[Event "Chicago Open"]
[Site "?"]
[Date "2007.05.26"]
[Round "2"]
[White "Brock, Bill"]
[Black "Nienart, Chris"]
[Result "0-1"]
[ECO "D94"]
[WhiteElo "2018"]
[BlackElo "2227"]
[Annotator "Bill Brock"]
[PlyCount "172"]

{After almost a year away from tournament play, I found myself in the awkward
position of having given up on USCF but very far from kicking my addiction to
chess, my patzerhood notwithstanding. Fortunately, there's this loophole
called "Tournament Membership" that allows one to play in an event without
becoming a USCF member. And so (after a fun warmup at an Evanston Action
event directed by Maret Thorpe), I found myself once again doing what I
usually do at the Chicago Open: making embarrassing blunders in the opening
(two this year!), winning ugly, and playing one or two decent games.... But
this year was different in one respect: I played two fascinating endgames,
both of which reached six-piece tablebase draws. However, my opponents, Chris
Nienart in Round 2 and Jim Dean in Round 5, applied tremendous pressure and I
lost both games after both players had used virtually all our alloted time. 
While I was pleased with my play in portions of these two games, Messrs.
Nienart and Dean have even more reason to be proud. Chess computer nerds
(and non-nerdy users such as you, gentle reader) are almost certainly familiar
with the Thompson and Nalimov databases that solve all five-piece endings. 
But the fact that all nontrivial SIX-piece endings have been solved, and are
readily available online, is less well known. (Five pieces against the bare
king is deemed "trivial.") You can follow along with, and perhaps correct, my
analysis by consulting the following website: http://www.shredderchess.com/
online-chess/online-databases/endgame-database.html Or simply Google
"Shredder endgame database"! Finally, thanks to Chris Nienart for
reconstructing the game score and his analysis of crucial variations.} 1. Nf3
d5 2. d4 Nf6 3. c4 c6 4. e3 {At least Kramnik and Topalov agree on something:
both used this move against their most hated opponent. Worried about
Chebanenko, Botvinnik/Anti-Botvinnik, Meran/Anti-Meran lines? Worry no more:
wimp out with 4.e3!} 4... g6 {A sensible choice.} ({In round 6, Tony Cao played
} 4... Bf5 {after which I bored him with} 5. Nc3 e6 6. Nh4 {
and an eventual draw.}) 5. Nc3 Bg7 6. b3 O-O ({
Black could try to blow open the long diagonal with} 6... c5 {
, but it doesn't look scary to me.}) 7. Bb2 {
"You are getting sleepy, sleeeepy..."} 7... Bf5 8. Be2 Ne4 9. O-O Nd7 10. cxd5
Nxc3 11. Bxc3 cxd5 12. Qd2 Bf6 13. Rfc1 Rc8 14. Ba5 b6 15. Rxc8 Qxc8 16. Rc1
Qb7 17. Bb4 Rc8 18. h3 h5 19. Rc3 Nb8 20. Ne5 Be6 21. Nd3 {
So White got a normal opening advantage after all: now Nf4 could be annoying.}
21... Bd7 {I was about 40 minutes ahead on the clock, a situation I'm not
accustomed to, especially when playing young whipper-snappers.} 22. Qc1 Rxc3
23. Qxc3 Bb5 {Trading his worst minor piece.} 24. a4 Bxd3 25. Bxd3 {I felt I ha
d a nice position: the open file (but what if Queens come off?) and the two
bishops (but with a symmetrical pawn structure, are the bishops that much
