Click HERE if you want
to purchase a chess book ... or other chess stuff. 
A.J.
Goldsby I (2200)  Rick Aeria (2035)


What follows can be called a bolt
from the clear, blue sky.
(They often used to say this about some
of Mikhail Tal's combinations!)
21.Nf5+!!, (Maybe  '!!!/!!!!')
{Diagram?}
A player sitting next to me nearly fell
out of his chair when I played this.
(He actually slipped and came partially
out of his seat.).
This is easily one of the most profound
moves I have ever played. It is also
one of the most surprising.
As a test, I offered this position to several strong computer programs.
I set each to search 3 nodes, set the hash tables to 25 MB, and then let them
crank. At least five of them did not list this move in their top three move choices
after exactly seven minutes of analysis. (Seven minutes being the limit set by
expert programmers for the average program, running on a Pentium III, to find
even the most astounding and amazing moves from known brilliancies.)
{This was true right after the game ...
it may NOT be true as the overall
technology of computers and the
associated chess programs
continually
advance!}
This shows that this move can not be completely calculated, at least not by
a human anyway. This means that a player making a move like this will have
to lean heavily on his experience,
judgment, and intuition.
I did do a great amount of calculation over the board and I was fairly certain
that Black could not play PxN; at any point, however.
(I also spent around 45 minutes on this move.)
[ White could not play:
21.000??
e5;
22.Ne2??
Nxb3+;
("/+")
{Diagram?} which wins White's Queen. ].
21...Kg8!; (Maybe  '!?')
The best move?
(I think it is.)
[ The alternatives were:
Var. # 1.)
21...Kf6!?; This looks very dangerous, and would be easy to reject on
general principle. But over the board I was unable to find a clear and
convincing refutation.
22.Qg5+, The computer says this is the most
accurate move, but I am not at all sure.
The alternatives to 22. Qg5+ are:
a.)
22.Rd1!?
Ne6;
(Definitely not: 22...gxf5??;
23.Qg5+
Ke6; 24.Qxf5#).
23.Ne3
Kg7; 24.Rh3!?;
(Or 24.Nc4!?; or 24.Nd5!?; or
24.g4!?)
24...Qb6; 25.Rg3
Qc5; 26.Kf1!?,
"~" (Or 26.Nf5+!?);
b.) 22.Qd4+!?
Ke6; 23.Ne3
Kd7;
24.e5,
">" ("With an attack.")
(Or 24.Nc4!?);
c.) 22.Nd4
Kg7; 23.Rh3;
(Or 23.Nf5+!?
Kf6; 24.Nd4
Kg7; 25.Nf5+,
with a draw
by repetition.). 23...Kh7; 24.Rg3,
"~" {"Unclear."}
Back to 21...Kf6; 22.Qg5+, and now:
22...Ke6; 23.Rd1,
"+/=" when White is probably a little better.
(I certainly would be at least a little uncomfortable with my King position,
if I were playing the Black
pieces!)
***
Or:
Var. # 2.)
21...gxf5?!; Maybe  '??'
(This move is simply too dangerous.)
22.Qg5+
Kh7; 23.Qxh5+
Kg7;
24.Qg5+
Kh7; This looks
forced.
Black
should probably NOT play: 24...Kh8;
25.Qh6+,
(25.Rh3!?)
25...Kg8; 26.Qg5+,
(Maybe also playable are: 26.exf5!?; or
26.Rh3!?),
26...Kh8;
"~" {"Unclear?"}
25.Qxf5+ Kg7; 26.Qg5+ Kh8; Forced again?
A little
too risky is: 26...Kh7!?; 27.Rh3
Nxe4;
( Definitely not 27...Rg8?;
28.Qh5+
Kg7; 29.Rg3+
Kf8; 30.Qh6+
Ke8; 31.Rxg8+,
and White is
winning easily. ("+/") ). 28.Qh5+
Kg7; 29.Qg4+
Kh8;
30.fxe4,
("+/")
and White should be
winning.
27.Rh3
Nxe4;
Practically forced.
But NOT:
27...Rg8??; 28.Qh6#.
