[Retros] Natch-2.5Beta

[Mario Richter]

Hello Pascal,


> rs1q1bsr/1ppkpppp/4p3/1K1b4/p7/8/PPPP1PPP/RSBQ1BSR

> 14 single moves. 15 + 16

> 

> The problem is caused by the moves Bc8-d7-c6-d5, necessary because of

> Pd7xe6, colliding with the move Ke8-d7.

> 

> If you play 7.Qd7 (or 7.Nd7) instead of 7.Kd7, there is no problem.

> The problem also disappears if the bK plays another move after passing

> by d7.  By adding 8.Bc4 Kd6, Natch finds again the solution.


Interestingly, if you raise the ply count to 16:

Natch 2.4      Copyright (C) 1997,98,99,2001-2005 Pascal Wassong
rs1q1bsr/1ppkpppp/4p3/1K1b4/p7/8/PPPP1PPP/RSBQ1BSR
16 single moves. 15 + 16

the first solution given by Natch 2.4 is:

Solution 1 (position 1)
1.Pe2-e3   Pa7-a6   2.Pe3-e4   Pa6-a5   3.Pe4-e5   Pa5-a4   
4.Pe5-e6   Pd7xe6   5.Ke1-e2   Bc8-d7   6.Ke2-d3   Bd7-c6   
7.Kd3-c4   Bc6-d5   8.Kc4-b5   Ke8-d7   


> The complete and technical explanation of this bug is a bit longer.


so I'm just wondering, if divisibilty by 4 or something similiar
did play a role ...

Best,

  mario