better?).} 25... Qc6 {Diagram #} 26. Bb5 $6 {Chris felt that this was my first
mistake: better to keep the queens on the board.} (26. Qd2) 26... Qxc3 27. Bxc3
Kf8 28. f3 a6 29. Be2 Ke8 {The Black king wants to walk over to the queenside
to help Mr. Knight protect the weak a- and b- pawns. So I had the bright idea
of attacking on the side being vacated.} 30. g4 hxg4 31. fxg4 ({If I'd understo
od that I wasn't getting the outside passed pawn of my dreams, I might have
preferred} 31. hxg4 {.}) 31... Kd7 32. Kg2 Kc6 33. Kg3 g5 $1 {Oops!} 34. e4 $6
{A cute but misguided idea: I'm trying to make the position more "bishopy" by
opening diagonals.} 34... e6 $1 (34... dxe4 35. Kf2 b5 36. axb5+ axb5 37. Ke3)
35. Kf3 b5 36. Ke3 Nd7 37. e5 ({In the last dozen moves, I had relaxed,
thinking that my small plusses would play themselves. Now that it's clear
that I've lost the thread of the game, maybe I should be grovelling for a draw
with} 37. axb5+ axb5 38. exd5+ exd5 39. Bb4 {. In the post mortem, Chris
criticized my allowing the space-gaining march of the b-pawn that now follows.}
) 37... Be7 38. axb5+ axb5 {
No more pawns on the a-file, the center is blocked: it's getting "knighty"!}
39. Bf1 Kb6 40. Be1 Nb8 41. Kd3 Nc6 42. Be2 b4 {Diagram # Very knighty. The
position is beginning to look like a French Defense gone horribly wrong for
White. Sometimes a bad bishop (the fellow on e7) can be very useful.) And
given the central pawn structure, isn't the "bad" bishop really a very good
bishop?} 43. Bd1 Kc7 44. Bc2 f6 45. exf6 ({
I did see that the apparently attractive opposite-color bishop ending after} 
45. Bg3 $2 fxe5 46. Bxe5+ Nxe5+ 47. dxe5 Bc5 $1 {was not very good for White,
but Chris showed me that it was absolutely winning for Black because of the
manuever Bc5-Bf2-Be1-Bc3-Be5!}) 45... Bxf6 46. Bf2 Kd6 47. Bd1 e5 48. dxe5+
Nxe5+ 49. Ke3 Nc6 50. Bg3+ Be5 51. Kf2 {"Aha!," I naively said to myself. "I
get the outside passed pawn after all! And I have the bishop!"} 51... Bxg3+
52. Kxg3 Ke5 53. h4 gxh4+ 54. Kxh4 d4 {Not Black's only possibility.} 55. Kg5 (
55. Bc2 $5) 55... Ne7 56. Kh6 d3 {Duh. Now I began to understand just how
pathetic my bishop is. It was not put on this earth to babysit the b3 pawn. 
But that's the job I gave it....} 57. g5 {Diagram #} 57... d2 $6 ({Chris was a
little self-critical here, as the d3 pawn was taking two squares away from the
Lame One on d1. Pushing the pawn to d2 ironically gives a couple squares back
to Mr. Bishop. In the post-mortem showed me some amazing studylike variations
after} 57... Nf5+ 58. Kh7 Ne3 59. Bh5 {In some lines, Black could only win by
bringing the knight to a2 to hold the b4 pawn.} ({
But instead of 59.Bh5, the simple} 59. g6 {appears to draw!})) ({
Chris also suggested} 57... Kf4 58. g6 Ke3 59. g7 Kd2 60. Bh5 Kc1 61. Kh7 ({
But can White still hold here with something like} 61. Kg5 $1 d2 62. Kf6 Ng8+
63. Kf7 Nh6+ 64. Kg6 $5) 61... d2 62. g8=Q Nxg8 63. Kxg8 d1=Q 64. Bxd1 Kxd1)
58. Kg7 {Back in the game: destination, b4.} 58... Nf5+ ({
But Black still seems to be winning:} 58... Kf5 $3 {
Chris rattled off the following variation blindfold in an email to be} 59. Kf7