Or 27...f6?; ('??') 28.Qh5+
Kg7;
29.Rg3#
28.Qh5+
Kg7; 29.Qg4+
Kh6;
30.fxe4, "+/"
(Maybe ... "+/")
when White may just be winning.
***
Or: Var. # 3.)
Definitely NOT: 21...Kh7??; 22.Qh6+
Kg8; and now White
can
play ... 23.Qg7#,
and Black is mated. ].
White's next move is a nice, but not unheard of, Rook luft.
22. Rh3!, (Maybe  '!!')
A very good move ... and judging
from the look on my opponent's
face 
it was a complete surprise.
[ 22.Qg5 ].
22...Ne6; (Maybe  '!')
{Diagram?}
Time and time again, my opponent plays some very good defensive moves.
[ Very bad is: 22...gxf5??;
23.Qg5+
Kh7;
(Maybe even worse is: 23...Kh8;
24.Qh6+
Kg8; 25.Rg3#.)
24.Qxh5+
Kg7;
(Or 24...Kg8; 25.Rg3#)
25.Rg3+
Kf6; 26.Qxf5#
].
23. Rg3
Rc5!;
24.Qh6!,
{Diagram?}
Beginning what I thought was the
final phase of the attack.
[ Also interesting was:
24.000!?
Rfc8; 25.Qh6
Rxc2+;
26.Kb1, "White has Compensation for
the material sacrificed." ].
24...Rxf5!!; (Very
nice!) {Diagram?}
I saw this sacrifice but had completely missed the reason for the followup.
I honestly thought this was one of those, "lastgasp, desperation" type of
sacrifices that a player makes just before resignation.
[ One alternative is: 24...Qd7!?;
25.000
Rfc8; 26.Rd2,
"/\" ("Initiative.")
But not: 24...Rxc2?!; 25.Rxg6+!!
fxg6;
26.Qxg6+
Kh8; 27.Qh6+!
Kg8;
28.Qxe6+
Kh7; This looks
forced. (But not 28...Rf7;
29.Nh6+, "+/")
29.Qh6+
Kg8; 30.Qg7#,
{Analysis diagram.}
This is one of the many lines I
had actually analyzed over the board. ].
25. exf5
Qd5!;
Now Black regains all his material.
{I had seen this move also, but NOT all of the consequences. If White
tries to play PxP/g6??, ('?'); he might lose ... or just get mated!}
[ 25...Nc5??; 26.fxg6, "+/" ].
26. fxe6[],
(Forced?)
I thought this to be the only winning try.
[ Very bad is: 26.fxg6?
(Maybe
 '??') This looks natural
but it fails
totally and completely. 26...Qe5+;
27.Kf2
Qd4+; 28.Ke1,
(Not pretty.)
White could also try:
a.) 28.Kf1
Qxa1+; 29.Kf2
Qd4+; 30.Ke1
fxg6; 31.Qxg6+
Ng7,
"/+"
b.) 28.Ke2 Nf4+, "/+"
c.)
28.Qe3
Qxa1; 29.Qh6;
(29.gxf7+
Kxf7; "/+")
29...Qd4+;
30.Kf1
Qd1+;
31.Kf2
Qxc2+; 32.Kg1
fxg6;
33.Rxg6+
Kf7;
34.Rxe6
Rg8;
"/+"
After 28. Ke1, we have:
28...Qxa1+; 29.Ke2
Qe5+;
30.Kf1, (Hmmm.)
(Or 30.Kf2
Qd4+; 31.Kf1
fxg6; "/+" is pretty much the same thing.)
30...fxg6;
"/+" and Black, NOT White ... wins easily.
***
But definitely not:
26.Rd1?
Qxf5; 27.Rd2
Rc8; "=/+" Black has ample
material compensation for the exchange. A Pawn and a Knight vs. a Rook
and a great deal of play. (Also, White has almost zero coordination between
his pieces.) My instincts told me this was a dead end. The computers verify
that my intuition was probably correct.
***
I completely rejected 26.Rc1, "~" as being too passive. ].
26...Qe5+; 27.
Kf2, {Diagram?}
This is forced.
[ 27.Kd2
Qd4+; 28.Ke2
Qe5+;
29.Kd2
Qd4+; 30.Ke2
Qe5+, etc.
