Author: Christophe Theron
Date: 00:33:33 05/31/98
Go up one level in this thread
On May 30, 1998 at 11:21:50, Thorsten Czub wrote:
>The third game is interesting because of the endgame and the dramatical
>situation this endgame had for the 2
>operators. Ulf Lorenz operated his parallel machine P.CoNNerS, a program
>using Parallel Controlled Conspiracy
>Number Search. The program uses certain CONSPIRACY TARGETS and extends
>them.
>This program was able to search very deep, and if there would be
>something to calculate, it would better handle
>this. In theory, Conners would be able to see a silent move although it
>is 30 plies diep - if this is a target. So in fact
>Tiger (on a p2/266 Mhz) had a tough opponent. Ulf said he uses 24
>machines. I don't know exactly WHICH kind
>of machines.
ICCA journal says: 24 Pentiums.
>[Event "7th IPCCC"]
>[Site "Paderborn(Germany)"]
>[Date "12.2.98"]
>[Round "3"]
>[White "P.CoNNerS"]
>[Black "Chess Tiger"]
>[Result "0-1"]
>[ECO ""]
>[Opening ""]
>
>1.e4 e5 2.Nf3 Nf6 3.d4 Nxe4 4.Bd3 d5 5.Nxe5 Nd7 6.Qe2 Nxe5 {here Ctiger
>had to think for the first time} 7.Bxe4
>dxe4 8.Qxe4 Bf5
>9.Qxe5+ {out of book for Conners} Qe7 10.c3 Qxe5+ 11.dxe5 Bc5 12.Nd2
>0-0-0 13.Nb3 Bb6 14.0-0 Rhe8 15.Bf4
>h6
>16.Rad1 Be6 17.h4 a5 18.Rfe1 a4 19.Nc1 Bc5 20.b3 Be7 21.Bg3 Ba3 {this
>was outplayed with a very optimistic
>+0.58} 22.c4 Bb4 {+0.57}23.Rf1 Bd2 {+0.66}
>24.Ne2 g5 {+0,68} 25.f3 {with this move the game turns arround. Now
>black is really a little in advantage} Bf5
>{+0,66} 26.Be1 Be3+ {+0.58} 27.Bf2 Rxd1 {+0,62} 28.Rxd1 Bc2 {+0,32}
>29.Ra1 Rxe5 {+0,32} 30.Bxe3 axb3
>{0.14}
>31.axb3 Rxe3 {+0.16}32.Ra8+ Kd7 33.Nd4 Bg6 {+0.16} 34.hxg5 Rd3
>{+0.48}35.Nb5 c6 {+0.58}36.gxh6 cxb5
>{+1.23} 37.Ra7 Kc7 {+1.50}
>38.cxb5 Rxb3 {+2,01} 39.Kf2 {i have to explain that Conners was in time
>trouble. I don't remember where exactly
>the time problems occured. I explained ulf that i will not allow conners
>to lose on time, i would - before his time
>runs out, offer him a draw, since i don't think it is fair to misuse
>someones trouble into such a way. But on the other
>hand i asked the other programmers if they would continue such a game,
>and if this would be fair. Since all of them
>agreed that they would continue the game if it would be their program, i
>did so. But - i want to make sure that I
>would have offered ulf a draw if it would have been nessesary. } Rb2+
>{+2,12} 40.Kg1 Rxb5 {+2,17} 41.g4 Kd6
>{here we joked about databases/tablebases. And ulf said he was to lazy
>to implement tablebases before the
>tournament. He was so busy in doing more important things. It is obvious
>that murphy listened to this...} 42.Kh2 f5
>43.Kg3 fxg4 44.fxg4 Ke6
>45.Ra4 Kf6 {+1,82} 46.Kh4 Bh7 {+1,76} 47.Rd4 Rb1 {+1,78} 48.Rd7 Rh1+
>{+3,47} 49.Kg3 Bb1 {+3,47} 50.g5+
>Kxg5 {+4,92}51.Rxb7 Bg6
>52.h7 Bxh7 {and there it is, the technical endgame ! Can you imagine
>what happened now. Since Stefan Meyer
>Kahlen was there, the author of Shredder, I asked him if he would like
>to install Shredder2 on a PC next to our
>desk. I remembered how Frederic Friedel called ChessBase/Hamburg in a
>similar situation between Genius-Mchess
>in munich 1993. He called hamburg and asked about the chances to win or
>lose or draw this endgame. Now - since
>technology has made progress, we only had to install shredder2 on a
>machine and it told us exactly and live about
>how good the moves are !} 53.Kf2 {it must have been a nightmare. Anybody
>counted for the 50th move time
>control, since the last capture move was 52nd move, black would have
>time until move 102 to get the rook or mate}
>Kf4 54.Ke2 {can you imagine how ulf must have felt. Still conners was in
>time trouble. Again i was not sure what
>to do. Should i continue the game or offer a draw. I tried to find out
>what Christophe Theron would say if i would
>have told him: i offered a draw since i felt for ulf.
I would have killed you Thorsten. You can have concerns when your
opponent runs on a much slower machine and is short on time, but NOT
when your opponent runs on 24 processors and you have only one!!!
