Author: Dirk Frickenschmidt
Date: 09:56:18 09/18/98
Go up one level in this thread
On September 18, 1998 at 05:27:44, Christophe Theron wrote: Hi Christophe, >On September 17, 1998 at 16:54:07, Dirk Frickenschmidt wrote: > >>On September 17, 1998 at 06:29:50, Christophe Theron wrote: >> >>>On September 17, 1998 at 06:12:56, Dirk Frickenschmidt wrote: >>> >>>>Here's one of the test positions where I was surprised about Junior5's tactical >>>>strength. >>>> >>>>Fritz5 blundered this position standing well but accepting an exchange sacrifice >>>>leading to a black mate attack. And Fritz normally does *very* well in finding >>>>mate threats. Other programs like Mchess 7.1 and Rebel9 also have problems with >>>>this positions, as is mentioned below. I guess this could count for most >>>>programs - with the possible exception of de Konings programs guessed again, >>>>perhaps also some Kittinger prg - , so please test and show me I was wrong... >>>>;-) >>>> >>>>Junior5 had the best time of soving this problem by far from the four programs I >>>>tested so far. >>>> >>>>Fritz5 - Golubev,M [E98] >>>>Sparkassen open rapid bad tactical move >>>>[DF] >>>> >>>>1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.Be2 e5 7.0-0 Nc6 8.d5 Ne7 9.Ne1 >>>>c5 10.dxc6 bxc6 11.b4 d5 12.Bg5 d4 13.Bxf6 Bxf6 14.Na4 Bg7 15.Nc5 h5 16.f4 exf4 >>>>17.Ned3 h4 18.Rxf4 a5 19.b5 cxb5 20.cxb5 a4 21.Rb1 Bh6 22.Rf1 Be3+ 23.Kh1 f5 >>>>24.Bf3 fxe4 25.Bxe4 Nf5 26.Bxa8?? [This is the Fritz5 blunder losing the game. >>>>Mchess 7.1 sees problems in ply 8 after about 5' and begins dropping from +5.02 >>>>down to +2.10, considering g4 as alternative and finally switching to ot after >>>>about 15'. Rebel9 smells "something" after 1'17, dropping from +6.23 to 4.09 in >>>>ply 9, but needs ply 11 and about 20' to evaluate even less with +0.63, still >>>>not finding another move within 30'.} [But Junior5 sees the mess already after >>>>1'51 on 200MMX-32Mb Hash and switches to 26.g4 with a white winning position] >>>>26...Ng3+ 27.hxg3 hxg3 28.Rxf8+ Qxf8 29.Bd5+ Kh8 30.Qh5+ 0-1 >>>> >>>>Kind regards >>>>from Dirk >>>> >>>> >>>>P.S. I was not able to continue my Junior5 tournament testgames for some days >>>>for a - perhaps from your view - funny reason: >>>>My parents in law visit us for some days and the second computer, being used for >>>>non-chess purposes (my two sons playing and one of them beginning to program) >>>>during the day, is in the room where they sleep. >>>>Of course their healthy sleep, undisturbed by computer cooler sounds, is *much* >>>>more important to me than any game of chess ever will be! :-) >>>> >>>>So I ask for your understanding and will continue the tournament with the first >>>>game against Fritz5 next night probably. >>> >>>Could you please post the position in FEN format? >>> >>> >>> Christophe >> >>Hi Christophe, >> >>this is my first try to copy FEN within CB7. Hope it works. This is the position >>before the wrong Bxa8 (g4 is better, winning the game). >> >>r1bq1rk1/8/6p1/1PN2n2/p2pB2p/3Nb3/P5PP/1R1Q1R1K w - - 0 26 >> >> >>Kind regards >>from Dirk > >Yes it works. Posting the positions in FEN (together with the game if you want) >is better because some of us are not able to read PGN files with their programs >(it is my case). I will try to have that in mind when I post a position next time. Many programs support pgn, but some not, and I very well understand that other things are more important for you while developing a program. Yet I think implementation of pgn could be useful for yourself, because so much useful stuff is public domain in pgn (like on the pitt server). >Tiger doesn't solve the position in a reasonnable time I fear. On my P200+ 8Mb >hash it took 1364s (almost 23 minutes) to find g4. > >After Bxa8, it finds that Ng3+! is good for black in 8s, but then it takes 640s >(10mn 40s) to find b6, then 1364s to find g4. > B6 may work as well for winning. I haven't checked that yet, though Fritz also wants to play this move after finding the trouble with Bxa8. It's really a difficult position, some ply including a 'silent' move, to find the mating danger. But perhaps nice for improving programs? > Christophe Kind regards from Dirk
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.