This is, of course, a draw by
repetition. ].
27...Qd4+; 28.
Qe3[], {Diagram?}
This move has got to be forced.
[ 28.Ke2!? Qe5+; 29.Kf2 Qd4+; only repeats the position.].
28...Qxa1; 29.
Qe4!, (Nice.)
Threatening PxP/f7+, followed
by QxP/g6+, winning.
29...Qf6!?;
{Diagram?}
Time and time again, Black
plays the best defense.
[ Over the board, I thought I saw a win for White after:
29...f5!?; (Maybe  '!'
Hmmm.) 30.Qc4,
I guess this is the best.
Some of the alternatives for White are:
a.)
Not: 30.Rxg6+??
Kh7; 31.Qf4
Kxg6; 32.Qg5+
Kh7; 33.Qxh5+
Kg8;
34.Qg6+
Qg7; "/+"
Black gets back to defend, and wins easily.
b.) or 30.Qd5!? Rf6; 31.f4, "~" {Probably "unclear"}
c.) But not:
30.Qf4?
Rf6; 31.Qh6
Qd4+; 32.Kf1
Qd1+; 33.Kf2
Qxc2+;
34.Kg1
f4; "=/+" ...
when maybe Black is a little better here.
Returning to the main analysis line of, 29...f5!?; 30. Qc4, we have:
30...Qf6; 31.Rg5!?,
"~" but could not find it at home.
(This position might be, "+/=", or
even equal.)
***
Also interesting is:
31.Rh3!?, "~" A
fascinating position. Five different
strong programs gave five different evaluations of this position!
Also probably playable is: 31.Kg1!? {Unclear?}
***
Definitely not: 29...Qb2?; (??)
30.exf7+
Rxf7; 31.Qxg6+
Rg7;
(31...Qg7??; 32.Qxh5,
"+/") 32.Qe8+
Kh7; 33.Qxh5+
Kg8;
34.Qe8+
Kh7; 35.Rxg7+
Qxg7;
36.Qh5+
Kg8; 37.Qg5,
"+/" When
White should win the
King & Pawn ending without
too much trouble.].
30. exf7+
Kxf7; 31.Rg5!
e5[]; {Diagram?}
I was threatening to take Black's apawn.
[ Not: 31...Qe6?;
32.Qxe6+
Kxe6;
33.Ke2!,
"+/"
and White will win a pawn. (Maybe
more.) ].
Now comes a sequence which must be calculated very, very accurately.
Common sense would say with a Queen and a Rook on the board for each
side, that one misstep could be fatal.
32. Qd5+!
Kg7; 33.
Qxa5
Qf4; 34.
Qc7+
Rf7; 35.
Qxd6
Qxh4+;
36. Rg3
Qd4+; {Diagram?}
Pretty much forced. White's Queen was threatening b4, e5, and g6;
all at the same time.
[ Not: 36...Qf6?!; 37.Qxb4,
"+/" (or "+/") Or ...
Another bad line is: 36...Rf6?;
37.Qe7+
Kh6; 38.Qxe5,
"+/" ].
37. Qxd4
exd4; 38.
Ke2!, {Diagram?}
The only move.
I have to guard c2.
[ 38.Rg5?!, (Maybe  '?') 38...Rc7; 39.Rb5, "+/=" ].
38...Rc7; 39.
Kd2
Kf6; {Diagram?}
Keeping the King active seems
like the best bet.
[ Interesting is: 39...Rc3!?; but White comes out on top after  40.Rh3!,
(40.a5?
d3; 41.cxd3
Rxb3; "~" is less clear.)
and now 40...d3;
41.f4
dxc2; 42.Rxc3
bxc3+; 43.Kxc2,
"+/" and White should win
...
the K + P ending relatively easily. ].
40. f4!,
{Diagram?}
The most accurate.
This move is a violation of principle, as White exchanges pawns and breaks
up his own pawn formation. But in the process, he activates his Rook!
[ Not as good is: 40.Rh3!?
Rc3;
41.Rh1
d3; 42.cxd3
Rxb3;
43.Rh4, "+/" {Diagram?}
White should be better, but it
is
not 100% clear that he is winning. ].
