Jaki you are sooo pleasant. Thank's for the prolongation
After two trials, now I take my third chance. Others not seen to play chess and our champion @freiheit50 seems to be also busy in another matter. Take my third strike (no computers aid needed): c3-b5!, a6xb5 (c6xb5 is obsolete), e4-e5 and neither white nor black is able to win the game. So easy…
You are viewing a single comment's thread from:
Congratulations!
Yes, white has to play 1. Nb5+! . Black has now only two possible moves: After 1. ... cxb5? he will be checkmated by 2. c6# (check of the bishop). Therefore the only possibility is: 1. Nb5+ axb5. But then white plays 2. e5 and the position is blocked forever. As the black bishop is moving on the light squares the own pawns are like an impenetrable fence. White only need to wait and move its king around.
Why computers cannot find this solution? Because there are still many possible black moves. They cannot recognize though that there is no chance (and also will not come in future) to break through the pawn chain. In this position it is important to recognize an unchangeable structure on the board instead to calculate very fast.
Great blocking strategy! I guess you got inspired by the famous chess problem of Rudolph, W., where 1. Ba4+! saves the draw:
Yes, indeed!
And the computer will never end the game as drawn? That's not good.
The game will end because of the "fifty-move rule":
The chess software will recognize it cannot win some moves before the fifty-move limit is reached.
As I said, normally chess programs play great chess ... This chess problem was designed to point out an amazing weakness of current chess software (as you proved yourself humans are much better in this special kind of positions). https://en.wikipedia.org/wiki/Fifty-move_rule
Did you fiddle out this funny disfunction or is it a common knowledge of chess professionals?
This special weakness of chess software is well known by insiders. But the position I gave you was created by myself (I managed to combine the feature of a blocked position with a tactical sacrifice of the knight on a square where it can be taken by two different pawns).
The problem in a real game is that the human player very rarely reaches these kind of positions which the software doesn't understand.
I did explain above why the engines cannot find it. They don't recognize long lasting structures and patterns. They are unbeatable when they have to find long forced tactical variations but here after the second white move black can still long time shuffle around its pieces and try many maneuvers, but all that doesn't change the fact of the impenetrable pawn chain.
all current rating lists of chess programs.
But of course I cannot guarantee that there is no single engine which could solve the problem (there are many hundred chess programs available nowadays). I do have in mind two engines with detectors of blocked positions: you could try "Crafty" and / or "Shredder" and lets see if they may solve the problem. I tried the strongest available chess engine (Stockfish, https://stockfishchess.org/), which is leading