In this issue I continue with the games submitted by former Pitt student Eric
Johnson, begun in the January issue.  Johnson's rating, though "only" 2115 at
the time of his letter, had peaked earlier in the 2190's.  "For that," he says,
"I thank all my chess-playing friends in Pittsburgh for those early (and
painful) lessons."

Johnson writes, "By 1992 I was back over 2100 again, and had my best tournament
result to date:  a tie for first with NM Roll in the 1992 World Open Warm-up.
Particularly satisfying was the fact that I played sharply (if not necessarily
correctly!) in the final round and earned an honorable draw with the co-winner."
The opening, an important gambit in the Exchange Gruenfeld, leads to fast
action with no letup in the middlegame.  Time pressure takes its toll toward
move 40 as the players trade serious errors.

   White: Eric Johnson
   Black: Craig Roll
   World Open Warm-up, Philadelphia, 1992
   Gruenfeld Defense, Exchange variation

 1 d4     Nf6
 2 c4     g6
 3 Nc3    d5
 4 cxd5   Nxd5
 5 e4     Nxc3
 6 bc3    Bg7
 7 Nf3

   This move was rediscovered in the 70's.  Before then, it was thought that
   Black's eventual ... Bg4 would at least equalize.  But it's not so simple.

 7 ...    c5
 8 Rb1    O-O
 9 Be2    cxd4
10 cxd4   Qa5+
11 Bd2

   A theoretically important gambit.  Both sides have played logically since
   move 7.  If White now allows the exchange of queens, Black soon catches up
   in development and equalizes:  11 Qd2 Qxd2+ 12 Bxd2 b6 13 O-O Bb7 14 d5 Ba6
   (Gaprindashvili-Levitina, Smederevska Palanka 1987), or 13 Rc1 Bb7 14 Bd3 Na6
   (de Boer-Mikhalchihin, Cascais 1986).

11 ...    Qxa2
12 O-O    Qe6
13 Qc2    Qc6!?

   Possibly an improvement on Conquest-Korchnoi, Lugano 1986, which continued
   13 ... Qd7 14 d5.

14 Bc4

   The alternative was 14 Qd3.

14 ...    Nd7

   14 ... Bg4 15 d5 and 16 Ng6 is no better.

15 d5

   White can consider 15 Qa2, since 15 ... Qxe4 16 Rfe1 and 15 ... Nb6
   16 Bb3 (16 Bd3 Qa4) Qxe4 17 Rfe1 both leave Black with problems to solve.

15 ...    Qc7
16 Qa2    Nb6
17 Bb3    Bg4
18 Ng5    Be2!?
19 Rbc1   Qd7
20 Rfe1   Bb5
21 Be3

   21 e5!? seems like the logical sequel to White's 15th move.  Black can't
   trade his way to safety with 21 ... Ba4:  22 Bxa4 Qxa4 23 Qxa4 Nxa4 24 Rc7,
   or 22 Bxa4 Nxa4 23 e6 fxe6 24 Nxe6 Rf7 (24 ... Rfc8 25 d6) 25 Rc7.

21 ...    Ba4
22 f4     Rac8
23 Bxb6   Bxb3
24 Qxb3   axb6
25 e5

   White must have thought this was more accurate than 25 Qxb6, but that move
   doesn't look bad.  After 25 ... Qa4 26 Qxb7 Rxc1 27 Rxc1 h6 28 Nf3 Qxe4
   29 Re1 Qxf4 30 Qxe7, White is OK, while less forcing lines can still lead to
   trouble for Black.

25 ...    Rxc1
26 Rxc1   Rc8
27 Rxc8+

   27 Rd1 looks interesting (White will play 28 d6).  Perhaps the clock was
   playing a role at this point.

27 ...    Qxc8
28 Qd1

   28 Qxb6!? looks strange but may hold the draw:  28 ... Qc1+ 29 Kf2 Qxf4+
   30 Nf3 Bxe5 31 Qd8+ Kg7 32 Qxe7.  At this point, one variation that shows
   some of the intricate tactical complexity is 32 ... Bd4+ 33 Ke2 b5
   34 Nxd4 Qxd4 35 d6 Qf6 36 Kd3 b4 37 Kc4 Qc3+ 38 Kd5, etc.

28 ...    Qc4
29 g3     b5
30 Nf3

   Not 30 d6? exd6 31 exd6 Qd4+.

30 ...    b4
31 d6     exd6
32 exd6   b3?

   32 ... Bf6 immediately, so as to answer 33 d7 Bd8 34 Qe1 Qe6, would have
   given White headaches.

33 d7     Bf6
34 d8Q+   Bxd8
35 Qxd8+  Kg7
36 Qe7?

   36 Qd2 Qc2 37 Kf2, etc., should win.

36 ...    Qc1+
37 Kg2    b2
38 Qe5+   Kg8
39 Qe8+   Kg7
40 Qe5+   Kg8
41 Qe8+
    Drawn

Whelan-Johnson, Philadelphia 1992, Sicilian Defense, Taimanov variation:
1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e6 5 Nb5 d6 6 c4 Nf6 7 N1c3 a6 8 Na3 Be7
9 Be2 O-O 10 O-O b6 11 Bf4 Bb7 12 Qd2 Ne5 13 Qd4 Qc7 14 Be3 Bc6 15 f4 Ned7
16 Kh1 e5 17 Qd3 ef4 18 Rxf4 Nc5 19 Bxc5 bxc5 20 Nd5 Bxd5 21 exd5 Nd7
22 Raf1 Bf6 23 Qf3 Ne5 24 Qe4 Ng6 25 Rg4 Rfe8 26 Qc2 Qe7 27 Bd3 Qd7 28 Bf5 Qb7
29 b3 Re5 30 Re4 Rae8 31 Rfe1 Qb4 32 Rxe5 Qxe1+, Resigns.

