt 0:00:00]} Bxc3 {[%emt 0:00:08]} 10. Bxc3 {
>[%emt 0:00:00]} Ne4 {[%emt 0:00:12]} 11. O-O-O {[%emt 0:00:00]} Nxc3 {
>[%emt 0:02:20]} 12. Qxc3 {[%eval 95,18] [%emt 0:11:03]} O-O {
>(dxc4) [%emt 0:00:11]} 13. Bd3 {[%eval 81,17] [%emt 0:09:53]} c5 {
>(dxc4) [%emt 0:02:36]} 14. Kb1 {[%eval 88,15] [%emt 0:05:45]} cxd4 {
>(dxc4) [%emt 0:00:14]} 15. Qxd4 {[%eval 55,16] [%emt 0:06:15]} dxc4 {
>[%emt 0:04:16]} 16. Qxc4 {[%eval 37,16] [%emt 0:00:53]} Nc5 {[%emt 0:01:25]}
>17. Bc2 {[%eval 55,16] [%emt 0:03:32]} b5 {[%emt 0:02:45]} 18. Qf4 {
>[%eval 54,17] [%emt 0:02:13]} Bb7 {[%emt 0:00:09]} 19. g5 {
>[%eval 47,16] [%emt 0:02:40]} Bxf3 {[%emt 0:02:04]} 20. Qxf3 {
>[%eval 50,16] [%emt 0:00:42]} Rac8 {(g6) [%emt 0:01:42]} 21. Rd4 {
>[%eval 40,15] [%emt 0:02:47]} g6 {[%emt 0:00:10]} 22. h4 {
>[%eval 36,16] [%emt 0:02:33]} Nd7 {(e5) [%emt 0:04:03]} 23. Rgd1 {
>[%eval 33,15] [%emt 0:02:39]} Nb6 {[%emt 0:02:35]} 24. Be4 {
>[%eval 37,15] [%emt 0:00:00]} Nc4 {(Rfd8) [%emt 0:02:25]} 25. Qe2 {
>[%eval 14,16] [%emt 0:02:39]} e5 {(Rc7) [%emt 0:00:10]} 26. Rd5 {
>[%eval 43,16] [%emt 0:02:41]} a6 {[%emt 0:02:38]} 27. Ka1 {
>[%eval 47,17] [%emt 0:00:00]} Qb4 {(Rfd8) [%emt 0:02:39]} 28. Bb1 {
>[%eval 75,15] [%emt 0:02:22]} Rc7 {[%emt 0:02:51]} 29. Rd8 {
>[%eval 42,17] [%emt 0:02:26]} Rc8 {(Qe7) [%emt 0:00:08]} 30. R8d7 {
>[%eval 83,16] [%emt 0:01:41]} Na5 {(Nb6) [%emt 0:01:23]} 31. h5 {
>[%eval 98,16] [%emt 0:01:40]} Qa4 {[%emt 0:00:57]} 32. R7d3 {
>[%eval 106,15] [%emt 0:00:42]} Rc6 {(Rcd8) [%emt 0:06:31]} 33. f4 {
>[%eval 113,15] [%emt 0:01:40]} exf4 {[%emt 0:16:30]} 34. hxg6 {
>[%eval 283,18] [%emt 0:00:00]} fxg6 {[%emt 0:04:45]} 35. exf4 {
>[%eval 303,17] [%emt 0:00:00]} Qxf4 {[%emt 0:02:24]} 36. Rd7 {
>[%eval 303,16] [%emt 0:00:00]} Qxg5 {[%emt 0:02:04]} 37. Qe1 {
>[%eval 336,16] [%emt 0:00:00]} b4 {[%emt 0:07:00]} 38. Qxb4 {
>[%eval 375,16] [%emt 0:00:00]} Rc5 {[%emt 0:02:15]} 39. Qa3 {
>[%eval 409,15] [%emt 0:00:00]} Qh5 {[%emt 0:02:24]} 40. Qe3 {
>[%eval 476,16] [%emt 0:00:00]} Re5 {[%emt 0:01:48]} 41. Qc3 {
>[%eval 533,16] [%emt 0:00:24]} Rfe8 {[%emt 0:03:14]} 42. a3 {
>[%eval 588,16] [%emt 0:00:00]} Qg5 {[%emt 0:02:47]} 43. Re1 {
>[%eval 660,16] [%emt 0:00:00]} Qf6 {[%emt 0:02:18]} 44. Rxe5 {
>[%eval 961,18] [%emt 0:00:24]} Qxe5 {[%emt 0:02:32]} 45. Ba2+ {
>[%eval 958,17] [%emt 0:00:16]} Kh8 {[%emt 0:01:13]} 46. Rd8 {
>[%eval 1189,20] [%emt 0:01:43]}
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.