Die Schwalbe

1 problem(s) found in 6625 milliseconds (displaying 1 problem(s)). [PROBID='P0527781'] [download as LaTeX]

1 - P0527781
Fadil Abdurahmanovic
4664 feenschach 76 08/11/1985
In memoriam Albert H. Kniest
2. Preis
P0527781
(3+13) C+
h#3
2.1...
1) 1. Tc5 0-0 2. Dd7 Tf4 3. exf4 Te1#
2) 1. Lc5 0-0-0 2. Td7 Td4 3. exd4 Te1#
play all play one stop play next play all
VL: Only one castling is legal. Therefore presently (since 2009), due to Art.16 (3) of the Codex, this diagram, as is, represents a sound retro-problem (PRA) with one solution consisting of two partial ones.
Reprints: (13) StrateGems SG19, p.158, 07/2002 (2017-11-25)
Henrik Juel: ... so you might replace
'1)' by 'If Ta1 made last move', and
'2)' by 'If Th1 made last move'
If a skeptic asks: What if Ke1 made last move?
the answer is a reference to the ancient convention that a castling is considered legal unless it can be shown that it is illegal (2017-11-25)
A.Buchanan: A problem comprises diagram and stipulation. The stipulation says 2 solutions, so the problem is unsound under PRA, which would just have a single solution with two parts.
The convention says: "If...a solution is not possible according to the PRA convention..." Was this intended to mean here "if it's not possible to find a single h#3" or to mean: "if is not possible to resolve the problem including stipulation" i.e. to find *two* h#3. I assume the latter.
However if the stipulation had asked: "How many solutions?" then it would have altered from being RS to being PRA. I don't think this would be an issue, because the chess here would work the same way. (2018-03-25)
A.Buchanan: Arts 2 & 3 say: "A chess composition consists usually of a position on the chess board, a stipulation in the form of words and the solution, which usually consists of one or more sequences of moves which satisfy the stipulation." So the Codex appears to be using "solution" in the broader usage, which is consistent with the latter interpretation. (2018-03-25)
Yoav Ben-Zvi: As I understand it PRA means the solution is a "Strategy" (in the sense borrowed from Mathematical Game Theory) that contains a valid solution for every possible game that leads to the stated position. In plain language: when the game score is revealed the solver has a solution ready for it no matter which game score it is. If this is correct then there is no PRA solution because there is no solution in case White has lost both castling rights, which is a valid possibility. The fact that there is no circumstance in which White retains both castling rights means there is no case with a double solution but it does not change the fact that this problem has "No Solution".
I seem to be the only one to see it this way which makes it likely that there is something wrong in the above reasoning. I dont think Henrik's reply to the "Skeptic" points to such a flaw. (2018-03-26)
A.Buchanan: Here's how PRA was taught me in 2000 by composer, mountain climber and wonderful human being, Ronald Turnbull.
To understand the interaction of any castling and en passant rights, consider in turn each subset of the set of these variable moves Suppose this is the set of permissible moves. Is it consistent with the full application of the castling and en passant conventions? If no, we discard it; if yes we keep it and call it a 'part'. The problem divides into these independent parts, each to be solved separately.”
So e.g. here, S={wQc,wKc}. S itself cannot be the set of permitted moves, because at least one of the castlings must be impossible. {wQc} is maximally consistent, because you can't add {wKc}, although the castling convention tries to. So this is a "part". Similarly {wKc} is a "part". But the empty set {} is *not* a part, since you can add either {wKc} or {wQc}, without becoming inconsistent.
So there are two parts: {wQc} & {wKc}.
Please tell me if this makes sense, or how you would improve it. (2018-03-26)
VL: This problem is originally merely a helpmate with no retro: its hidden retro-content could be and was ignored. On the contrary, some problems with similar incompatible w castlings, stipulated appropriately (e.g., P0007077 by W.Keym) are retro-problems of the PRA-genre (PRA of a specific kind, sometimes called Typ Keym or Ad Libitum...). But the problems of this sort have never been considered previously (before Piran-58, after Piran and after Imola-73) as retro-strategic: that time, RS was formulated and understood more restrictedly (here I only can refer to
numerous available old articles devoted to RS or to RS vs PRA, etc.). Restipulated "one solution, pRA" (the corresponding version of) this problem would have become a sound retro-problem with the same but reinterpreted solution. See more details and similar examples in my
article "Retrovariants in helpmates with two white castlings" (StrateGems, 2002, SG19). Further, Art.16 adopted in Jurmala in 2007 made this problem, and similar ones, unsound. I was the first to revealed this unexpected
negative consequence, and I suggested a way (that has other, more important, advantages) to return their soundness, this time as RS-problems. It has been successfully implemented in the current wording of Art.16(3) (the main ordering idea goes back to my article "An order in controversial retro-genres", Shakhm. Kompoz., 1993, No.6, (in Russian)). Thus, we've luckily restored the soundness and enriched the content of such problems by
a retro-strategic scenario. However, instead, as before, one could restipulate "one solution" (already without the mark "pRA"), what, again, made this problem a sound retro-problem but now of the PRA-genre. (Here as always in such contexts, "this problem" is meant in the broad sense: it
literally or an appropriate version/variation of it.) There is a third competitive ways to leave this problem sound; namely, by adding a technical wP that could made the last move: such a version would preserve both the initial genre (purely helpmate) and the solution. In my opinion, all these options make sense although I myself prefer the first, newest and fortunate, retro-strategic one. (2018-03-29)
A.Buchanan: If we stick to the current conventions (1) this problem works fine as RS, or (2) we could change the stipulation to make it fit PRA (but why bother), or (3) we could change the diagram to make these retro concerns go away (again why bother). I view Keym's convention as a modus vivendi: to allow RS & PRA problems to appear without special stipulation (the price being paid by newcomers who aren't aware of what's going on). If the stipulation asked: "how many solutions?" then I agree that the problem would be both PRA & RS, and we would have to choose one. Keym's convention protects problems from conventions. (2018-03-29)
comment
Keywords: Sacrifice of white pieces, Reihen-Echo, Castling (wb)
Genre: h#
Computer test: Popeye C-Version 3.47 (1024 KB)
FEN: 2r5/q1r2b2/1b1pkp2/3ppp2/7p/p7/8/R3K2R
Reprints: 809 FIDE Album 1983-1985 1992
(13) StrateGems SG19, p. 158, 07/2002
Input: Hans-Jürgen Schäfer, 1998-03-11
Last update: Rainer Staudte, 2019-08-11 more...
Show statistic for complete result. Show search result faster by using ids.

https://pdb.dieschwalbe.de/search.jsp?expression=PROBID%3D%27P0527781%27

The problems of this query have been registered by the following contributors:

Hans-Jürgen Schäfer (1)