Antimarscirce: castlings are not executed

[thomas-maeder]

From yesterday's solving tourney in Andernach:

BeginProblem Stipulation h#2 Condition AntiMarsCirce Pieces black Kf6 Bg1 Rh5 Pe6d7 white Kf1 Rh1 Sh4 Bf5 EndProblem

Popeye only finds 1.Rh5h4 Bf5e6 2.Rh4-h8-e8 Rh1-h1-h6 # but neither the second author's solution 1.Rh5f5+ Kf1g1 2.Rf5-a8-e8 0-0 # nor my cook 1.0-0-0 Rh1-h2 2.R-h8-e8 R-a1-a8 #

[thomas-maeder]

This may be a duplicate of bug #489

[WalterL-wL]

Indeed this is a duplicate of my #489. How would you understand the wording of your indicated 3rd solution "...nor my cook: 1.0-0-0..."? Do you think the black castling is legal when BOTH pieces have to be reborn in advance in the SAME non-capturing move, although castling by definition is a king's move only? So, what would be the justification for the rook's advance rebirth? - Thanks!

[WalterL-wL]

Considering your III. solution with 1.0-0-0: 1.Kf6[→e8],Rh5[→a8]-0-0-0 Rh1[→h1]-h2 2.Rd8[→h8]-e8 Rh2[→a1]-a8#

this seems to result in four subsidiary cooks: a) 'coz 1... Rh1×Bg1 2.Rd8[→h8]-e8 Rg1[→a1]-a8# b+c) 'coz 1... Kf1[→e1]-d1,d2 2.Rd8[→h8]-e8 Bf5[→f1]-a6# d) 'coz 1... Kf1×Bg1 2.Rd8[→h8]-e8 Bf5[→f1]-a6#

[WalterL-wL]

Though I personally feel in Antimarscirce this kind of extensive interpretation of the scope of the king's move in castling (including advance-rebirth also of the rook) is pushed too far, it still is an interesting and provoking thought! Referring to my former comments (especially concerning those subsidiary cooks in III, also in equally adapting my w0-0 solution) I'd like to suggest a little modification of the diagram, this should result in 3 "clean" solutions with no cooks under this extensive interpretation: shift Rh1→b5! 3 Solutions: I) 1.Rh5×h4... 2... Rb5[→h1]-h6# II) 1.Rh5×f5... 2... Kg1[→e1],Rb5[→h1]-0-0# (!!) III) 1.Kf6[→e8],Rh5[→a8]-0-0-0 Rb5[→h1]-h2 2... Rh2[→a1]-a8#

[WalterL-wL]

Sorry, further amendments: With a diagram shift Rh1→b5 you'll even have a nice 4th solution: IV) 1.e6×Bf5 K×Bg1 2.Rh5[→a8]-e8 R×f5#

If you only want those former 3 solutions, then you'd better shift Rh1→d3

If shift Rh1→c4 you will only have 2 solutions (no b0-0-0 allowed)

If shift Rh1→a6(or)e2 there's another way to imply b0-0-0 in a 4th solution: IV) 1.Kf6[→e8],Rh5[→a8]-0-0-0 Ra6(or)e2×e6 2.Rd8[→h8]-e8 Re6×Re8#

[WalterL-wL]

A big, big sorry! The suggested versions wRh1→b5 as well as wRh1→d3 have cooks! Only the version wRh1→a6(or)e2 seems to show correct 4 solutions, though III and IV unfortunately start with the same move. Actually wRh1→c4 with an added +wPc5 should result in those correct 3 solutions in this extended interpretation of castling with Antimarscirce...

[thomas-maeder]

I called it "my cook" because I was the only participant who found it at the Andernach solving tourney.

The chess rules define castling as a king move in the context of the touch-move rule; i.e. if a player does an illegal castling, he is required to move his king instead, not the rook. This rule isn't relevant in the context of chess composition; e.g. in Madrasi, castling isn't possible with a paralysed rook.

[thomas-maeder]

this seems to result in four subsidiary cooks:

Indeed. I should have found R*g1 as alternative to Rh2; if I had given this as an additional solution, I might have won the Andernach solving tourney :-)

[WalterL-wL]

Thanks! As I understand you see your extended interpretation with this 3rd solution b0-0-0 as correct, so the original two solutions show a cook - I'm still undecided but rather open minded on this. I would fully agree if this will finally be a wider accepted interpretation, currently there still seems to be some opposite views on that, even some written reactions to Andernach 2024 seem to be announced... One could even refer to FIDE rule 3.8.2: "Castling - This is a move of the king and either rook of the same colour along the player’s first rank, counting as a single move of the king..."... but there's also this (orthodox) restriction "along the player's first rank"...? Of course FIDE doesn't set the legal context for chess composition insofar, but where's THE genuine link then?

Maybe there's another unresolved difference: In contrast to your comment, according to Märchenschachlexikon, in Madrasi castling with a paralysed rook IS explicitely allowed (because it's a king's move)! In Chamäleonschach castling doesn't change the rook to queen, because it is seen as a king's move. Whereas in Einsteinschach castling DOES change the rook to bishop (or queen in Reverse Einsteinschach). Especially this latter rule (in combination with Antimarscirce) might create some rather quirky moves of the rook as a proper participant in castling with such an extended interpretation, because a rook(a3) shouldn't be able to move non-capturing "as a rook" in his advance-rebirth to a1, because he would be changed to bishop in this advance-move - what kind of move should it be else!? So, would it require a queen(a3 or even a4???) to move to a1 (becoming a rook in this advance-move) for executing w0-0-0? And as you such define the rook's move as a separate subpart of the move, then after castling again the rook would automatically change to bishop in Einsteinschach... Lots of unclear consequences with such an extended interpretation in my opinion...

And to add another thought: If you treat both "parts" of the castling move in this extended way separately, then what about all the other combined moves, consisting of non-capturing part-moves, like e.g. double-move queen (Doppelzugdame), Oogaplex, the hurdle with Bulgarian/Dobrich hopper...? Would each one of the part-moves have to start with an advance-rebirth move of its own?

[WalterL-wL]

On second and third thoughts I want to withdraw my apprehension of quirky moves of a rook in a combination (Reverse) Einsteinschach + Antimarscirce in this extended interpretation of castling! The superordinate rule of Antimarscirce defines that rebirth already happens in advance of the underlying non-capturing move, and on reference to Marscirce the rebirth square is defined by the current square before rebirthing... So, only a w rook proper (dark square) can be advance-reborn on a1 for performing w0-0-0 (the rook would finally be changed to bishop / queen only after castling). No problem!