>So - i continued
>the game. But i again told ulf that when we see
>that conners would run out of time i would offer a draw a few seconds
>before the clock runs out of time. I am sure
>this did not made him happier, but ... i thought it would only be fair.
>} Rh2+ 55.Ke1 {would tiger be able to get the
>rook ? Fast enough ? the cheerleaders arround the board told us which
>moves have to be played, and how many
>moves until the rook is lost. It went up and down. There was only a
>chance for tiger to win by chance - avoiding the
>3time repetition or by blunder of Conners} Ke3 {the white king is so
>near at the edge, isn't there a solution for
>black ?!?} 56.Rf7 Bd3 57.Re7+ Be4 58.Rf7 Ra2
>59.Rd7 Rg2 60.Rf7 Rb2 61.Rd7 Re2+ 62.Kd1 Rg2 63.Kc1 Rc2+ 64.Kd1 Rc6
>65.Ke1 Rc1+
>66.Rd1 Rc2 67.Rd7 Rb2 68.Rd8 Rg2 69.Rf8 Rc2 70.Rd8 Rh2 71.Rf8 Re2+
>72.Kd1 Rd2+
>73.Ke1 Ra2 74.Rd8 Re2+ 75.Kd1 Rg2 76.Kc1 Rc2+ 77.Kd1 Rc3 78.Ke1 Bg6
>79.Kf1 Rc2
>80.Kg1 Be4 81.Rf8 Rg2+ 82.Kf1 Rb2 83.Kg1 Rc2 84.Rf7 Rg2+ 85.Kf1 Rg5
>86.Ke1 Rc5
>87.Rd7 Rh5 88.Rf7 Ra5 89.Rd7 Ra1+ 90.Rd1 Ra3 91.Kf1 Ra5 92.Re1+ Kf3
>93.Rc1 Bd3+ {mate in 9}
>94.Kg1 Rg5+ {mate6} 95.Kh2 Kf2 {mate4}96.Kh3 0-1
>
>In the end P.Conners made a mistake that made it possible for black to
>win shortly before the 50 move rule.
>Unbelievable. It was horrible. I left so much sweat, and Ulf also, that
>we both lost 4 kg or more weight.
>But there was enough coke and in the evenings we ate roast potatoes with
>eggs , called bratkartoffeln in a famous
>pub called "lötlampe". I am sure we compensated the weigth-loss to
>anybodies confidence.
>
>So Tiger had 2.5 from 3 and my promise to win this tournament for
>Christophe was still possible.
>Nimzo98 was maybe the strongest program / opponent in this competition.
>It made w whitewash in the netherlands
>champioships. And I was sure somebody would have to stop nimzo,
>otherwise nimzo will win almost any game.
>But it was still to early. And maybe nobody really seriously thought
>that Tiger would have chances to win the
>tournament, despite myself. A reason was that I played manually
>tournament games before Paderborn. And my
>results were quite ok-ish !
>
>My old chart said the following bean-counted results:
>
>Chess Tiger 11.2 against
>
>Opponent ELO Time Control Results
> win draw loss
>Hiarcs6 2532 40/120 2 0 1
>Nimzo98 2530 40/120 2 2 0
>Junior 4.6 ? 40/120 2 1 1
>Rebel9 2533 40/120 + 60/60 1 1 1
>Mchess7 2516 40/120 + 60/60 2 1 1
>Shredder2 2503 40/120 2 1 1
>Fritz5 ? 40/120 0 0 3
>CM5000 >2387 40/120+60/60 1 1 1
>Genius5 2498 40/120 0 2 1
>Virtual2 ? 40/120 2 2 0
>-----------------------------------------------------------------
>35 games: 19.5-15.5 against ELO 2522 average 14 11 10
>
>Ok - obviously Tiger had a problem with Fritz5 ! I don't think this had
>something to do with playing strengh since
>it has had no problems with nimzo and not with other strong programs.
>But fritz5 did not participated in paderborn.
>
>BTW: the elo-numbers came from ssdf-list
>from november or at this time I played the games. I was very sure that
>Tiger had chances to win the tournament.
>Not only because of the results, but even more because the games were
>very good. It is still a question to me how
>Christophe was able to program that many progress into his engine from
>the waterloo-result of tiger in paris to
>december, january. Strange.
I felt so ridiculous and angry in Paris that I HAD to make some real
progress. I guess working on the endgames has made Tiger a lot stronger.
But remember that Tiger 11 (32 bits) is born in June 1997. So in Paris I
had a 4 month old baby-Tiger.
>If he would have participated with tiger
>11.2 in PARIS, he would have had chances to
>win there too. What a pity the progress came in the engine AFTER paris.
I would have been glad just to be in the top 10 in Paris.
(Well... at least I made it in the blitz tournament!)
Christophe
>In the next round I had to play against Nimzo. I was in no way afraid.
>My games before paderborn with the
>commercial version of Nimzo98 were ok. Of course Chrilly came with a new
>monster, the paderborn version,
>another progress for nimzo. But - i was not afraid.
>
>So - to be continued with the fourth round, ChessTiger 11.2 vs.
>Nimzo-Paderborn...
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.