Author: Albert Silver
Date: 18:23:50 10/07/00
Go up one level in this thread
On October 07, 2000 at 10:50:11, Vincent Diepeveen wrote:
>Hello here an attacking position for tiger:
>
>[D]2rr2k1/pp3pp1/4b3/2qNp1b1/4P3/1B1Q4/PPP5/1K3R1R w - -
>Qf3!! Diepeveen - H.H. Hagen, corr. game
>
>Qf3 is a simple tactical win at the king side. Diep has no probs
>with it and after a few hours score goes even up to nearly 2 pawns
>for Qf3, thereby being pawns better as other moves.
Simple tactical win? Are you sure? It's possible this wins, but it is hardly
simple. Your opponent helped you greatly by playing Bxd5 allowing the very nice
shot Bxd5 (kudos for seeing it of course). What would you have played after
1.Qf3 Rc6 2.Qh5 Bh6 3.Qxe5 Rdc8 instead of 3...Bxd5? I didn't see any killer
blows, but perhaps I didn't look deep enough.
Still, you wished to see Tiger Gambit's attacking ability, so here is an
impressive game played yesterday on the server. BTW, I should be fair in
pointing out that there are plenty of counter examples of it exaggerating and
then losing, but the game does show the enormous promise of Gambit IMO. Frankly,
when you look at it, it's hard to believe White was played by a program.
Time control was 40 min/KO (no increment)
[Event "?"]
[Site "?"]
[Date "7/10/2000"]
[Round "?"]
[White "SubtleOne"]
[Black "UltraMaster"]
[Result "1-0"]
[WhiteElo "2530"]
[BlackElo "2474"]
[Opening "E15 Queen's Indian: Nimzovich Variation"]
1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.g3 Ba6 5.Nbd2 Bb7 6.Bg2 c5 7.e4 cxd4 8.O-O d6 9.Nxd4
Qc7 10.b3 a6 11.Bb2 Be7 12.f4 Nc6 13.Nc2 O-O 14.g4 Nd7 15.g5 Nc5 16.Rf3 e5
17.Rh3 Bc8 18.f5 Bxg5 19.Nb1 Ne7 20.Nc3 Bf4 21.Qh5 h6 22.Nb4 Bb7 23.f6 gxf6
24.Nbd5 Nxd5 25.Nxd5 Bxd5 26.exd5 Nd7 27.Rf1 Rfe8 28.Rxf4 exf4 29.Qxh6 Qc5+
30.Kf1 Re1+ 31.Kxe1 Qg1+ 32.Kd2 Qxg2+ 33.Kc3 Qg7 34.Qxf4 Ne5 35.Rg3 Ng6 36.Qxd6
Re8 37.Kd3 Qh8 38.Qc6 Rf8 39.Qxf6 Qxf6 40.Bxf6 Re8 41.d6 Re6 42.Be7 Kh7 43.Rxg6
Rxg6 44.d7 {UltraMaster resigns} 1-0
An impressive game for a program.
On the other hand, it is hardly infallible, and in a 15min/KO game after saccing
the bishop on h7 (so White is down a piece for the moment), reached the
following position:
[D]1rb2r2/p2n1pp1/4p1k1/1p1pP1N1/1nq2P2/2N5/PPP3PP/2KRQ2R w - -
On a fluke of intuition, I found a stronger move than the one played by Gambit.
The programs I tested it on take a long time to see it, so it probably makes a
good test position. To leave you the pleasure of finding it, I placed the answer
below in a P.S.
Albert
P.S. Nope, nope, the answer is a little below...
P.P.S. Gambit played the normal looking 1.Qg3 and did go on to win, but it
missed a stronger, more spectacular win with 1.Rd4!!!!! Not only is White down a
piece but now offers a rook. Amazing. :-)
>
>Qg3 is a draw, so is Qe2. You win a pawn with both Qg3 and Qe2 at g7,
>but you keep left with opposite bishops and blacks king safety is ok
>with king on e7.
>
>Qf3 is a simple tactical blow for example game went like:
> Qf3 Rc6 Qh5 Bh6 Qxe5 Bxd5 Bxd5!! Qxc2+ Ka1 Rc7 a3! b5 Rxh6 and my opponent
>resigned a move later (note there are several different wins after gxh6)
>
>I think next programs for sure have no problems finding
>Qf3:
> Diep, The King, LambChop, ZarkovX, Crafty
>
>Not sure about other programs. with search depths above 10 ply nowadays
>it should be peanut
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.