Andreiko-Mogiljanski 1975

[Rein Halbersma]

<img src="http://fmjd.org/dias2/save/11855598524.png">
black to move: draw?

This diagram, from the game Andreiko-Mogiljanski 1975 (decision match Soviet championships), is given in Sijbrand's monography on Andreiko (p.361). Sijbrands thinks it's still a draw, mainly due the somewhat incovenient position of white's piece on 47. However, he claims that with a white piece on 48, the position might be lost for black.

<img src="http://fmjd.org/dias2/save/11855600245.png">
black to move: win for white?

Could Gerard or Ed run these positions against their databases?

Rein

[BertTuyt]

Rein, according to Damage 8-13 is a draw for Black (in the second diagram).
But this is based on 6-piece database search.
So im not sure if Ed or Gerard have similar results

Bert

[Rein Halbersma]

Rein, according to Damage 8-13 is a draw for Black (in the second diagram).
But this is based on 6-piece database search.
So im not sure if Ed or Gerard have similar results

Bert

do you get a "datase draw" for all variations, or do you get an "evaluation draw" (with e.g. a 7-piece position that you can't convert yet into the 6-piece database)? With Truus I also get 0.00 after 25 ply.

[BertTuyt]

Rein, i get an evaluation draw (score 0.0) after 25-26 ply.
Till the 0.00 score the score is positive for white.
So apparently black is able to reduce this no advantage, which for me is an indication that this is a draw.
But certainty is only possible when 7-piece databases are used (better is [img]images/smilies/icon_cool.gif[/img].

Bert

[TAILLE]

With only the 6 pieces endgame db Damy is not able to show a db draw. I obtain only an evaluation draw.

Supposing the result is really a draw I think it is almost impossible to prove this draw with only a 6 pieces endgame db because you cannot escape the variants where white sacrifices a man in order to make a king and and where black is able to make latter also a king. That way you obtain a 7 pieces endgame and the program will quickly eliminate this variant which give the advantage to black. The point is that it may happen that such variant is a win for white but it is too difficult to find it without the 7 pieces endgame db.

By the way I tried to find a win with white to move but here again I did not find it.

If white can play two moves the win exists :

Black to move : white wins

In the above diagram if you put the white man 48 on the 47 square it seems that the win disappears

Gérard

[Ed Gilbert]

Could Gerard or Ed run these positions against their databases?
Rein

Hi Rein,

Both of the positions are database draws in the 8-piece database.

-- Ed

[pontel]

Hi,

After almost 14 years, with the help of Kingsrow's 8-piece bases, we can categorically affirm that both positions are tied.

Greetings...