AMU: Mate is possible only in the case, when mating piece is attacked by the opposite piece. If the piece is attacked twice, then it loses again the ability to give mate. (or from Juraj Lörinc’s website: Condition AMU means that the piece making the mating move must be before this mating move attacked by exactly one opposite unit.)
No.1293Hubert Gockel Germany original – 30.03.2018
white Qa3 Ba8 Re5f8 Sg6 Bg3 Kg8 Pb6c7f5
black Kd6 Sb5c4 Qc2 Pc5d7e6
#2 (10+7)
AMU
1.R:e6#?? / Q:c5#?? / c8=S#??
These thematic mates are illegal because each mating piece is observed twice.
In each of three defences Black withdraws two observations of two pieces in cyclic fashion so that two mates mates should be possible at the same time. But thanks to cyclic dual avoidance only one is. (Author)
1.Sg6-e7 ! threat:
2.Se7-c8 #
1...Sc4*b6
2.Re5*e6 # {A (2.Q:c5+? B Q:c5!)}
1...Sb5-a7
2.Qa3*c5 # {B (2.c8=S+? C S:c8!)}
1...Kd6*e7
2.c7-c8=S # {C (2.R:e6+? C d:e6!)
(C+ by Popeye 4.79)}
Instructive example of cyclic dual avoidance and a new proof of AMU condition usability for various effects.
Geoff Foster
May 14, 2018 01:04
An interesting problem with a good flight-giving key. The dual avoidance uses 3 different mechanisms: opening of black line, guard of mating square, and move of King so that white move is not a double-check.
Geoff Foster
May 14, 2018 01:08
However it is a pity that after 1…Kxe7, 2.Qxc5 would not be mate anyway.
Instructive example of cyclic dual avoidance and a new proof of AMU condition usability for various effects.
An interesting problem with a good flight-giving key. The dual avoidance uses 3 different mechanisms: opening of black line, guard of mating square, and move of King so that white move is not a double-check.
However it is a pity that after 1…Kxe7, 2.Qxc5 would not be mate anyway.
Can d7 be shifted to f7?
@seetharaman: I don’t understand what you mean about shifting d7 to f7. The bPd7 is necessary in order to block the square d7.
I meant if a suitable guard could be found for d7.