>From the Club Championship

The recent Club Championship featured a game that was so hard-fought and
instructive that I am annotating it rather than one of my own games.  It was
the sixth round, and Bill Hughes had White against Jeff Schreiber.  The two
were tied for second place with four points, but a draw would have been
slightly more favorable for Schreiber, as he had already faced both David
Gertler and me, while Hughes had yet to play Gertler.

Bill Hughes is one of my generation--we played a game in my first rated
tournament, the 1965 Pittsburgh Metro.  In this game, he uses the slow,
python-like fianchetto style that has characterized his opening repertoire with
the White pieces for many years.  Jeff Schreiber is the most recent addition to
the ranks of Pittsburgh's masters.  His progress in the last few years has been

 1  c4   Nf6
 2  g3    e5
 3 Nc3   Nc6?!
 4 Bg2

    Black's third move is a common mistake.  Who can resist developing knights
    before bishops?  To understand what's wrong, think of the opening as a
    Sicilian with colors reversed.  Black can't play it open Sicilian style
    because he can no longer play ... d5.  But if he plays it closed Sicilian
    style, his king knight gets in the way; it's just on an awkward square,
    as we'll see in the game.

    I used to play this way against the English opening myself.  Naresh
    Jhunjhnuwala vs. Leverett [[note to editor:  that's the right spelling]],
    Hong Kong 1981, went 1 c4 e5 2 g3 Nf6 3 Bg2 Be7 4 Nc3 O-O 5 e4 Nc6?
    (5 ... d6 followed by 6 ... Nbd7 is OK) 6 Nge2 d6 7 d4.  I thought for
    a long time, wondering if I had a single useful move on the whole board.
    I couldn't find one, and played 7 ... exd4 8 Nxd4 Nxd4 9 Qxd4, with a
    lifeless position that I eventually lost.  This is the game that finally
    convinced me that I had to learn something about the English!

    Here's a more recent game with a similar opening:  Leverett-Schreiber,
    D & K Electronics Open, November 1990.  1 c4 Nf6 2 Nc3 e5 3 Nf3 Nc6
    4 g3 d6? 5 Bg2 g6 6 O-O Bg7 7 d4!  White gets a substantial advantage in
    space no matter how Black plays.  Jeff chose complications with 7 ... e4
    and went down in flames.  If he had learned the right lesson from this game
    he would have played a better defense against Bill Hughes!

    I don't mean to imply that Black's third move in Hughes-Schreiber is fatal.
    Every player of the Nimzovich defense (1 e4 Nc6) or some kinds of King's
    Indian Attack (such as 1 e4 e5 2 Nf3 Nc6 3 g3?!) knows that putting knights
    on awkward squares is not a mortal sin, as long as you know what you are
    getting into.  But 3 ... Bb4 or 3 ... c6 or 3 ... d5 would have been much

 4 ...    g6
 5  e3   Bg7
 6 Nge2  O-O
 7 O-O    d6
 8  d4

    White's method of taking advantage of Black's error is as good as the
    method used in Leverett-Schreiber.  Compare this position with a similar
    position from the "book" variation of the Closed English:  1 c4 e5
    2 Nc3 Nc6 3 g3 g6 4 Bg2 Bg7 5 e3 d6 6 Nge2 Nge7 7 O-O O-O.  It's exactly
    like the game, except that Black's king knight is on e7 instead of f6.  If
    White then plays 8 d4, Black equalizes easily with 8 ... exd4 9 exd4 Bg4.
    If now 10 h3 Bxe2 11 Nxe2 Nf5, Black's pressure on d4 forces 12 d5, giving
    up control of c5 and e5.  But in the text, 8 ... exd4 9 exd4 would simply
    surrender the center for nothing.

 8 ...   Bd7
 9  d5   Na5?

    A serious error.  9 ... Nb8 or 9 ... Ne7, despite their drawbacks, were

10 Qd3!

    Threatening to trap the knight with 11 b4.

10 ...    b6

    The only alternative was 10 ... c5 11 dxc6 Bxc6, and now either 12 Nd5
    or 12 e4, in either case forcing an eventual ... b6, a horrible weakening
    of Black's queenside.

11  e4   Nh5
12  b4   Nb7

    Not only is Black's queenside hopelessly passive and weak, but also this
    knight has no easy way of getting back into the game.

13 Be3    f5
14 exf5  Bxf5
15 Qd2   Qd7
16 Ne4    a5

    Black begins a plan of counterplay along the a-file.  This could have
    backfired.  But passive play, or exchanging with 16 ... Bh3, was not
    necessarily better.

17  a3   axb4
18 axb4  Ra4
19 Ng5   Rfa8

    Black may have rejected 19 ... Nd8 because of 20 c5.

20 Rxa4  Qxa4
21 Ne6   Rc8
22 Nxg7  Nxg7
23  h3!  Ra8
24  g4

    Now both Black knights are paralyzed.

