It's Not Fair--I Was Winning!
Chess players have a perverse desire to tell stories, not about their wins,
but about their agonizing losses: the games where they were beating the
opponent, usually a stronger player, but he got away. Jeff Schreiber has sent
me two such games: a loss to Mark Eidemiller from last April, and a recent
loss to Jerry Meyers from the Pittsburgh Chess League. What strength of
character--but it would have demonstrated even more strength of character if
Jeff had won the games. Here's the game against Mark Eidemiller; the game
against Meyers will appear in a future issue.
Master-Expert Open
Round 5, 4/20/93
White: Mark Eidemiller
Black: Jeff Schreiber
Queen's Indian Defense
1 d4 Nf6
2 Nf3 e6
3 c4 b6
4 g3 Bb7
5 Bg2 Be7
6 Nc3 d5
Black has three plausible moves at this point: 6 ... Ne4, 6 ... d5, and
6 ... c5. For many years 6 ... Ne4 was considered to lead almost straight
to a draw after 7 Nxe4 Bxe4, and so Black was rarely allowed to play the
Queen's Indian. Nowadays, White will generally play 7 Bd2, and the burden
is on Black to demonstrate that he doesn't get a significant disadvantage
in territory.
By playing 6 ... d5, Black solves the territory problem immediately, but
creates pawn weaknesses. Either his c-pawn will be backward on the
half-open file, or (if he pushes it to c5) his d-pawn will become weak,
or the d-pawn and c-pawn together will become "hanging pawns". This is
not a fatal disadvantage and the resulting games, like this one, can
become quite double-edged.
The third possibility, 6 ... c5, is definitely bad. White answers 7 d5! exd5
8 Nh4 and 9 cxd5 with a very favorable Benoni-like pawn structure.
7 cxd5 Nxd5
A typical example of the "hanging pawns" game is 7 ... exd5 8 O-O O-O
9 Ne5 Na6 10 b3 c5 11 Bb2. Instead, Black trades a pair of knights, slightly
simplifying the game, but not solving the basic problem of pawn structure.
8 O-O Nd7
9 Nxd5 exd5
If 9 ... Bxd5 10 Qc2 O-O 11 e4, or 10 ... Nf6? 11 Ne5, is not satisfactory
for Black, so he can't avoid hanging pawns. Now the players are on their
own resources. Uncharacteristically, it is Eidemiller who loses his way.
10 Qa4?!
This would have made sense with the knights still on the board, because of
a tactical trick. To see this, go back to move 7 and play 7 ... exd5
8 O-O Nbd7 9 Qa4! O-O 10 Ne5 Nxe5 11 dxe5 Ne8 12 Nxd5! Bxd5 13 Rd1 and
14 e4. This is a transposition to Najdorf-Wexler, Buenos Aires 1965, quoted
in MCO-13. But in our game, the queen belongs on c2.
10 ... O-O
11 Rd1
11 Ne5 Nxe5 12 dxe5 gives White less than nothing.
11 ... Bf6
12 Bf4
Unless White can follow up with Ne5, this move is risky. White can no
longer play e3 without trapping his own bishop. 12 Qc2 followed by 13 b3
and 14 Bb2 was still OK.
12 ... Re8
13 Rac1 c5
14 Bh3 Nf8
15 dxc5
Last chance to dig in with 15 Qc2. But now White starts a faulty
combination.
15 ... bxc5
16 Be3 d4
17 Rxc5?? Qe7
Black wins a whole piece.
18 Qc4 Bxf3
18 ... dxe3 19 Rc7 Ba6! was OK, or even better.
19 exf3 dxe3
20 Rc7 exf2+
21 Kg2
21 Kxf2 Qe3+ 22 Kg2 Qe2+ is even more hopeless.
21 ... Qe1
22 Qf1 Qxd1
23 Qxd1 Re1
24 Qd3 Bd4
Some people are never satisfied. Black can just cash in with 24 ... f1Q+.
After 25 Qxf1 Rxf1 26 Kxf1 Bxb2 he would have an easy win. Instead he tries
to win even more material. But if he had calculated correctly, he would have
seen that this accomplishes less than nothing. Of course, Black is now
threatening mate in one.
25 f4 Rae8
26 Rc2 Rg1+?
27 Kf3 Re3+
28 Qxe3 f1Q+?!
Last chance to bail out to a piece-up ending with 28 ... Bxe3. Even then,
Black would have had his hands full after 29 Kxe3 f1Q 30 Bxf1 Rxf1 31 Rc7.
