Author: Baldomero Garcia, Jr.
Date: 04:04:07 04/20/98
I have played against several computer programs using the Stonewall
Attack, and I've had a lot of success. Some of the victims have been
Knightstalker (386), Crafty 14.6 (Mac, 120 MHZ, 6MB hash),
MacChess 4.0, Gnuchess Mac 4.?, and HIARCS 6.0 (MAC, 120 MHZ)
Granted, the hardware has not been the fastest, but the findings are
similar: the computer programs have a very hard time playing in a
constricted environment.
Check out this game, played under a time control of 40/100 with
increment of 30 sec. After the game, I had HIARCS analyze the game
at a rate of 10 mins per move. I deleted some of the annotations that I
felt were not necessary. My comments are given for entertainment value
only (: I'm just a patzer.
Event "Individual Match"]
[Site "Yokosuka, Japan"]
[Date "1998.04.19"]
[Round "1"]
[White "Garcia, Baldomero"]
[Black "HIARCS 6.0, Mac Performa 6320"]
[Result "1-0"]
1.d4...Nf6 2.e3...d5 3.Bd3...e6 4.f4
{Baldo: the beginning of the Stonewall Attack. This move closes the
center and ensures that White will have control of the e5 post for his
Knight}
4...c5 {H6: 10 4...c5 b3 Nc6 Nf3 Ne4 O-O Be7 Nbd2 Nc3 Qe1 = 34
Baldo: HIARCS responds well. Black should attack the center
immediately.}
5.c3 {Baldo: this is key in the Stonewall. It appears that the
dark-squared Bishop is *really* bad. However, the plan is for the
Bishop to do its long tour via Bd2, Be1, Bh4, where it can either be
traded for another minor piece, or join the attack. The c3 pawn
protects the base of the skeleton}
5...Bd7 {H6: 09 5...Bd7 b3 Na6 Ba3 Ne4 Bxe4 dxe4 Qc2 Bc6 Ne2 = 46
Baldo: Ugh. Even after looking at HIARCS's analysis, I have trouble
understanding the purpose of the move. Since Black is out of its
opening book, we can see another computer weakness: it doesn't
understand opening development very well}
6.Nf3 {H6: 09 6.b3 Bd6 Nf3 O-O O-O Nc6 Ba3 Qa5 Ng5 = -54
Baldo: A regular development move. Yet, it's powerful because it
increases the power exerted over the e5 square. Balck finds itself
helpless to open the center.}
6...c4 {H6: 10 6...c4 Bc2 Bd6 O-O O-O Nbd2 Nc6 Ne5 Qb6 e4 = 50
Baldo: the sequence of HIARCS's analysis doesn't seem bad at all.
But why does it think it has 50 centipawns of advantage? This move is
IMHO very wrong. c4 gets rid of the central tension. In its attempt to
gain more space, it gives up the opportunity to open up the center.
However, this move is very typical of computer programs.}
7.Bc2 {H6: 10 7.Bc2 Bd6 b3 Qc7 Ne5 cxb3 axb3 O-O Ba3 = -41}
7...Bd6 {H6: 10 7...Bd6 b3 Qc7 Ne5 cxb3 axb3 O-O Ba3 Na6 O-O Bb5 Bxd6
Qxd6 = 38}
8.Ne5 {H6: 10 8.b3 Qc7 = -38
Baldo: A bit premature. But the move sequence doesn't seem to make much
of a difference. The knight would get to e5 eventually.}
8...O-O {H6: 10 8...O-O O-O Qc7 Nd2 Bb5 e4 Nbd7 Qe2 Rac8 = 43}
9.O-O {H6: 10 9.b3 cxb3 axb3 Qc7 Ba3 Bxa3 Rxa3 Nc6 O-O a5 = -16}
9...Bb5 {H6: 09 9...Bb5 Na3 Ba6 Rf3 Qa5 Rh3 h6 Bd2 Bxa3 bxa3 Qxa3 =
38
Baldo: this is another move that I find suspect. It may prevent me from
trying to open up the queenside, but my gameplan is nowhere near the
queenside.}
10.Nd2 {H6: 09 10.Na3 Ba6 Rf3 Nbd7 Rg3 Qe7 b4 cxb3 axb3 = -30}
10...Nbd7 {H6: 09 10...Nbd7 e4 Rc8 Nxd7 Nxd7 Qf3 Nb6 e5 Be7 Qg4 = 8}
11.Rf3 {H6: 09 11.e4 Rc8 = -8
Baldo: I can't say this move is my invention. I've seen others do it.