Ng6 60. Ke8 Nf4 61. Kd7 Nd3 62. Bc2 Kxg5 63. Kd6 (63. Kc6 Kf4 64. Kb5 Ke3)
63... Kf4 64. Kd5 Nb2 65. Kd4 Kf3 66. Kc5 Nd3+ 67. Kd4 Nc1 (67... Nf2 68. Kc4
Ke2 69. Kxb4 Ng4 70. Kc5 Ne3) 68. Kc4 Ke2 $1 69. Kxb4 Nd3+ 70. Kc3 Ne1) 59. Kf7
{Diagram # Take a few minutes to analyze this position: how many moves ahead
can you see?} 59... Ne3 60. Be2 ({During the game, we both saw that} 60. g6
Nxd1 61. g7 Nc3 62. g8=Q d1=Q 63. Qb8+ Qd6 {(Nienart) was hopeless for White.})
60... d1=Q 61. Bxd1 Nxd1 62. g6 Ne3 63. g7 Nf5 64. g8=N $1 {Box-tastic! One
doesn't often see a defensive underpromotion in practical play. This position
is a tablebase draw (if only we could access the tablebase during play...)}
64... Nd4 65. Ne7 $1 {(The exclams for White are Nunn Convention--the exclams
for Black are compliments to Chris's endgame mastery...)} 65... Kd6 66. Kf6 ({
Strangely,} 66. Ke8 {draws, too.}) 66... Nxb3 67. Nf5+ $1 Kc5 $1 {When the king
is three squares away from the enemy knight on rank or file, it's a little
domination. (The Big Domination is two squares away on a diagonal.) And
because knight pawns are near the edge of the board, knights have difficulty
defending against knight pawns.} 68. Ke5 $1 {Diagram #} 68... Na5 69. Ke4 b3
70. Kd3 Kb4 71. Ne3 Ka3 72. Nd1 Nb7 $1 {In his best patriarchal voice,
Dvoretsky writes, "The deflecting knight sacrifice is the almost universally
employed technique in [knight] endings." Chris knows this, and quickly
repositions his knight to a square in which more deflections may be possible. 
I began to worry about an eventual ...Nf2!} 73. Kc3 Nc5 {Diagram #} {From a pra
ctical perspective, this is the critical moment of the game. My move was
theoretically good enough to draw, but forced me to find a string of "only"
moves that I was unlikely to find over-the-board. Al Chow and Angelo Young
buttonholed me after the game, asking why I didn't simply play 74.Kd2!? I
was worried about a deflection sacrifice aimed against the Nd1 (the eventual ..
.Nf2), but I should have been even more worried about allowing the b3 pawn to
reach b2.} 74. Ne3 $6 (74. Kd2 Ne4+ 75. Kc1 Ka2 76. Nb2 $1 {is a reciprocal
zugzwang position: White to play would lose by force, but Black to play can
only draw.}) 74... Na4+ 75. Kd2 Nb6 $1 {Again, Chris repositions the knight to
set up a possible future deflection sacrifice.} 76. Nc2+ Ka2 77. Nb4+ Ka1 {
Diagram #} 78. Kc3 $2 {The losing move.} (78. Nc2+ $1 Kb1 79. Nb4 Na4 $5 80.
Nd3 $1 Ka1 {Diagram #} 81. Ke2 $3 (81. Kc1 $2 Nc3 $1 82. Kd2 Nd5 $1 83. Kd1 Nf4
84. Nc5 b2 85. Nb3+ Ka2 (85... Kb1 {
is actually one move faster per the database, but not a very human move)}) 86.
Nc1+ Kb1 87. Nb3 Ne6 88. Nd2+ Ka1 89. Nb3+ Ka2 90. Nc1+ Ka3 $1 91. Kc2 Nc5 $1
92. Kb1 Ne4 $1 93. Kc2 Nc3 {wins elegantly}) 81... Kb1 (81... Nc3+ 82. Ke1 $1 (
82. Ke3)) 82. Kd2 $5 ({The bizarre-looking} 82. Ke3 {also draws)}) 82... Ka1 (
82... Nb2 83. Nc5 Nc4+ 84. Kc3 b2 85. Nd3) 83. Ke2 $3) 78... b2 79. Nc2+ Ka2
80. Nb4+ Ka1 81. Nc2+ Kb1 82. Na3+ Kc1 83. Kb3 Nd7 84. Ka2 Nc5 85. Nb1 Ne4 86.
Na3 Nd2 {And I flagged in this hopeless position.} 0-1

[Event "?"]
