|
No.1349 |
Original Fairy problems |
|
No.1349 Arno Tüngler |
Solution: (click to show/hide) |
|
white Ke8 Pa7d3d4f3h2h5
black Kd5 Qg7 Ba1b1 Rb2e3 Se6 Pc4c6f4g3h4h6
ser-!=66 (7+13) C? |
|
No.1349 |
Original Fairy problems |
|
No.1349 Arno Tüngler |
Solution: (click to show/hide) |
|
white Ke8 Pa7d3d4f3h2h5
black Kd5 Qg7 Ba1b1 Rb2e3 Se6 Pc4c6f4g3h4h6
ser-!=66 (7+13) C? |
what is the cook when the a-pawn is lower than a6 ?
For example: 1.Ke8 2.hxg3 3.gxf4 7.f8=B 8.Bb4 13.f8=B 14.Bfe7 18.Ka6 20.Bxh6 21.Bf4 24.h8=R 26.Rxc6 27.Rxc4 29.Ra8 30.Bb8 32.Ba5-c7 33.Ka7 35.a6 !=
So, with wPa6 that should already not be possible.
Is it so important to say « without capture « ?
Is there already a longer position « with captures » ?
Yes, the longest has 173 moves with and 119 moves without promoted force. See PDB P1333509 and P1273550. In long series-movers captures are usually quite important to get small changes in the position that add length.
cook 1.hxg3 2.gxf4 3.f5 4.f4 5.f6 6.f5 7.f7 8.f6 9.fxg7 10.a8=R 11.Ra4 12.Rxc4 13.Rxc6 14.Rxe6 15.Kf8 16.Rxh6 17.Rh8 18.h6 19.h7 20.Rg8 21.h8=B!=
Thanks to Václav for the cook that I should have seen…
Would like to correct this with wBd7 instead of wPa7 and one move shorter – a ser-!=65. The solution remains unchanged besides after 1.Bc8 2. Ba6 everything is one move less. Maybe you can test again?
A full test of this version is not possible by either Alybadix or Gustav. I tested this sd-auto=65 with Gustav 4.2c with the additional condition that pawn promotions can only be on the bishop, and the problem passed this test! Even this partial test required over 60 hours of computer time.