However, in my early days of playing the Stonewall, I normally just
place the Nd2-f3. What I found out later is that the queen's rook plays
no role in the upcoming attack. So, my plan is to place the king's rook
in h3, where it'll attack h7 -a point already attacked by the B-c2, then
place the Nd2-3, followed by the dark-squared Bishop dance to h4.}
11...Rc8 {H6: 09 11...Rc8 Rg3 Qb6 a4 Qa5 e4 Bc6 Nxc6 Rxc6 = 17
Baldo: OK. It moves the rook towards the center. But what is it
doing?}
12.Rh3 {H6: 09 12.e4 Qb6 Nxd7 Nxd7 e5 Be7 Rh3 g6 a4 Qa5 = 2
Baldo: Notice HIARCS's continued recommendations to push the e3-e4 pawn.
This is totally against the Stonewall strategy! It may be possible to
do it at some point, but we're not trying to open up the center to
attack. We want to keep it closed because we know the computer is not
able to play well against it.}
12...g6 {H6: 09 12...g6 g4 Qc7 Rh4 Kg7 g5 Nh5 e4 = 7
Baldo: introducing the first major weakness. I am a big fan of Irving
Chernev. I read his book "Logical Chess: Move by Move" three times from
cover to cover. If you want to learn how to attack the kingside, and
you're a beginner-class C, you should read it. It's a lot of fun.
Plus, you'll learn a lot. In his book, Chernev considers just about
every pawn movement in the Kingside to be an invitation for the opponent
to attack. g6 leaves big holes in h6, g7, f6. Now you can see the
importance of the B-c1.}
13.Ndf3 {H6: 08 13.e4 Qa5 Nxd7 Nxd7 Qf3 Nb6 e5 Be7 Nf1 = 12}
13...Qe7 {H6: 09 13...Qe7 Ng5 Kg7 b4 h6 a4 Ba6 Nxd7 Qxd7 = -3
Baldo: Go back and check HIARCS's evaluations up until this point.
How many have been negative against itself so far?}
14.Bd2 {H6: 09 14.Ng5 Kg7 b4 h6 a4 = 3}
14...Kg7 {H6: 08 14...Kg7 Qe1 Kg8 Ng5 a6 = -18}
15.Be1 {H6: 09 15.Qe1 h6 b4 cxb3 axb3 a6 = -15}
15...Ng8 {H6: 08 15...Ng8 b4 cxb3 axb3 a6 e4 f6 Nxd7 Qxd7 Bd2 = 9
Baldo: I smiled when I saw this move. Just as my pieces are starting to
come out of the first rank, Black's pieces are moving backwards.}
16.Bh4 {H6: 08 16.b4 cxb3 axb3 a6 e4 f6 Nxd7 Qxd7 Bd2 = -9}
16...f6 {H6: 08 16...f6 Rg3 Nxe5 dxe5 Bc5 exf6+ Nxf6 Ne5 Qe8 Bg5 =
7}
17.Qe2 {H6: 08 17.Rg3 Nxe5 dxe5 Bc7 Nd4 Bd7 Qg4 Qf7 = -3}
17...b6 {H6: 08 17...Nh6 a4 Ba6 e4 dxe4 Qxe4 Nf5 Nxd7 Qxd7 = 27
Baldo: In the game, HIARCS played b6, which does not help its position.
In the 10 minute analysis, it prefers Nh6. It's hard to say what I
would have played. g4 seems tempting, though.}
18.Rf1 {H6: 08 18.Rg3 Qe8 Bxg6 hxg6 Nxg6 Qxg6 Rxg6+ Kxg6 a4 Bc6 g4 =
-20
Baldo: In retrospect, I think I wasted a lot of time in the next few
moves. My original idea was to have a rook in the f-file if Balck takes
on e5. However, Kh1 followed by Rg1 may have been better.}
18...Rfd8 {H6: 09 18...Qe8 a4 Ba6 Nxd7 Qxd7 Bg3 Qc7 Nd2 f5 Nf3 = 37}
19.Bb1 {H6: 08 19.e4 Nf8 = 8
Baldo: another move wasted. I was lured by a subsequent Qc2, but that's
too simplistic, and very easy to defend against. The only advantage the
move has is that it doesn't weaken White's position. Notice, when you
attack, that you shouldn't try to break the position open until you're
ready. As far as I can see, Black's position is passive, but I can't
break-through yet.}
19...Nf8 {H6: 09 19...Nf8 Bc2 a5 Nd2 a4 Ng4 h5 Nf2 a3 bxa3 Bxa3 = 40
Baldo: another grin on my part. Black's passivity is reaching its max}
20.Nd2 {H6: 09 20.Bc2 Qe8 e4 Ba6 Qd2 h5 Ne1 g5 Bxg5 fxg5 fxg5 = -55
Baldo: another wait-and-see move. At this point the break e3-e4 seems
more palatable, since I already have all my artillery aiming at the
Kingside. However, the break is *still* wrong. Black has no chance of
counterplay in a closed position. It's better for me to position my
pieces better than to give any chances to the Black side.}
20...Qb7 {H6: 09 20...Qe8 Ng4 h5 Nf2 Nh7 Re1 Ne7 Bc2 Nf5 = 49
Baldo: this is a sign that the computer doesn't know what it's doing.
