Author: Vincent Diepeveen
Date: 23:25:43 06/17/01
Go up one level in this thread
On June 17, 2001 at 15:25:02, Uri Blass wrote: >On June 17, 2001 at 14:35:53, Vincent Diepeveen wrote: ><snipped> >>Junior is only searching the tactical search space as it positionally >>will of course never be in the same league as todays programs. >> >>I'm amazed it's still scoring so well. Of course junior is tactical >>one of the greatest as it is focussing on seeing tactical lines very >>deeply. > >I do not believe in it. >I believe that in most of the cases Fritz is faster than Junior in seeing >tactics. All tactical tricks i tried Junior finds faster as Fritz. >It is easy to check it by test positions and I do not talk about known test >positions that programs can be tuned for them but about test positions from >games that were played after both Junior and deep Fritz were released. >It is possible to find cases when programs outsearched the opponent in the last >ICC tournament(When Fritz and Ferret won). >It is possible to see if junior is better in tactcis by testing both Junior6 and >Deep Fritz on these positions. This is exactly what i said about junior, it lacks positional depth. If you search positional lines only a ply or 5, then that's in quiet positions not very rewarding. > >> >>This whole junior system will fall apart as soon as you try to improve >>its knowledge. The reason for that is that one needs to do more near >>the leafs then. So not using a qsearch is out of the question then. >>So are very dubious ways of pruning. >> >>But in chess there is only one truth when talking about score, and >>that's the lemma that the weakest chain of a program decides whether >>you win or lose. So far junior team did a very good job in not having >>weak chains, but IMHO the positional chains are already its weakest >>chain quite some time. > >I do not believe in it. >I do not believe that Junior is positionally weaker than other top programs like >Deep Fritz or Tiger. You use 2 hell of an examples by the way. A program which is still partly preprocessor (tiger) and a program with an evaluation which is so small that it fits inside any L1 cache (fritz). How about comparing with the other loads of programs that exist on this planet. Let's take Deep Shredder or shredder5. >Here is a game that demonstrates Junior's superiority against Tiger when the >time control is 13.5 hours per game on PIII500 > >This game is based on a correspondence game of my friend moshe glazman. >He analyzed a position from his correspondence game by engine-engine games I happen to receive daily loads of correspondence things, and in 99% of all analysis the correspondence players are analysing anti-positional moves which might get played based upon extreme accurate tactics. If i sometimes carefully ask them then: "but why not playing a positional more sound plan, don't you win all of your games a lot easier then?" Then a part of them is bigtime insulted, though i didn't mean it like that then. Most can't even explain why they play a certain move, but well: "all the programs agreed on big search depths!" >He says that he regret for buying tiger because it is stupid and this game is >the proof for it(tiger lost also with white from the same position) >Tiger14 lost from more than +4.37 pawns advantage against Junior6 and moshe >glazman claims