Hello folks.
Finally here is the third part of my BCM article entitled 'Computers and
Ideas'. Apologies for the even later delay.
This is the hard one. Needing desperately to win I examined the position below
and came up with a paradoxical idea. However, it needed my opponent's
co-operation. Can you find the idea? Rest assured that it will take time and a
critical study of certain important lines. What's more, don't be disappointed
if you don't find it. Many strong players that I showed this to also failed to
find the trick.
Happy studying.
Tyson
"There's a debate about "How far ahead can strong players see?" and various
numbers are bandied about with the importance of seeing one move further also
thrown. I think it's more representative to say that seeing one idea further is
important. My third example will demonstrate this point.
I belonged to the well-known National Correspondence Chess Club. Then computers
came along and it became apparent that some opponents were using their programs
to play their games. One Natcor member wrote scathingly of players who did this
in the club magazine 'En Passant'. He had my full support and upon meeting him
face-to-face I said I would write an anti-computer article. This is it!
White: Stan Cranmer
Black: AT Mordue
White: King f2, Rook d2, Bishop e2, Pawns a3, b2, e3, f4 and h2.
Black: King c6, Rook b8, Bishop e6, Pawns a4, c5, e4, f5 and h4.
At the time of this game Stan probably had an OTB grade between 160-175. It was
at this point I decided that my only chance of victory would arise if my
opponent tried winning it. Things to consider are the two open files and the
points at which the Black rook can penetrate. Also the queenside pawns are not
locked and in certain circumstances this is a significant factor. The weakness
of the b3 square is also to be noted. So your challenge is to find my concept
45...Bb3
46.Kg2 Rg8+
47.Kh3
Stan takes up the challenge. There's nothing wrong with winning the h4 pawn.
Meanwhile I made some forward moves.
47...Rg1
48.Kxh4 c4!?
Apparently locking in the bishop. Maybe Black intends .c4-c3 breaking up the
Queenside? After the text White can safely play 49 Kh3 and do nothing.
Alternatively he can activate his own Bishop with 49 Bh5 but if it then moves
down the h5-e8 diagonal then the e3 pawn is vulnerable to .Re1. Stan plays the
natural move.
49.Kh5 Kc5
50.h4?
Playing for the win yet in turn giving Black excellent winning chances. With 50
Kh4 White can remain passive and challenge Black to do something. However, even
after the text move many people that I showed this position to still didn't see
what my idea was.
50... Rg3
51.Kh6 Rxe3
52.Kg6
At the time I thought that 53 Kg7 was a more important move for a reason that
will become apparent. That's a subtle hint that this is your final chance to
work out what's really going on here. CM doesn't have a clue.
52...Rd3!!
One exclamation mark for this being a good move and another for the shock
value. This surprised nearly every human I showed it to. Stan was at first
surprised but subsequently delighted. CM was certainly shocked as it switches
its evaluation from +0.26 to -2.28! This is the square that I'd been aiming for
all along. Normally a Rook goes vertically on a file to occupy an outpost, yet
this one has come in horizontally!
White's rook is effectively trapped. The point of 45...Bb3 was to control the
c2 and d1 squares and .c5-c4 was needed for the double support of d3. Of course
the only way the Rook could gain safe access to d3 was via e3, to e3 via g3 and
only then if the h2 pawn had moved. Why would White want to do that? So he
could promote it of course. After all he was a pawn up and completely safe .
apparently. Now get a computer to come up with that idea!
So how did I come up with this concept? I noticed that back on move 45 the
White Rook though apparently active was actually short of squares. It was also
necessary for it to hold the second rank if ever the White king went walkies.
As explained above my idea only works if the Black rook can access the g3
square, so it was necessary to throw a carrot, the h4-pawn, to White so I could
get my stick in. Fortunately it worked, and the moves 45...Bb3 and 48...c4
probably did a lot to confuse White. They don't look as if they are parts of a
coherent plan.
If now 53 Rxd3 cxd3 then White's Bishop is overwhelmed by the Black Pawns so
Stan must take with the Bishop. After the recapture with the c-pawn Black has
two connected passed Pawns with close support of king and bishop.
53.Bxd3 cxd3
54.h5 Bg8!
Perfect timing. White has to lose tempi to take the prelate to ensure his own
pawn's promotion but Black has two passed pawns and can no longer be denied.
Our dull position has exploded into life. There are now plenty of lines but
they are all in Black's favour. "Initiative over material every time." Thanks
Keith.
Had White chosen 52 Kg7 then he would save a tempo if Black went 52...Rd3
immediately but 52...Rg3+ first throws a spanner in the works. If 53 Kf8 Rh3 54
h5 Rh2 and now 55...e3 is going to win a piece. Obviously moving back onto the
h-file is no help while 53 Kf7?? loses to the nasty discovery 53...c3+. That
leaves 53 Kf6 but now .Bg8 is back as a candidate move.
55.h6 Kc4
56.Kg7 e3
57.Kxg8
Preferring this to 57 Rxd3 Kxd3 58 Kxg8 e2 59 h7 e1=Q 60 h8=Q Qe8+! forcing a
easily winning pawn ending. I'd calculated this ages ago but at this point most
computers are reduced to hysterics. It's no good calculating every line five
moves deep if the decisive finesse comes twelve moves deep.
57...exd2
58.h7 d1=Q
59.h8=Q Kb3
With the d-file blocked Black no longer has the skewer. However, he has his
advanced Pawn and adequate cover for the King. White has nothing. Note that I
referred to the weakness of the b3 square earlier. First the bishop, now the
king .
60.Qe5 Qc2
61.Qb5+ Ka2
62.Qxf5 Qb3+
. and now the queen with decisive effect. Black is getting yet another queen so
it's time to draw the curtain.
0-1
Dare to imagine!"