Perhaps Be8 would have provided more protection, although the position
would be even more passive.}
21.Ng4 {H6: 09 21.Ng4 h5 Nf2 Qf7 Nh1 Nh6 = -23
Baldo: now that the pin from the f6 pawn is gone, it's time to attack
it!.}
21...h5 {H6: 09 21...h5 Nf2 Qf7 Nh1 Qc7 Bc2 a5 Rg3 Kf7 = 18
Baldo: The voice of Irving Chernev is becoming louder. This move
introduces yet another weakness and another target.}
22.Nf2 {H6: 09 22.Nf2 Qf7 = -18}
22...Qf7 {H6: 08 22...Ne7 g4 hxg4 Nxg4 Ng8 e4 dxe4 Nxe4 Be7 Bc2 = 18}
23.g4 {H6: 09 23.Nh1 a5 e4 dxe4 Nxe4 Be7 Bc2 Nh7 Rhf3 = 5
Baldo: at last. This break proves to be decisive. Notice HIARCS's
recommendation of Nh1. If you're a postal player, and your opponent
does something like this, then you know why.}
23...hxg4 {H6: 10 23...hxg4 Nxg4 a5 e4 dxe4 Nxe4 Be7 = 29}
24.Nxg4 {H6: 09 24.Nxg4 a5 = -29}
24...Rc7 {H6: 08 24...a5 e4 dxe4 Nxe4 Be7 Bc2 Rd5 = 13}
25.Kh1 {H6: 09 25.e4 Be7 e5 f5 Bxe7 Qxe7 Ne3 Rcc8 b3 Qc7 = 61
Baldo: the plan is to load the heavy pieces in the g-file. The camel's
back is about to break.}
25...a5 {H6: 08 25...Rdc8 Ne5 Bxe5 fxe5 f5 e4 a5 Bg5 Nh7 = -8
Baldo: the move is aimless. Although Rdc8 doesn't seem to be much
better either.}
26.Rg1 {H6: 09 26.Ne5 Qe7 = 218
Baldo: I've not had a chance to see if Ne5 is more accurate/powerful
than Rg1.}
26...Rb8 {H6: 08 26...Rb8 e4 dxe4 Nxe4 Bc6 Rf3 Bxe4 = 7
Baldo: HIARCS continues the queenside play with no plan in mind.}
27.Nf3 {H6: 08 27.e4 dxe4 Nxe4 Bc6 Rf3 Be7 Bf2 b5 Rh3 = -35
Baldo: bringing back the last piece to join the attack. Patience will
eventually pay off.}
27...Qe8 {H6: 09 27...Qe8 Qg2 a4 = 15}
28.Qg2 {H6: 08 28.Qg2 Rcb7 = 32}
28...Be7 {H6: 08 28...Be7 Bxf6+ Bxf6 Nxf6 Nxf6 Qg5 N6d7 Rh6 Qf7 Bxg6
Qf6 Qh5 = -138}
29.Nxf6 {H6: 09 29.Nxf6 Bxf6 = 204
Baldo: and so the game is won with a simple exchange of pieces.}
29...Nxf6 {H6: 08 29...Nxf6 Qg5 Kg8 Bxg6 Nxg6 Qxg6+ Qxg6 Rxg6+ Kf8 =
-334}
30.Bxg6 {H6: 09 30.Qg5 Kg8 Bxg6 Nxg6 Qxg6 = 334}
30...Nxg6 {H6: 08 30...Nxg6 Ng5 = -472}
31.Rg3 {H6: 08 31.Ne5 Nd7 = 482}
31...Kh8 {H6: 08 31...Kh8 Rxg6 Qf8 Ne5 Bd8 Bxf6+ Bxf6 Qh3+ Rh7 Qxe6 Re7
Qxf6 = -534}
32.Rxg6 {H6: 08 32.Rxg6 Nh7 = 825}
32...Nh7 {H6: 08 32...Nh7 Rg7 Bxh4 Rxc7 Qg8 Qc2 Qxg1+ Nxg1 = -914}
33.Rg7 {H6: 08 33.Rg7 Bxh4 Rxc7 Ng5 fxg5 Qh5 Qh3 Be8 Qxh4 = 1192
Baldo: and the game is over.}
This page took 0.01 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.