Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: a strange positional evaluation

Author: Simon Finn

Date: 16:49:54 10/01/01

Go up one level in this thread


On October 01, 2001 at 15:14:22, Dieter Buerssner wrote:

>Simon, thanks for all the analyses. I followed some of your lines, and also
>other lines - allways I could find a win with engine help.
>
>You seem to be a very strong chess player.

I haven't played OTB for years. When I did, I had a British grading
of about 200 (roughly 2200 FIDE).

> Perhaps you could give some hints.
>So, how to understand this (from an chess engine point of view)? You made one
>interesting comment about the undefended pawnd (g4 vs. g5). What other ideas
>would you have to take into consideration (by just looking at the pos). That the
>B has the color of the promotion square of the a-pawn? What else?

Widely-separated passed pawns. Centralized king. How advanced is the Black pawn?

Blockade is the main defence opposite bishop endings,
so play on both flanks is essential to have any winning chances at all.

In the original position, Black's advanced e-pawn is a major asset
because it (indirectly) restrains White's g-pawn. (If White pushes
the g-pawn, the Black king rushes back to the king-side and swaps
the g-pawn for the e-pawn.) This allows the Black king to wander
as far as the c-file without leaving the square of the g-pawn.
It also allows Black to counter-attack the a-pawn if the White
king goes to the king-side.

Perhaps Black loses only because his bishop is badly placed.
If it were on b6 (rather than d2), could White win? (I suspect not.)

If the Black pawn were on d4 rather than e3 then White could safely
push his g-pawn to g6 (with the bishop on d3) and the win would
be much easier (I think).

I originally thought that White could win fairly easily, but the position
turned out to be very tricky. White has to find a sequence of 14 "only"
moves (except that the first two can be played in either order) to prove
the win.

So it would be reasonable to evaluate the initial position as "probably drawn"
(but White has a lucky tactic!) but the final position should be evaluated
as a "probable win", even though it takes several more moves to actually pick up
any material.

In the final position, White's bishop can be placed on e2 where it
simultaneously defends both pawns (the a5 pawn can be pushed to a6)
and blockades the Black pawn. Then the White king can break through
to one side or the other and win the Black bishop.

The hard part is to distinguish between positions like

[D]8/b1k5/8/P2K4/6P1/4pB2/8/8 b - - 0 1

which White wins because there's no blockade, and positions like

[D]8/8/3k4/PK6/2Bb2P1/4p3/8/8 b - - 0 1

where White's attempts to break the blockade fail by a single tempo in lines
like:

1...Bc5 2. Ka6 Kc7 3. g5 Kd6 4. g6 Bd4 5. Kb7 Ke7 6. a6 Kf8 7. a7 Bxa7 8. Kxa7
Kg7 9. Bd3 e2 10. Bxe2 Kxg6

Simon








>
>Regards,
>Dieter



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.