View Single Post
  #9  
Old September 3rd 07, 07:27 AM posted to rec.games.chess.analysis,rec.games.chess.misc,alt.chess,rec.games.chess.computer,rec.games.chess.politics
help bot
external usenet poster
 
Posts: 7,975
Default Winners for Aug 2007 at GetClub Chess.

On Sep 2, 9:59 am, Taylor Kingston wrote:

White: Sanny's program (at "normal" level)
Black: Fritz8 (in blitz mode -- the numbers following each move are
the time each machine took, in seconds)


It is not clear as to whether or not Fritz8 was thinking
on the opponent's (leisurely) time.



1. Nf3 Nf6


Who says Sanny's program can't go toe-to-toe with the
big boys in the opening?


2. Nc3 {127} d5 {0} 3. d4 {205} g6 {0} 4. Bg5 {180} Bg7 {0}
5. Qd3 {211} O-O {8} 6. O-O-O {477} Re8 {0} 7. e4? {118} -- Perhaps
the losing move. Fritz now shows its greatly superior calculating
ability. 7...dxe4 {19} 8. Nxe4 {157} Bf5 {0} -- A very troublesome
pin. White can't break it by 9.Nxf6+, since 9...exf6 wins a piece. 9.
Bxf6 {194} exf6 {0} 10. Re1 {191} Nc6 {14} 11. Kb1 {308} Qd5 {0} --
Fritz rated the position at about -1.45 here. 12. Nfd2 {104} Bxe4 {0}
13. Rxe4 {174} f5 {0} 14. Rxe8+? {238} -- Relatively best was 14.Re2,
accepting the loss of a pawn. Now Fritz rates the position at about
-3.10. 14...Rxe8 {6} 15. Nf3 {272} Qxf3! {4} -- The queen sac. Of
course if White retakes on f3 then 16...Re1+ and mate quicky. 16. Qd2
{187} Qe4 {0} 17. d5?? {172} -- There was no good move, but this is
very bad. Fritz's now saw a forced mate. 17...Qd4! {4} -- Threatening
18...Qxb2# and therefore forcing further material loss. 18. Qc1 {132}
Re1 {0} 19. c3 {141} Rxc1+ {141} 20. Kxc1 {159} Bh6+{159} 21. Kc2
{188} Qa4+ {188} -- At this point Sanny's program resigned. Black has
a forced mate in 4: 22. b3 Qxa2+ 23. Kd3 Ne5+ 24. Kd4 Qxf2+ 25.
Kxe5Bg7# 0-1


Perhaps there is a material limit which triggers the
program to resign? I find it hard to believe that Sanny's
program can spot a mate-in-four, since it nearly always
has trouble finding mate-in-one against itself in my
games.



Pretty much a hare-and-tortoise affair, except that here the hare
never let up. Sanny's program simply doesn't have the speed and
calculating ability to stand up to Fritz8 even in its quickest-moving
mode.


Does Fritz8 have a way of weakening the program to
USCF-rating equivalents? If it does, you might be able
to give some kind of estimate for the strength of the
program at GetClub.

Personally, I find it very hard to estimate, because at
times I feel I am playing a tactically dangerous computer,
yet at other times I feel like I am playing Sanny himself.


-- help bot






Ads
 

Xbox 360 - Loans - Credit Card - Loans - Web Advertising