Johnson entered the 1993 U.S. Open and, on the way to a plus score, played an
interesting draw against IM Walter Shipman.  My own files include four games
against Shipman, who has been active in New York City chess for decades.

   White: Walter Shipman
   Black: Eric Johnson
   U.S. Open, Philadelphia, 1993
   Catalan Opening

 1 d4     d5
 2 Nf3    e6
 3 g3     Nf6
 4 Bg2    Be7
 5 O-O    O-O
 6 c4     dxc4
 7 Qc2

   The more natural 7 Nbd2 could be answered by 7 ... b5 holding onto the pawn.
   But after 7 Qc2 b5? 8 a4 c6 9 axb5 cxb5 10 Ng5! wins.

 7 ...    a6
 8 a4     Nbd7

   This old-fashioned move allows White to regain his pawn the "preferred" way,
   with 9 Nbd2.  That's also why the natural 8 ... c5 is unpopular:  9 dxc5 Nc6
   10 Na3! leaves White with easy and natural development, while Black will have
   a tough time solving the problem of the queen bishop.  Nowadays moves such
   as 8 ... Bd7 or 8 ... Nc6, strange-looking but sharper, are seen more often
   here.  But now White, instead of capitalizing on this inaccuracy, returns
   the favor, playing routinely and missing some typical Catalan subtleties.

 9 Qxc4   b5!
10 Qc6

   Of course, 10 axb5 axb5 costs White a rook.

10 ...    Rb8

   One of my own games, played against Kevin Plesset in 1977, reached this
   position by transposition.  Plesset played 11 axb5 and after 11 ... axb5
   12 Ra7, he had good pressure on the queenside, which I was never able to
   shake off.  Black could answer 11 axb5 with the interesting 11 ... Rxb5!?.
   Shipman may have decided that this was worth avoiding.

11 Qc2    Bb7

   11 ... c5 might have been more accurate.

12 axb5   axb5
13 Bg5

   This discourages 13 ... c5.  Black makes a courageous decision to push the
   game in another direction altogether.

13 ...    h6!?
14 Bxf6   Bxf6
15 Nbd2   e5
16 dxe5   Nxe5
17 Nxe5   Bxg2!?

   Another courageous decision.  I think that 17 ... Bxe5 would have been
   adequate to hold the position:  18 Qe4 Rb8 19 Qxe5 Qxd2 20 Rfd1 Qb4, or
   18 Nb3 Rb8, don't seem to leave White with anything tangible.  Instead,
   Johnson elects to try to take the initiative, at the cost of a pawn.

18 Nxf7   Rxf7
19 Kxg2   Qd4
20 Rab1   Re7
21 e3     Qb4
22 Qb3+

   This doesn't succeed in breaking the bind on the queenside.  22 b3 may have
   been an improvement.

22 ...    Qxb3
23 Nxb3   Rd8!

   23 ... Ra2 is too greedy:  24 Nc5 Rxb2 25 Rxb2 Bxb2 26 Rb1 Ba3 27 Rxb5 Re5
   28 Rb8+, or 24 ... Bxb2 25 Nd3, enables White to regain his extra pawn.

24 Nc5    Rd2
25 Rfd1   Rc2
26 Rdc1   Rd2
27 Rd1    Rc2
28 Rdc1   Rd2
    Drawn

Johnson-Carroll, Philadelphia 1994:  1 Nc3 d5 2 e4 dxe4 3 Nxe4 e5 4 Bc4 Nf6?
5 Ng5 Nd5 6 d4 Bb4+ 7 c3 Nxc3 8 Bxf7+ Ke7 9 bxc3 Bxc3+ 10 Kf1 Bxa1 11 Ba3+ Kd7
12 dxe5+ Kc6 13 Qf3+ Kb6 14 Qb3+ Kc6 15 Qc4+, Resigns.

Cox-Johnson, Bethlehem 1994, Caro-Kann Defense:  1 e4 c6 2 d4 d5 3 Nc3 dxe4
4 Nxe4 Nf6 5 Nxf6+ gxf6 6 Be3 Bf5 7 g4? Be4 8 f3 Bg6 9 Ne2 Na6 10 Nf4 Nb4
11 Ng6 hxg6 12 c3 Nd5 13 Bd2 Qd6 14 Qe2 Bh6 15 O-O-O Bxd2+ 16 Rd2 g5
17 Kb1 O-O-O 18 Qf2 Rh4 19 Bc4 Rdh8 20 Qe1 e6 21 Bxd5 Qxd5 22 Rf2 Qd6 23 Qg1 Kb8
24 a3 Qf4 25 Qc1? Rxg4 26 fxg4 Qxf2 27 Rf1 Qg2 28 Rxf6 Rxh2 29 Ka2 Qd5+
30 c4 Qd4 31 Rxf7 a5 32 Re7 e5 33 Rg7 e4 34 Rxg5 e3 35 Rf5 Rd2, Resigns.