[Site "Chicago"]
[Date "2007.05.27"]
[Round "5"]
[White "Dean, Jim"]
[Black "Brock, Bill"]
[Result "1-0"]
[ECO "D37"]
[WhiteElo "2250"]
[BlackElo "2018"]
[Annotator "Bill Brock"]
[PlyCount "168"]

1. d4 {0} 1... Nf6 2. c4 e6 3. Nc3 d5 4. Nf3 Be7 5. Bf4 O-O 6. e3 c5 7. dxc5
Bxc5 8. a3 Nc6 9. cxd5 Nxd5 (9... exd5) 10. Nxd5 exd5 11. Bd3 h6 (11... Bb6 12.
O-O d4 13. e4 Bg4 14. h3 Bh5 15. g4 Bg6) 12. O-O Be6 13. Rc1 Bd6 14. Bxd6 Qxd6
15. Bb1 Rac8 16. Qd3 {0} 16... g6 17. Nd4 Nxd4 18. Qxd4 {Diagram # I've been
outplayed in the opening: the Be6 is pathetic and White has better pawn
structure. So I start looking for counterplay....} 18... Rxc1 19. Rxc1 Rc8 $5
20. Rxc8+ {12} 20... Bxc8 {2} 21. h4 ({The point is that after} 21. Qxa7 {
Black can exploit White's temporary first-rank weakness with} 21... Qc6 22. Bd3
Qc1+ 23. Bf1 Qxb2 24. Qb8 Qc1 {and Black is almost equal.}) 21... Be6 22. Ba2 (
{And now} 22. Qxa7 Qc7 23. Bd3 Qc1+ 24. Bf1 Qxb2) 22... b6 23. Qf6 Qc7 24. g3 {
Having achieved equality, how can I screw it up?} 24... a5 $2 {
Jim doesn't miss the shot.} 25. e4 $1 Qc1+ ({Duh. The b-pawn hangs after} 
25... Qd7 26. exd5 Bxd5 27. Bxd5 Qxd5 28. Qxb6 {
The game continuation reaches a similar position.}) 26. Kh2 Qd2 27. Bxd5 Bxd5
28. exd5 Qxd5 29. Qxb6 a4 {
Forced: if White gets a passed pawn on the queenside, it will run very quickly.
} 30. Qf6 h5 31. Qf4 Qc5 32. Kg1 {Diagram #} 32... Qc6 $2 ({
During the game, I thought that} 32... Qc2 33. Qd4 {
was too passive, but it's harder for White to make progress than I thought:}
33... Qb3 34. Kg2 Qc2 35. Kf3 Qb3+ 36. Ke4 Qc2+ 37. Kd5 Qb3+ 38. Kc5 Kh7 39.
Qc3 Qb8 {might be messy for White (though it should still win).}) 33. b3 $1 {
Dean's second shot is fatal: now I'm completely busted.} ({Fritz says} 33. b4
$6 Qd5 {is also fine for White, but it's silly for a non-silicon brain to
allow Black to dream of counterplay against a3.}) 33... axb3 34. Qb8+ Kg7 35.
Qxb3 Qc1+ 36. Kg2 Qc6+ 37. Qf3 Qa4 38. Qc3+ Kh7 39. Qb4 Qc6+ 40. Kh2 Qf3 41.
Qb2 {Diagram # And now I went into a twenty-minute huddle.} 41... g5 $5 ({
Passive resistance is hopeless.} 41... Qa8 42. Qa2 Kg7 43. a4 Qa5 44. Qb2+ Kf8
45. Qb5 Qa7 46. Kg2 Qa8+ 47. Kg1 Ke7 48. a5 Qa7 49. a6 {
So I start playing for cheapos.}) 42. Qc2+ ({
I believe that White missed a clear win here with} 42. hxg5 $1 h4 43. Qf6 $1 {
for example:} 43... hxg3+ 44. Kh3 $1 ({A Fritz-move: the human} 44. fxg3 {
wins too}) 44... Qh1+ (44... Qxf6 45. gxf6 gxf2 46. Kg2 Kg6 47. a4 {wins}) 45.
Kxg3 Qg1+ 46. Kf3 Qd1+ 47. Kf4 Qc1+ 48. Kf5 Qc8+ 49. Ke4) 42... Kg7 43. hxg5
Qxa3 {Now all the pawns are on one side: good for the defense.} 44. Qf5 Qa1 45.
