Author: Jeroen van Dorp
Date: 15:07:00 10/16/03
Go up one level in this thread
On October 16, 2003 at 17:15:28, swaminathan natarajan wrote: >Thanks! >WoW it is human like program just like Delfi but stylish than latter >is it possible to load that rebel file in arena? >Kudos to Ed Rebel is a winboard compatible engine, and comes with a winboard compatible interface, ChessPartner 5.3 and with a ERT-enabled tournament manager for those who think these tournaments are interesting and fun. I only can add to Stephen that I think he's right. Not being a strong chess player, I can't show you typical examples of play where the "regulars" lose their focus and Rebel doesn't, because I'm neither strong enough to challenge each of the batches. But playing Rebel gives a strange human feel to the game. I see a logical move build up. It might be that sometimes Rebel misses a deep tactical combination Fritz will find, but OTOH in human play such deep tactical combinations are not that interesting because there's hardly someone around - except for the world top players - having the ability to calculate on par. I only have a few examples where I tried out Fritz 8 and Rebel 12 when I bought the latter for a renewed hello, and indeed Rebel was able to identify positional characteristics much faster, or identify it while Fritz 8 didn't identify them at all: [D]rn1q1rk1/pbp2ppp/1p2pn2/8/1bBP4/2N1P3/PP2NPPP/R1BQ1RK1 w - - 0 9 Here f3 starts limiting the mobility of the b7 bishop, quite important for a solid defense or counter attack. Fritz 8 changes opinion quite often before settling to f3 as the top candidate. 9.f3 Nbd7 10.Qc2 Bd6 11.Nf4 = (0.25) Depth: 3/14 00:00:00 = (0.13) Depth: 5/17 00:00:01 8kN 9.Qd3 c5 10.a3 Bxc3 11.Qxc3 cxd4 12.Nxd4 Nbd7 = (0.16) Depth: 5/17 00:00:02 16kN = (0.16) Depth: 6/19 00:00:06 51kN 9.Qc2 c5 10.dxc5 Bxc5 11.e4 Nbd7 12.Bf4 Rc8 = (0.19) Depth: 6/19 00:00:07 65kN = (0.03) Depth: 7/19 00:00:10 109kN 9.Qd3 a6 10.Rd1 b5 11.Bb3 Bd6 12.d5 b4 = (0.06) Depth: 7/20 00:00:10 126kN = (0.19) Depth: 7/22 00:00:13 200kN 9.f3 c5 10.Qc2 cxd4 11.Rd1 Nc6 12.Nb5 Rc8 13.Nexd4 Nxd4 = (0.22) Depth: 7/22 00:00:14 250kN = (0.03) Depth: 9/27 00:00:19 1341kN 9.Qd3 a6 10.a3 = (0.06) Depth: 9/27 00:00:19 1534kN = (-0.06) Depth: 10/34 00:00:22 3668kN 9.Qa4 = (-0.03) Depth: 10/34 00:00:23 4304kN 9.Qb3 = (0.00) Depth: 10/34 00:00:25 4945kN 9.Qc2 = (0.03) Depth: 10/34 00:00:26 5881kN 9.f3 c5 10.a3 Bxc3 11.bxc3 Qc7 12.Bd3 Ba6 13.Bxa6 Nxa6 14.e4 cxd4 15.cxd4 = (0.06) Depth: 10/34 00:00:27 6707kN = (0.22) Depth: 12/38 00:00:51 23940kN Rebel goes straight to f3 and stays with its assessment: 00:00:00.4 -3,22 1 9 Bxe6 fxe6 00:00:00.4 -2,99 1 23 Nd5 Bxd5 00:00:00.4 0,13 1 25 Bd2 00:00:00.4 0,13 1 83 f3 00:00:00.4 0,44 2 190 f3 Qd6 00:00:00.4 0,65 