Author: Uri Blass
Date: 03:10:30 05/28/02
Go up one level in this thread
On May 27, 2002 at 23:32:15, pavel wrote:
>HLCC - 3
>
>=====================================
>HLCC - 3 (Clash of the Titans)
>
>HLCC = (High Level Computer Chess)
>=====================================
>
>Participants:
>
>Fritz 7 (updated)
>The Highest rated program in last SSDF
>list,(http://w1.859.telia.com/~u85924109/ssdf/list.htm). The author of the
>Program is Frans Morsch.
>
>Homepage of Fritzz7: (http://www.chessbase.com/catalog/product.asp?pid=85)
>
>Hiarcs 8:
>One of the strongest Chess Programs out there. The Author of the program is Mark
>Uniacke.
>
>Homepage of Hiarcs8: (http://www.chessbase.com/catalog/product.asp?pid=112).
>
>==============================
>The game can be seen LIVE here --->
>http://members.bellatlantic.net/~vze33v6m/live/
>(Auto-update is set to 10mins)
>==============================
>
>Fritz 7
>Book --> Fritz7.ctg (optimized)
>Hash --> 128mb
>TB --> 3,4,5men
>Ponder = off
>
>
>Hiarcs 8
>Book --> Hiarcs8.ctg (Tournament version)(optimized)
>Hash --> 128mb
>TB --> 3,4,5men
>Ponder = off
>
>==============================
>Game Condition -->
>
>Time Control = 12hrs/side (also can be said 720mins/side)
>A set of 10 games will be played.
>Machine = Pentium III 1 Ghz/ 512mb ram / Windows XP.
>
>
>You can see results from previous HLCC tournaments here: --->
>http://members.bellatlantic.net/~vze33v6m/tournaments/tournaments.html
>
>===============================================================================
>[Event "HLCC-3"]
>[Site "home"]
>[Date "2002.05.27"]
>[Round "2.1"]
>[White "Fritz 7"]
>[Black "Hiarcs 8"]
>[Result "1-0"]
>[ECO "D30"]
>[WhiteElo "2500"]
>[BlackElo "2500"]
>[PlyCount "101"]
>[EventDate "2002.05.25"]
>[Source "Mahamood"]
>
>1. d4 {0} 1... d5 {0} 2. Nf3 {0} 2... Nf6 {0} 3. Bg5 {0} 3... e6 {0} 4. c4 {0}
>4... h6 {0} 5. Bxf6 {0} 5... Qxf6 {0} 6. e3 {0} 6... Nc6 {0.30/13 2167} 7. Nc3
>{0.38/16 1723} 7... Bb4 {0.38/14 2116} 8. Qb3 {0.38/15 1092} 8... dxc4 {0} 9.
>Bxc4 {0} 9... O-O {0} 10. O-O {0} 10... Bd6 {0} 11. Ne4 {0} 11... Qe7 {0} 12.
>Qc3 {0} 12... Bd7 {0.51/13 1089} 13. Rfd1 {0.50/16 2436} 13... Bb4 {
>(Rfe8) 0.48/13 1212} 14. Qc2 {0.59/16 920} 14... f5 {(Bd6) 0.55/13 862} 15. Ng3
>{0.63/15 1015} 15... Kh8 {0.54/13 1192} 16. Ne2 {(e4) 0.69/15 1083} 16... g5 {
>(Bd6) 0.35/13 1152} 17. Ne5 {(Nc3) 0.69/15 1525} 17... Nxe5 {0.23/13 830} 18.
>dxe5 {0.72/14 197} 18... c5 {(Rad8) 0.23/13 1079} 19. Nc3 {0.59/15 1359} 19...
>Rad8 {(Rfd8) 0.19/13 1378} 20. Qb3 {0.47/15 1026} 20... Bxc3 {
>(Bc6) 0.21/14 2066} 21. Qxc3 {(bxc3) 0.56/16 230} 21... b5 {0.17/14 843} 22.
>Be2 {(Bb3) 0.50/17 1073} 22... b4 {0.24/14 1130} 23. Qc2 {(Qc4) 0.53/16 673}
>23... Bc6 {0.27/13 731} 24. Rac1 {(Rdc1) 0.53/15 1031} 24... Rxd1+ {0.25/13 845
>} 25. Rxd1 {0.53/15 169} 25... Bd5 {0.25/13 817} 26. f3 {0.44/16 1259} 26...
>Qc7 {(c4) 0.19/12 1099} 27. e4 {(b3) 0.88/17 759} 27... fxe4 {0.55/14 1130} 28.
>fxe4 {0.78/17 587} 28... Bxa2 {0.55/13 489} 29. b3 {(Qa4) 0.88/18 884} 29...
>Bxb3 {0.32/14 918} 30. Qxb3 {0.78/16 377} 30... Qxe5 {0.15/14 669} 31. Qc4 {
>(Qh3) 0.66/15 498} 31... Kg7 {(Rb8) 0.26/12 845} 32. Bg4 {(g3) 1.06/16 1457}
>32... Rf4 {(Rf6) 0.00/12 508} 33. Qb5 {(Rd7+) 1.34/15 703} 33... Qxe4 {
>0.10/12 789} 34. Bh5 {1.25/15 654} 34... Qe3+ {(Rf8) 0.10/12 575} 35. Kh1 {
>0.91/9 3} 35... Rd4 {0.13/12 556} 36. Qb7+ {(Ra1) 1.66/13 247} 36... Kf6 {
>0.00/11 160} 37. Qf7+ {(Rf1+) 1.69/13 356} 37... Ke5 {0.00/6 0} 38. Qc7+ {
>1.66/12 264} 38... Kd5 {0.00/12 399} 39. Rg1 {(Qb7+) 1.75/13 370} 39... g4 {
>0.00/12 519} 40. Qxa7 {1.88/13 724}
This is what happes when a knowledge based program like Fritz meets a search
based program like Hiarcs.
You can see that the searched based program knows nothing about the position and
evaluates it as 0.00 at moves 36-39 when the knowledge based program see small
improvements in the score 1.66 at move 38, 1.75 at move 39 1.88 at move 40 and
the small improvement continues in the next moves.
40... Re4 {(Rf4) 0.33/11 752} 41. Be8 {
>2.00/13 492} 41... g3 {(Qf2) 1.49/12 2979}
Hiarcs found a fail low of more than 1 pawn when Fritz only continues it's small
score improvements.
42. hxg3 {2.22/12 204} 42... Rg4 {
>1.57/11 495} 43. Rd1+ {2.78/13 650} 43... Rd4 {4.88/13 3153}
Here Hiarcs found a big fail low when Fritz need some more plies.
44. Rf1 {
>3.22/13 409} 44... h5 {5.09/10 332} 45. Qb7+ {(Qd7+) 3.97/12 278} 45... Ke5 {
>6.03/12 1480} 46. Qc7+ {2.59/7 389}
only 7 plies in 389 seconds?
It seems that something was broken in the hardware of Fritz(it could calculate
12 plies in the previous move in less time) but it does not help Hiarcs.
Uri
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.