Kg2 Qa8+ $2 ({
Another careless move: Black has good drawing chances after the more accurate} 
45... Qd4 $1) 46. Qf3 $1 Qd8 47. Qxh5 {White is clearly winning again. But
because the queen had to be decentralized to eat the pawn, Blacks gets some
counterplay.} 47... Qd5+ 48. Kh2 Qf5 49. Qh6+ Kg8 50. Qf6 $1 Qh7+ $1 51. Kg2
Qe4+ 52. Qf3 Qe5 {
How does White stop the checks on the white squares and guard the g5 pawn?} 53.
Kh3 $1 {That's how!} 53... Qa1 (53... Qxg5 $2 54. Qg4 Qg7 55. Qxg7+ Kxg7 56.
Kg4 {is a won pawn ending.}) 54. Kg4 Qb1 55. Qf5 Qd1+ 56. Qf3 Qd4+ 57. Kh3 Qh8+
$1 58. Kg2 Qe5 59. Qa8+ Kg7 {Diagram #} 60. g6 $5 {
Perhaps not the best move, but an excellent practical choice.} 60... fxg6 (
60... Kxg6 $4 61. Qg8+ $1 Kf6 62. Qh8+ {with a won pawn ending.}) 61. Qa7+ Kg8
62. Qd7 Qe4+ 63. Kh3 Qe2 64. Qd5+ Kg7 65. Kg2 Qe1 66. Qd4+ Kg8 67. Kf3 Qh1+ 68.
Kf4 Qh6+ 69. Ke4 Qh1+ {Persistence is a virtue.} 70. f3 Qe1+ 71. Kf4 Qc1+ 72.
Qe3 {Diagram # At this point, I had a minute left, and my hand was itching to
play 72...g5. But it was safe to assume that Dean had not overlooked this
shot.} 72... g5+ $5 {
Cute, but is it best? Was it worth forty seconds? You tell me.} 73. Ke4 {
Forced.} 73... Qc6+ 74. Kf5 $1 {Forced.} (74. Ke5 $2 Qe8+ 75. Kd4 Qd7+ 76. Ke4
Qe6+ {
is a position that Black should be able to draw even with the flag hanging.})
74... Qd5+ 75. Kg4 Qd7+ 76. Kxg5 {
Diagram # And once again, we're in Nunn Convention land.....} 76... Qg7+ 77.
Kf5 Qf7+ (77... Qh7+ $11) 78. Kg4 Qd7+ $1 {
The key move: Black cannot allow White to sneak back into his shelter via h3.}
79. Kh4 Qh7+ $1 80. Kg4 Qd7+ $1 81. Kf4 Qf7+ 82. Ke4 Qe7+ 83. Kd3 $1 {
The only practical try, and a very good one.} 83... Qd6+ 84. Ke2 Qxg3 {
Diagram # and this position is a tablebase draw! But as a practical matter,
it's far from easy for Black to hold this position with four seconds left on
the clock, even with the delay. White cuts off the Black king on the back
rank and creates two concrete threats: either mate on g7, or an exchange of
queens on one of the many squares in front of the f3 pawn.} ({I'm frankly not
sure what the final moves of the game were, but the following is a reasonably
close approximation.} 84... Qxg3 85. Qe7 Qg2+ ({Maybe} 85... Kh8 {was a better
practical try, as it sets up the counterthreat of a kamikaze stalemate on g6
or f7.}) 86. Ke3 Qg1+ 87. Kf4 Qc1+ 88. Kf5 {I finally realize that I've done
nothing but improve White's position, and hide the queen from potential
exchanges on the f1 square.} 88... Qf1 89. f4 Qb1+ $2 {Now it's a tablebase
loss. Why? Don't ask me--queen and pawn endings are too hard.} (89... Qg1 $14
) (89... Qd1 $14) 90. Kf6 Qb6+ $2 {A blunder in an already lost position...}
91. Qe6+ {And I didn't need to use my remaining four seconds plus delay to
realize that the king and pawn ending was lost.}) 1-0