3 1220 f3 c5 e4 00:00:00.5 0,56 4 3006 f3 c5 e4 Nc6 00:00:00.6 0,70 5 25469 f3 Nh5 a3 Bd6 00:00:00.8 0,58 6 88250 f3 Nc6 b3 Qd6 Bb2 00:00:01.1 0,63 7 232524 f3 c5 a3 Ba5 e4 cxd4 Qxd4 Nbd7 00:00:01.7 0,65 8 522112 f3 c5 a3 Bxc3 bxc3 Qc7 e4 cxd4 Qxd4 Rd8 00:00:02.9 0,62 9 1156184 f3 c5 a3 Bxc3 Nxc3 cxd4 Qxd4 Qxd4 exd4 Nc6 00:00:04.2 0,62 10 1855976 f3 c5 a3 Bxc3 Nxc3 cxd4 Qxd4 Qxd4 exd4 Nc6 00:00:08.0 0,59 11 4038431 f3 c5 a3 Bxc3 Nxc3 cxd4 Qxd4 Qxd4 exd4 Nc6 Be3 Rad8 Rad1 00:00:43.0 0,68 12 23560240 f3 Be7 e4 c5 00:01:06.8 0,60 13 37152941 f3 c5 a3 cxd4 axb4 dxc3 Qc2 Qc7 Bd3 Nc6 bxc3 Rad8 Rd1 00:01:48.0 0,62 14 61318622 f3 c5 a3 cxd4 axb4 dxc3 bxc3 Qc7 Bd3 Nc6 Qb3 Here's another: [D]rn1qkb1r/pp3ppp/2p1pn2/3p4/3PP3/2N2Q1P/PPP2PP1/R1B1KB1R w KQkq - 0 7 Bd3 leads to an efficient pawn sacrifice to create an open d-rank. Fritz 8 looks around and sticks with an equal assessment: 7.Bg5 h6 8.Bxf6 Qxf6 9.Qxf6 gxf6 10.a3 dxe4 11.Nxe4 f5 12.Nd2 Rg8 13.b4 Bg7 14.c3 ² (0.44) Depth: 7/22 00:00:00 175kN = (0.25) Depth: 13/36 00:00:53 40586kN Rebel identifies the possibilities for white and choses Bd3 with a far better evaluation: 00:00:00.4 0,41 1 16 exd5 cxd5 00:00:00.4 0,60 1 37 Bd3 00:00:00.4 0,78 1 49 Bg5 00:00:00.4 0,37 2 405 Bg5 Bb4 00:00:00.4 0,37 2 554 Bd3 00:00:00.4 0,97 2 677 Bd3 Bb4 00:00:00.5 0,97 3 1762 Bd3 Bb4 e5 00:00:00.7 0,88 4 4985 Bd3 dxe4 Nxe4 Qxd4 c3 00:00:00.8 0,78 5 12670 Bd3 dxe4 Nxe4 Qxd4 c3 Qe5 00:00:00.9 1,15 6 86029 Bd3 dxe4 Nxe4 Qxd4 c3 Qe5 Bf4 00:00:01.2 1,19 7 297627 Bd3 dxe4 Nxe4 Qxd4 c3 Qd8 Nxf6+ gxf6 Bf4 00:00:01.5 1,00 8 447293 Bd3 dxe4 Nxe4 Qxd4 c3 Qd5 Bf4 Nxe4 Qxe4 Qh5 00:00:02.2 1,05 9 839604 Bd3 dxe4 Nxe4 Qxd4 c3 Qd5 Nxf6+ gxf6 Be4 Qc4 Bf4 00:00:03.1 1,05 10 1377565 Bd3 dxe4 Nxe4 Qxd4 c3 Qd5 Nxf6+ gxf6 Be4 Qc4 Bf4 00:00:04.9 0,98 11 2447105 Bd3 dxe4 Nxe4 Qxd4 c3 Qd5 O-O Nbd7 Rd1 Nxe4 Bxe4 Qb5 00:00:30.5 1,01 12 18268304 Bd3 Qb6 exd5 cxd5 Ne2 Nc6 c3 00:00:53.0 1,00 13 32435558 Bd3 Qb6 O-O Qxd4 exd5 Nxd5 Be3 Qe5 Nxd5 cxd5 Bf4 Qf6 Rae1 Nc6 IIRC on the old Rebel homepage there also were a few positions with better positional assessment. If you look for an analysis engine, I guess Rebel shouldn't be your top choice, and if you're out to play zillions of engine-engine games, maybe you should pass by, but if you're interested in a nice game of chess against a feel-good engine, Rebel is your first pick. I prefer it over Hiarcs 8, and value it certainly more than (also fun program) CS Tal II, which has a group of staunch supporters/believers. The CP interface is stable, simple, for 95% intuitive, and can link you up to the Internet as well. It connects flawlessly to my DGT-board. And it gives you a swell interface for your Winboard and UCI engines. J.
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.