that he never saw a chess program doing things like it in the >past. >Unfortunately moshe's opponent was smarter than tiger and the game was drawn >Do you think that Tiger lost at this time control because of tactics? >I think that Tiger simply overevaluated the passed pawns. I wrote a long thing and you completely deleted what i said about about todays high bluff scores and what i have said about tiger in a load of emails. - queen near opponent king gets +2.0 pawns or so - passers on 6th + 7th rank sometimes get +3.0 each I'm not going to repeat what i said about tiger, but you removed those crucial few lines from your quote!!!!! >I posted this game some weeks ago but nobody replied(probably it was hidden by >other posts and most people did not read it. > >Junior 6.0 - Chess Tiger 14.0 [B78] >ACER_6100, 240'/40+360'/20+210' Ramat-Gan (1), 20.04.2001 > >W=16.9 ply; 523kN/s; 76,611 TBAs >B=13.3 ply; 127kN/s; 1 TBAs >1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6 6.Be3 Bg7 7.f3 0-0 8.Qd2 Nc6 >9.Bc4 Bd7 10.0-0-0 Rc8 11.Bb3 Ne5 12.h4 Nc4 13.Bxc4 Rxc4 14.h5 Nxh5 15.g4 >Nf6 16.Nb3 Re8 17.e5 Nxg4 18.fxg4 Bxg4 19.Rdg1 dxe5 20.Qh2 h5 21.Nd2 Rxc3 >22.bxc3 Be6 23.Nb3 Qc7 24.Qd2 Rd8 25.Qe1 Rc8 26.Bd2 a5 27.Rh4 b5 28.Kb1 Qd6 >29.Qe3 a4 30.Nc1 Rc4 31.Rxc4 Bxc4 32.Nd3 f5 33.Nb4 e4 34.a3 Kh7 35.Kb2 e5 >36.Bc1 Qc7 37.Rd1 Qb7 38.Qg5 f4 39.Qd8 Be6 40.Rd6 Qf7 41.Rc6 Kh6 42.Bd2 e3 >43.Be1 Qd7 44.Qb6 Bf5 45.Bh4 Be4 46.Rc5 Qf5 47.Rc7 Qh3 48.Bd8 e2 49.Qc5 Qf5 >50.Bh4 g5 51.Be1 h4 52.Qxb5 f3 53.Qb6+ Bf6 54.Rc4 Kg6 55.Qe3 h3 56.Qf2 Ba8 >57.Rc5 Qe6 58.Ra5 Bb7 59.Rb5 Qd7 60.Rb6 Qc7 61.Rxf6+ Kxf6 Both last book >move 62.Qg3 -0.85/18 47:49 Qd7 -4.37/16 1:45:26 63.Nd3 -0.85/18 >24:18 e4 (Qf5) -4.18/15 34:38 64.Nf2 -0.26/17 28:19 Qb5+ -4.02/15 >35:08 65.Kc1 0.00/18 1:00:5 Qc5 -3.26/15 2:56:19 66.Qxh3 0.00/17 >35:38 Qxa3+ (e3) -1.71/15 2:07:20 67.Kb1 -0.10/17 21:24 Qe7 (Qd6) >-0.55/15 51:47 68.Ng4+ 0.31/17 1:18:22 Kf5 (Ke6) -1.88/13 6:50 >69.Bg3 (Ne3+) 0.48/17 18:18 69...Qg7 -0.18/13 11:34 70.Ne3+ 0.08/17 >29:58 Kg6 -0.05/14 12:50 71.Qf5+ 0.25/17 18:24 Kh5 -0.03/18 12:11 >72.Qe6 (Qh3+) 0.05/17 15:10 72...Qf8 -0.22/15 29:44 73.c4 (Qh3+) >0.15/17 14:06 73...Bc8 -2.13/14 9:42 74.Qxe4 0.15/17 7:51 Qf7 >-1.89/14 31:29 75.Nd5 (Bf2) 0.16/17 18:21 75...Kh6 -2.69/12 11:30 >76.Qd4 (Qe3) 0.21/16 16:41 76...Be6 (Bf5) -2.41/12 8:19 77.Be1 (Bf2) >0.28/16 9:51 77...Qg7 (Bf5) -2.56/12 6:36 78.Qb6 0.28/15 6:57 Qd7 >(Qg6) -0.38/15 37:25 79.Kb2 0.54/16 22:08 Kg6 (Qg7+) -0.22/15 28:04 >80.Ka3 (Nc7) 0.74/16 16:21 80...Kf7 (Qf7) -1.68/13 15:35 81.Qd4 >0.76/16 9:50 Qd6+ (Bxd5) -0.80/12 15:56 82.Kxa4 1.07/17 28:54 Qd7+ >-0.18/13 11:00 83.Kb4 (Ka3) 1.51/16 18:52 83...Qb7+ (g4) -0.04/12 >1:37 84.Kc3 1.54/17 15:37 Kg8 (Bxd5) 1.84/14 8:00 85.Qe3 (Kd2) >1.97/14 4:47 85...Qg7+ 1.28/11 37 86.Kd2 (Kd3) 2.25/17 8:22 >86...Bxd5 3.02/12 59 87.cxd5 2.30/14 1:50 Qf6 3.44/12 55 88.Bg3 >(Qe6+) 2.35/17 13:02 88...Qf5 (Kh7) 3.58/11 58 89.d6 2.65/15 12:58 >Qa5+ 4.14/10 5:34 90.Kd3 3.33/15 24:08 Qd5+ (Qb5+) 4.34/10 1:00 >91.Qd4 (Kc3) 3.52/15 4:29 91...Qf5+ (Qxd4+) 6.18/12 2:35 92.Qe4 >4.32/15 4:32 e1Q (Qf7) 6.44/10 15 93.Bxe1 (Qxf5) 4.75/14 2:52 >93...Qd7 6.97/10 23 94.Qg6+ (Qd5+) 5.38/15 5:16 94...Kf8 9.02/11 18 >95.Ke3 (c4) 5.56/15 4:48 95...Qa7+ (f2) 8.00/9 15 96.Kxf3 5.76/14 >4:22 Qd7 7.72/9 13 1-0 > >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.