White: Ryan Milisits
Black: Don Meigs
Exhibition game, Pittsburgh Chess Club
June 27, 2000
French Defense, Classical Variation
 1 e4    e6
 2 d4    d5
 3 Nc3   Nf6
 4 Bg5   Be7
 5 e5    Nfd7
 6 Bxe7  Qxe7

   White's pawns on d4 and e5 cramp Black's game.  Black should fight back by challenging
   White's center.

 7 f4    a6

   The immediate 7 ... c5 would be answered by 8 Nb5, and the threats of 8 Nd6+ and 8 Nc7
   can't both be stopped.

   To prepare ... c5, Black can play the text (7 ... a6) or 7 ... O-O.

 8 Nf3   c5
 9 a3

   This move is a necessary preparation for White's next move:  if 9 Ne2 immediately, then
   9 ... cxd4 10 Nxd4? Qb4+.  But the whole plan is too sluggish; White should play 9 dxc5
   and proceed with development.

 9 ...   Nc6

   The immediate 9 ... cxd4 might be more accurate, but after 10 Nxd4 Nc6 11 Nce2, Black
   may not have anything better than 11 ... Nxd4, transposing to the game.

10 Ne2

   White threatens 10 c3, a move that looks modest but effectively shuts Black's knight
   on d7 out of useful service.

10 ...   cxd4
11 Nexd4 Nxd4
12 Nxd4

   As often happens in the Classical French defense, White has a knight on d4 that can't
   easily be dislodged and is very annoying.  King's Indian aficionados know that in that
   opening, one often sees a similar knight on e4; in fact, the Classical French pawn
   structure is like a King's Indian pawn structure viewed in a mirror.  Black should not
   panic, for he still has excellent resources.

12 ...   b5?

   There is nothing much to do on the queenside at present, so this move is a waste of
   time.  Black should castle and prepare to play ... f6.  After 12 ... O-O, White might
   play 13 Bd3.  Black can then play 13 ... f6 immediately, because if 14 exf6 Qxf6
   suddenly White would be in deep trouble.  Or White could play 13 Qd2, so that after
   13 ... f6 14 exf6 Qxf6 he could just castle, with advantage.  Also if 14 ... gxf6
   15 f5! wreaks havoc:  if 15 ... e5 16 Ne6, or if 15 ... exf5+ 16 Be2, and Black is
   helpless.  But Black can play 13 ... Nc5, and White can't stop 14 ... f6 with a fine
   game.

13 Bd3   Qh4+?

   When chess players can't find a defensive plan, they sometimes react like a deer in
   a car's headlights.

14 g3    Qe7
15 Qe2

   White could consider 15 c3.

15 ...   Nc5

   White's buildup will eventually threaten f5.  Black's basic method of coping with this
   is to castle and play ... f6.  Because he has lost so much time, this won't be enough
   to equalize, but it's still the best strategy.  The text move is less accurate.  White
   could now play Tom Martinak's suggestion of 16 b4, and now 16 ... Nxd3+ 17 Qxd3 leaves
   Black completely blockaded, although White must still demonstrate a way to break
   through.  Also, 17 ... Na4 puts the knight out of play indefinitely, and 17 ... Nd7 is
   an admission of further error.

16 O-O   g6

   16 ... O-O was still possible.

17 Rae1

   This move is unnecessary, and the tempo wasted eventually comes in handy for Black.

17 ....  Bd7

   17 ... Bb7 allows the tempting sacrifice 18 f5:  18 ... gxf5 (or 18 ... exf5) 19 Bxf5!,
   or 18 ... Nxd3 19 fxe6 Nxe1 20 exf7+ Kd7 21 e6+ Kc7 22 Qxe1 (Martinak).  White may not
   have enough for the rook here ( or after 20 ... Kf8).  But in this variation, he can
   play 19 f6 Qc5 20 Qxd3, with a complete blockade, as in the note to Black's 15th move.

18 g4    Rg8
19 Kh1   b4!

   If Black's bishop were still on c8, this move would be answered by Nc6.  This is why
   White's waste of a move at move 17 was significant.

20 f5    bxa3
21 bxa3  gxf5
22 gxf5  Qh4
23 Qe3   Rc8

   Diagram:  2r1k1r1/3b1p1p/p3p3/2npPP2/3N3q/P2BQ3/2P4P/4RR1K

24 Rf4?

   An innocent-looking move, but now instead of winning, White is struggling to survive.
   The suggestion of 24 Rg1 (Martinak) looks logical.  After 24 ... Rf8 or 24 ... Rh8,
   White might build more pressure with 25 Rb1.  In spite of White's impressive buildup,
   it's not easy to find a breakthrough.  But Black would be nearly in zugzwang; for
   instance, if he exchanges knight for bishop, White can bring a rook to b7, and with the
   other rook on g7, he could then play fxe6 to wreak havoc on the seventh rank.

24 ...   Qg5
25 Qf2   Nxd3
26 cxd3  Rc3
27 Rf3

   White could counterattack by 27 Rb1!  After 27 ... Rxd3 28 Rb8+ Bc8 29 Rxc8+ and now:
   (A) 29 ... Kd7 30 Rc1 Rxd4 31 fxe6+ fxe6 32 Rf7+ Kd8 33 Rf8+ Rxf8 (33 ... Kd7
   34 Qf7+ Qe7 35 Qxg8 wins) 34 Qxf8+ Kd7 35 Qc8+ Ke7 36 Rc7 mate; (B) 29 ... Ke7
   30 Nc6+ Kd7 31 Qa7+ Kxc8 32 Qb8+ Kd7 33 Qb7+ Ke8 34 Qc8+ and mate next move.  Of
   course, Black doesn't have to play 27 ... Rxd3.

27 ...   Qe7
28 Rb1   Qc5

   The immediate 28 ... Qxa3 looks OK.  Black was probably already in severe time
   pressure.

29 Qe3   Qxa3
30 Rg3   Rxg3

   Meigs reports that he now had 2 minutes to play his remaining 10 moves.

31 Qxg3  Rc1+
32 Rxc1  Qxc1+
33 Kg2   Qd2+

   This is better than the suggestions of 33 ... Qh6 or 33 ... Qc5, which would allow
   White to pick up one of his pawns by 34 Qg8+ Qf8 35 Qxh7.

34 Kf1   exf5
35 Ne2

   35 Qg8+ Ke7 36 Qg7, threatening perpetual by Qf6+ and Qh8+, may be insufficient, after
   36 ... Qxd3+ 37 Ke1 Qe4+ 38 Kf1 Bb5+ 39 Nxb5 Qd3+ and 40 ... Qxb5.

35 ...   Qh6?

   Meigs suggests 35 ... Bb5 or 35 ... Be6.  After 35 ... Bb5 36 Nf4 Qd1+ 37 Kf2 Qg4!
   White would be desperate indeed.  After the text move the Queen is shut out of play.

36 Nf4   Be6
37 d4    Qf8

   Meigs writes:  "Again time pressure.  The queen should stay where it attacks the
   knight.  The king should move - to d7 say.  Now Black throws away the probable win and
   a draw is inevitable."  He is correct, but I have to add that after 37 ... Kd7 38 h4
   there is no obvious way for Black to make progress.

38 Qd3!

   There's no way to save the a-pawn.

38 ...   Qb4
39 Qxa6  Qc4+
40 Qxc4  dxc4

   Draw agreed.  Meigs now gives 41 d5 Bc8 42 Ke2 Ba6, and if 43 Ke3? c3 tying down
   White's pieces, but 43 Kd2! assures the draw.