29 Bxf1 Rxf1+
30 Ke2 Bxe3
31 Kxf1
It now looks pretty drawish. But with both players stirring the pot, the
game soon gets complicated.
31 ... g6
32 Ke2 Bd4
33 Kd3 Ne6
34 b3 Kg7
35 Rc6 Bg1
36 Rc2 Kf6
It's pretty easy for White to stop ... Kf5. Perhaps 36 ... f5 would have
been more accurate.
37 Ke4 Nc5+
38 Kf3 h5
39 h3
39 Kg2 Bd4 40 Kf3 was safer, as we shall soon see.
39 ... Ne6
40 Ke4 a5?!!
This move doesn't make much sense at first sight. I believe that Black made
this move to distract White from the kingside, that is, to set a crude trap;
and White fell right in.
41 g4 g5!
White's kingside is now vulnerable. For instance, 42 fxg5+ Nxg5+ 43 Kd5 h4!
44 Rc3 Bf2 followed by ... Bg3 and ... Nxh3 looks very dangerous. White
finds a clear path to a draw, but at the cost of his rook.
42 f5 Nf4
43 h4!? hxg4
44 Rc6+ Kg7
45 hxg5 Bh2!
46 Rc1
There's nothing better.
46 ... g3
47 Kxf4 g2+
48 Ke4 g1Q
49 Rxg1 Bxg1
50 Kd5
White can get to the a-pawn "just in time." A logical finish now would be
50 ... Be3 51 g6 fxg6 52 fxg6 Kxg6 53 Kc4 Kf6 54 Kb5 Bd2 55 a3 and 56 b4,
draw. Instead Black goes completely haywire.
50 ... Kf8?
51 Kc4 Ke7?
52 Kb5 Be3
53 g6 f6??
53 ... fxg6 54 fxg6 Bd2 still draws. Notice that if White had played
53 Kxa5, he would have gotten only losing chances; after 53 ... Bxg5, there's
enough time for Black's king and bishop to blockade the queenside pawns and
win them.
54 Kxa5 Bh6
55 Kb5 Kd6
56 a4 Kc7
57 Kc5 Bf8+
58 Kd5 Resigns
Prison Chess
The State Correctional Institution, Pittsburgh (SCIP) has been the location of
an active chess club and a series of tournaments starting about a year and a
half ago. In that time they have developed some strong players. Ken Davenport
is at Expert strength or near to it, and has taken the scalps of Bill Hughes,
Don Meigs, and other strong area players. Gildo DeMichele, though currently
rated only 1751, has also beaten some Expert and Class A players. It's quite
an accomplishment for any chess program to develop such strong players "from
scratch" in a short time. How do I join? (Maybe I shouldn't ask.) This month
I feature Davenport's instructive win against my fellow Transarc employee
Ron Kownacki. In a future issue I'll feature a game by DeMichele against
many-time Club Champion Don Meigs.
SCIP-Open V
May 16, 1993
White: Ron Kownacki
Black: Ken Davenport
French Defense, Steinitz Variation,
by transposition from Alekhine's Defense
1 e4 Nf6
2 Nc3
How to handle Alekhine's Defense? Even strong players often resort to moves
like 2 Nc3 because they are afraid to enter lengthy and complicated "book"
variations. No doubt that's what you get into with the Four Pawns Attack
(2 e5 Nd5 3 c4 Nb6 4 d4 d6 5 f4), or some other wild lines. But my
recommendation, even for players of average strength, is to go ahead and
play the main line: 2 e5 Nd5 3 d4 d6 4 Nf3. You'll find that it's pretty
easy to play, and not so terrifying once you're in it.
The problem with 2 Nc3 is that it's insipid. Black can play 1 ... e5,
transposing to the Vienna Game, which is quite all right for him. The move
played in the game is also good.
2 ... d5
3 e5
3 exd5 is also playable.
3 ... Nd7
There are two plausible alternatives, 3 ... d4 and 3 ... Ne4. The former
leads directly to an endgame after 3 ... d4 4 exf6 dxc3 5 fxg7 cxd2+
6 Qxd2 Qxd2+ 7 Bxd2 Bxg7 8 O-O-O. Black isn't out of the woods in this
position, although he may be able to equalize. 3 ... Ne4 leads to
complications after 4 Nce2! (threatening 5 d3 Nc5 6 d4 Ne4? 7 f3) 4 ... d4
5 c3. In this position Black must generally gambit a pawn, because of
variations like 5 ... dxc3? 6 Qa4+ or 5 ... Nc6 6 cxd4 Nxd4? 7 Qa4+. He
gets unclear compensation. The move played in the game is the soundest.