24 ...   Bd7
25 Ng3   Qa2
26 Qxa2  Rxa2
27 Ne4   Ne8
28 Rb1?!

    White could have ended resistance with 28 f4!  The threat of opening the
    f-file and playing Bh6 and Rf8 mate is very strong.  For instance,
    28 ... exf4 29 Bxf4 Rc2 30 Bh6 Ng7 31 Nf6+, or 29 ... Kg7 30 Bg5 Rc2
    31 Be7 Rxc4 32 Bf8+, etc.  Black can defend the mate by 29 ... Nd8, but
    then 30 c5 breaks through in the center and queenside.  Also notice that
    28 ... Re2? 29 Bd2 traps the rook.

28 ...   Ba4!?
29 Nc3   Bc2
30 Rc1?

    And here White could have called Black's bluff with 30 Nxa2 Bxb1 31 Nc3 Bc2
    (Bd3 is no better) 32 Bf1 and 33 Be2.  White will invade by bringing his
    king to b5 via f1, e1, d2, c1, b2, b3, and a4.  Black is remarkably
    helpless to prevent this.  Instead, White falls into passivity.

30 ...   Rb2
31 Bf1   Nd8!??

    Oddly enough this move, which looks like a blunder, may be just a trap.
    After the routine 31 ... Kf7, White would have a hard time making progress.
    He could try 32 Bd2 (with the idea of Be1, Nb5, and Ra1); then after
    32 ... Bd3 33 Bxd3 Rxd2 34 Bc2 at least his queenside would still be secure.

32 c5

    Finally breaking through--or falling for it?  Unless White can achieve
    something tangible with this break, he is just dissipating what's left of
    his advantage.  32 Bd2, as in the previous note, was plausible.  Time
    pressure was playing a role at this point.  The time control was at move 40.

32 ...   bxc5
33 bxc5  dxc5
34 Bxc5  Nb7
35 Ba3

    35 Be3 was more patient.

35 ...   Rb3
36 Nb5

    Threatening 37 Bc4.  The logical 36 Rxc2 leads to 36 ... Rxa3 37 Nb5 Ra8!
    38 Nxc7 Rc8! 39 Bc4!? Rxc7 (39 ... Nxc7 40 d6+ Kf8 41 Ba6 Nxa6 42 Rxc8+ Kf7
    is unclear) 40 d6+ Rf7 41 d7 Ned6 and Black is apparently hanging by a hair.

36 ...   Bd3
37 Bxd3  Rxd3
38 Nxc7  Nbd6?

    38 ... Nxc7 39 Rxc7 Rxa3 40 Rxb7 Rxh3 was suggested in the post-mortem.  It
    looks like a draw:  41 Re7 Rd3 42 Rxd5 h6! (preventing 43 g5 and
    threatening ... Kf7 and ... Kf6) 43 f4 Rd4! (but not 43 ... Rg3+ 44 Kh2 Rxg4
    45 d6 Rxf4 46 d7).

39 Bxd6  Nxd6
40 Kg2   Kf7

    The post mortem suggestion of 40 ... Rd2, threatening 41 ... Ne4, is
    interesting.  41 Kf3 is met by 41 ... Rd3+ 42 Ke2 (42 Kg2 repeats the
    position) 42 ... Rxh3 and now 43 Rc6 looks menacing, but after 43 ... Nf7
    44 d6 e4! Black's rook gets back into the game.

41 Rc2    h5!

    A finesse.  Black forces a trade of pawns (favorable for the defender)
    and does so on his own terms, since 42 gxh5 gxh5 would leave White's own
    pawns in ruins.

42  f3   hxg4
43 hxg4  Ne8

    A fateful decision.  Black's knight, rook, and king are all well posted.
    The most logical plan is to capitalize on this:  43 ... g5, with the idea
    of ... e4, ... Kf6, ... Ke5, etc.  (But not immediately 43 ... e4 because
    of 44 f4.)  Black's defense plays itself:  44 Rc6 Ke7 45 Ne6 e4
    46 Nxg5 exf3+ 47 Nxf3 Rxd5, etc.  It may also be possible to draw by doing
    nothing (moving the king back and forth), but in this case the more active
    plan is the safer one.  The text move should also have led to a draw, but
    it's a very different type of endgame.

    With the exchange of knights the game becomes more complicated, not less.
    Rook and pawn endgames can be very intricate, and this one is a good
    example.  At the same time, rook and pawn endgames are by far the most
    common type, and every player can benefit by studying them.

44 Nxe8  Kxe8
45 Rc5

    45 Rc6 is similar.

45 ...   Kf7

    45 ... g5! was best.  The point is to prevent White's next move and
    establish a blockade.  Of course Black must be careful not to lose the
    g-pawn or end up two pawns down:  46 d6 Kf7! 47 Rxe5 Kf6, or 47 Rc7+ Ke6
    48 Rg7 Kf6.