40...Kf5; 41.
Rg5+!, {Diagram?}
Keeping the Rook active.
[ Horrible is: 41.Rf3?,
making the most important piece in this ending,
a bored bystander. ].
41...Kxf4; 42.
Rxg6,
"+/" {decisive}
White is winning now.
42...h4!;
{Diagram?}
Stubborn defense.
43. Rg8!, (correct)
White puts Black in Zugzwang.
[ A line I calculated over the board was:
43.Rb6?
Ke4; 44.Rxb4
Rg7;
45.a5
Rxg2+; 46.Kc1
h3, "/+" Black
is probably much better! ].
43...Rc5; 44.
Rd8!
Ke4; {Diagram?} ("Never
say die!")
Black puts a lot resistance into this ending, even offering me a draw right
around here.
[ 44...Ke5; 45.Rh8 Rc3; 46.Rh5+ Ke6; 47.Rxh4, "+/" ].
45. Rh8!, {Diagram?}
The most accurate.
[ Not 45.Rb8
Rg5; 46.Rxb4??,
(46.Re8+!?).
46...Rxg2+; 47.Kc1
h3;
48.Rb8
h2; 49.Rh8
Rg1+;
50.Kb2
h1Q; 51.Rxh1
Rxh1;
"/+"
as Black should win. ].
45...Rg5; 46.
Rxh4+
Kd5;
White's Rook is now forced into total passivity.
47. Rh2,
My Rook is forced to a very poor square.
47...Ke4; Deeefence!!
Black continues to defend this
ending very staunchly.
48. a5!,
A nice decoy sacrifice.
48...Kd5;
Black gets his King closer to the dangerous aPawn.
[ Definitely NOT : 48...Rxa5??;
49.Rh4+
Kf5; 50.Rh5+,
(White wins the Black Rook.)
Analysis Position after, 50...Rh5+. 
"+/" (The "Skewer," or "XRay" Attack.) ].
49. a6, (Maybe  '!')
"White's got a runner."
(streetcorner chess)
White's apawn can actually be
attacked and won by Black. But
this would
be a decoy action, as
the time consumed in this effort
would lose the game
for Black.
[ Poor is: 49.Kd3?
Rg3+; 50.Kd2
Rg6;
51.Kd3??
Rg3+;
52.Kd2
Rg6; "=" {Diagram?}
which is a draw by repetition. ].
49...Rg7;
Black is almost in Zugzwang.
[ Winning the apawn costs too
much time. I.e.,
49...Rg6; 50.a7
Ra6;
51.Kd3!
Rxa7; 52.Rh5+
Ke6;
53.Kxd4, "+/"
].
50. Rh5+
Ke4; 51.
Rh4+
Kd5; 52.
g4,
{Diagram?}
A very important advance, giving
White some muchneeded space.
52...Ke5;
Black slowly runs out of good moves.
[Black is being forced back by Zugzwang, or the lack of good moves.]
[ 52...Kc6;  '?!/?'
53.Kd3,
"+/" Or 52...Ke4;
53.g5+
Kd5; {Box?}
(Probably bad is: 53...Ke5;
54.Kd3
Rxg5; 55.a7
Rg3+; 56.Kc4
Rg8;
57.Rh5+
Ke4; 58.Ra5,
"+/") 54.Kd3
Rxg5; 55.Rxd4+
Kc5;
56.a7!, ("+/") This wins for White. ].
53. Kd3
Rd7; 54.
g5!
Kf5; 55.
Rxd4
Ra7; 56.
Rd5+; and ...
Black [finally] Resigns. (10)
[ 56.Rxb4!? ]
I will play 57. Ra5, and then march my King up and win Black's bPawn.
The reason I was relieved when Black finally resigned is that he seemed very determined not to lose. He used almost all of his time in both time controls. At one point Black's resolve could almost be felt, like a tangible force. It took a nearly superhuman effort for me to overcome it. (And some good technique!)
A game both players can be proud of.
Copyright (c) {LM} A.J. Goldsby I
Copyright (c) A.J. Goldsby, 19952008.
Copyright © A.J. Goldsby, 2009. All rights
reserved.