4 d4
Also 4 Nxd5 Nxe5, or 4 f4 e6, or 4 e6!? fxe6 5 d4 c5, are playable and
interesting, but not bad for Black.
4 ... e6
We have transposed to the Steinitz Variation of the French Defense (1 e4 e6
2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7). Both players have made sensible moves and
an exciting game is in prospect. But the position on the board doesn't
tell the whole story. Kownacki's problem, as he told me afterward, was that
he was "out of book" already. By playing 2 Nc3 instead of 2 e5, he had only
jumped from the frying pan to the fire.
5 Nf3
The alternatives are 5 Nce2 and 5 f4, the latter being more common. The
line 5 f4 c5 6 Nf3 Nc6 7 Be3, in particular, has been the main variation of
the Steinitz for many years. The game position is sometimes reached by the
move order 1 e4 e6 2 Nf3!? d5 3 Nc3 Nf6! (sounder than 3 ... c5 or 3 ... d4)
4 e5 Nfd7 5 d4.
5 ... c5
6 Be3?!
By going on the defensive, White gives Black time to finish his development
and get comfortable equality. No better was 6 Bb5 as Alex Barbalat played
against me in a recent Pittsburgh Chess League game. The only serious try
for advantage is 6 dxe5. The game might then go 6 ... Nc6 7 Bf4 Bxc5
8 Bd3 f6 (but not 8 ... O-O? 9 Bxh7+!) 9 exf6 Nxf6 or Qxf6. A game featuring
this variation, Lein-Dvoretsky from the Moscow Championship of 1973, was
analyzed at length in the most recent issue of the American Chess Journal.
6 ... Nc6
7 Bb5 Be7
8 O-O O-O
9 Nce2
White would like to reinforce his center with 10 c3. It is obvious that
after 9 ... cxd4 10 Bxd4 (not 10 Bxc6 dxe3, or 10 Nexd4 Ncxe5) 10 ... Nxd4
11 Nexd4 Nc5, or 11 Qxd4 Qb6, Black would have nothing to fear. The plan
he selects is more double-edged.
9 ... Qb6!??
10 Bxc6 bxc6
11 b3 cxd4
12 Nexd4 c5
13 Ne2 h6?!!
More natural was 13 ... f6 or 13 ... f5. Black is setting a trap, but who
is falling into it?
14 Re1
14 h4 or even 14 g4!? could be considered.
14 ... Qc7
15 Bf4
And here 15 Ng3 would have been fine, since 15 ... Nxe5? 16 Nxe5 Qxe5
17 Bxc5 Qc7 18 Bxe7 Qxe7 19 Qxd5 wins for White.
15 ... g5
16 Bg3 g4
17 Nd2 Nxe5
Here it is--the point of Black's "trap". But White has plenty of
compensation for the pawn. He should play 18 Nc3, whereupon 18 ... Bf6
loses to 19 Qe2 and 18 ... Bd6 loses to 19 Nb5. So Black must play
18 ... f6. Now White must not regain the pawn immediately with 19 Bxe5 fxe5
20 Qxg4+, because then his attack stops and he gets overrun by Black's
center pawns. Instead 19 f4! gxf3 20 Nxf3 Nxf3+ 21 Qxf3 Qb7 22 Bf4 allows
White to pursue a promising attack with little risk.
18 Nd4 Bf6
19 Bxe5??
It's not too late for 20 Nb5 Qd7 (20 ... Qb8? 21 Qe2) 21 Bxe5 Bxe5
22 Rxe5 Qxb5 23 Qg4+, with material equality.
19 ... Bxe5
20 Qxg4+ Kh7
21 N4f3 Bxa1
22 Rxa1 e5
23 Qh5 f5
The pawn formation that every French Defense player lives for! The rest
of the game is just mopping up.
24 Re1 e4
25 Ng5+ Kg7
26 Nh3 Qf7
27 Qh4 f4
28 g3 Bxh3
29 Qxh3 fxg3
30 Qxg3+
Not 30 fxg3? Qf2+.
30 ... Kh7
31 Kf1 Qf4
32 Ke2 Qxg3
33 hxg3 Rae8
34 c4 d4
35 Kd1 e3
36 fxe3 dxe3
37 Nb1
37 Nf1? e2+.
37 ... Rf2
38 Re2
38 Nc3 puts up better resistance.
38 ... Rf1+
39 Kc2 Rf3
40 g4 Re4
41 Nc3 Rxg4
42 Nd5 Rf2
Resigns