46  g5!

    An inspired conception.  Even with only kings and one pair of rooks on the
    board, an advantage in space can be important!

46 ...   Rd2+!?
47 Kg3   Ke7
48 Rc6   Rxd5?!

    There was nothing wrong with 48 ... Kf7.  Sometimes the defender should
    just hold tight.  In this case after 48 ... Kf7 49 Rf6+ Kg7 50 d6 Kh7,
    White's king is boxed in and has no way to come to the support of the
    d-pawn.  For instance, 51 Rf7+ Kg8, or 51 f4 Rd3+ 52 Kg4 Rd4.

49 Rxg6  Kf7
50 Rf6+  Kg7
51 Kg4

    The sealed move.  White has a dangerous passed pawn.  Black's king is on
    the ideal square, right in front of the pawn, so that if the other two
    pawns were traded off, the endgame would usually be a draw.  (But as we'll
    see, this is not an ironclad rule.)  Black should strive to keep his rook
    mobile, but this must be combined with perfectly timed tactics:  defending
    his own pawn; attacking White's pawns; and an occasional cheapo.

    51 ... Ra5 is possible.  If 52 Kf5? e4+ draws, while 52 Kh5 Ra3 is also
    a standoff.  But 52 Rb6 is very dangerous.  For instance 52 ... Kf7?
    (52 ... Ra3 is OK) 53 Kh5 Ra3 54 Kh6! Rxf3 55 g6+ Kf8 (55 ... Kg8
    56 Rb8+ Rf8 57 Rxf8+ Kxf8 58 Kh7) 56 Rb8+ (56 g7+ Kf7 57 Rb7+ Kf6!) Ke7
    57 Rb7+! (57 g7 Rh3+) Kf8 (57 ... Kf6 58 Rf7+) 58 g7+ wins.  This variation
    shows the subtlety of rook and pawn endings even with the smallest amounts
    of material.

    The clearest draw would have been to proceed to the third rank immediately:
    51 ... Rd3; and stay there.  White has no way to make progress.  52 Re6 Re3
    (but not 52 ... Rd4+ 53 Kh5 Rf4 54 Re7+! transposing to the game)
    53 Kg3 Re2! keeps things under control.  52 Ra6 Rb3 53 Ra4 Kg6 54 Re4 Rb5,
    or 53 Ra5 Re3, is equally harmless.  Schreiber, giving credit to Tom Magar,
    Fernand Gobet, and Fred Sorensen for help with analysis, gives several
    interesting lines starting with 51 ... Rd3 52 Ra6 Kf7 53 Ra4 Rd1.  Now
    54 Re4 Rg1+ 55 Kf5 Rg3 56 Re3 e4! is a spectacular drawing combination.
    If 54 Kf5 Rf1? 55 Ra7+ followed by 55 Ra3 wins, but 54 ... Rd3 appears to
    lead to a draw against either 55 Ra7+ or 55 g6+.  But I think that in this
    line, 53 ... Kg6 is much clearer and safer.  Now back to the game.

51 ...   Rd1
52 Kf5   Rg1
53 Re6   Rg3
54 Re7+! Kf8
55 Rxe5  Rxf3+
56 Kg6

    White is winning.  Black's king has been pried off the queening square and
    can't get back:  56 ... Kg8 57 Re8+ Rf8 58 Rxf8+ Kxf8 59 Kh7.  Meanwhile
    White's king snuck into g6 at a time when Black's rook was too busy to
    check it away.  What went wrong?  It appears that Black's 51st move was the
    culprit, because I haven't found any saving alternatives for Black after
    that point.

56 ...   Rf1
57 Rb5   Rf2
58 Rb8+  Ke7
59 Rg8

    Is Bill trying to promote his rook to a queen?  Not to worry, he soon gets
    back on track.

59 ...   Rf1
60 Kh7   Rh1+
61 Kg7   Rg1
62  g6   Rf1
63 Rh8   Rf2
64 Rh4   Rg2
65 Re4+  Kd8
66 Kh7   Rh2+
67 Kg8   Rg2
68  g7

    This is the Lucena position.  The winning method is essential knowledge for
    the student of rook and pawn endings.  Black puts up the most resistance
    with 68 ... Rh2.  White then wins by 69 Kf7 Rf2+ 70 Kg6 Rg2+ 71 Kf6 Rg1
    (or 71 ... Rf2+ 72 Kg5 Rg2+ 73 Rg4) 72 Re5 followed by 72 Rg5.

68 ...   Rg1
69 Rh4   Ke7
70 Kh8   Resigns

Enough chess for three games!  Hughes went on to take clear second place by
holding Gertler to a draw in the last round.  Schreiber tied with Gertler for
third with five points.