# No.1201 (AB)

 No.1201 Andrew Buchanan (Гон Конг) Оригинальные сказочные задачи JF – 2017(I): январь – июнь

 No.1201 Andrew Buchanan Hong Kong original – 08.04.2017 Dedicated to Kjell Widlert Solution: (click to show/hide) white Kh3 Re1 Pb3c4e2g3g4 black Kg1 Rh1 Sf1 Bc5 Pc6d4e3f2h2g5 h=3                                      (7+9)  (C-) 1.d4-d3 b3-b4 2.d3-d2 b4*c5 3.d2-d1=Q Re1*d1 {= (C-; Popeye gives 3 sol. more d1=S/B/R (JV))} C+ Popeye 4.75, together with retro logic: We will use the following conventions: Article 16(2). An en passant capture on the first move is permitted only if it can be proved that the last move was the double step of the pawn which is to be captured. Article 17A. Unless expressly stipulated, the rule of dead position does not apply to the solution of chess compositions except for retro-problems. From Popeye's perspective, there are apparently 8 solutions: 1. d3 b4 2. d2 bxc5 3. d1=Q/S/R/B Rxd1= 1. dxc3ep b4 2. c2 bxc5 3. c1=Q/S/R/B Rxc1= However, evidently White has many potential last moves apart from the pawn double hop R: 0 ... c2-c4. Therefore we use Codex Article 16(2) to conclude that the e.p. capture is not permitted, and we can ignore the 4 apparent solutions which rely on it. Use of this convention is necessary, otherwise the problem has 4 cooks. (Even the Dead Position rule would not protect us from these, since 3. ... Rd1 is a living option whichever promoted officer is on c1.) So we do have a retro-problem on our hands. Therefore the Dead Position rule is turned on, by the lovely Codex Article 17A. So now let's turn to the 4 surviving apparent solutions, beginning 1. d3. 3. d1=Q! is alive. (3. ... Rxf1+ 4. Qxf1# is a win, even though 4. Kxf1= is just pat) Then 3. ... Rxd1=! 3. d1=R? is dead because 3. ... Rxd1= or 3. ... Rxf1+ 4. Rxf1=/Kxf1= are the only options, all drawn. 3. d1=B/S? is dead because 3. ... Rxd1= or 3. ... Rxf1+ 4. Kxf1= are the only options, all drawn. So in summary, there is just one real solution: 1. d3 b4 2. d2 bxc5 3. d1=Q Rxd1=, Queen Schnoebelen. This is a fairy problem because it is a help pat problem, in other respects it is orthodox retro. (Author) No.1201.1 Andrew Buchanan Hong Kong version of No.1201 – 15.04.2017 Dedicated to Kjell Widlert Solution: (click to show/hide) white Kh3 Re1 Pb2e2g4g3 black Kg1 Rh1 Bd1h6 Sf3f1 Pb3d2e3f2g7g6g5h2 h=3                                    (6+14)  (C-) 1...Re1*d1 2.Sf3-e1 Rd1*e1 3.d2-d1=Q Re1*d1 {= (C-; Popeye gives 3 sol. more d1=S/B/R (JV))} Solution: White has no last move (R: 0... g2-g3 would block wKB, one of 10 white units captured by bPs.) so it's White to move, and h=2.5. Therefore the Dead Position rule is turned on, by Codex Article 17A, and any drawing move must have alternative help play leading to a mate. According to Popeye, there are four solutions: 1. ... Rxd1 2.Se1 Rxe1 3.d1=Q/R/B/S Rxd1=, but if Black were not stalemated with the final move, only the Q promotion is enough to avert the looming stalemate of White by: 3. ... Rxf1+ 4. Qxf1#. So we have Queen Schnoebelen. Note that the blocked Bh6 prevents a cook: 1. ... Rxd1 2. Sh4 gxh4 3. gxh4 g5= (3. ... R~ is not a draw). Better than C-, through Popeye 4.75 + basic retro reasoning. (Author)

### 35 комментариев: No.1201 (AB)

1. Juraj Lörinc пишет:

I dare to say there is no retro element involved. E.p. solutions are excluded by codex and normally no one would consider them, seeing many possible last moves by White.

Also as a consequence, Dead Position rule must be explicitly stated under diagram, otherwise there are 3 cooking duals, ruining Schnoebelen Queen idea.

2. andrew buchanan пишет:

The diagram is missing bPg5, I don’t know why Julia removed it! I have explained that it is necessary. Imagine that it is there until it really comes back 🙂

Thanks for your comment, Juraj. The term “retro” has deliberately been left up in the air by Kjell, and I think that’s general the right decision because the meaning can’t be 100% nailed down for a good while. But it does mean that we can have debates like this.

Application of the codex rule requires (very slight) consideration of the history of the game: that makes this a retro problem.

The old “retro corner” site is a good a place to look for guidance. It states, reasonably: “Retrograde Analysis is a genre of chess problems where the legality of the position is a key element.” Here, the question of legality of the position without e.p. enabled is absolutely at the core of the problem.

I have another version of this which is help retractor then h=1. I guess you believe that retractors are retro? 🙂 But I think this is the right position to publicize because it does push us to think what the term “retro” means. This for me is comfortably in the retro zone. If the soundness of the solution did not depend upon the en passant status, then I would say it’s not a retro, but by construction it does.

Bottom line: just because retro logic is very easy doesn’t make it not retro.

Thanks again for your time, Juraj.

3. Julia пишет:

I’m sorry, I’ve missed bPg5 by mistake. Haven’t recognized it because testing without bPg5 gave the same result. Corrected now.

4. dupont пишет:

Dear Andrew,

Your point was probably to evacuate superfluous “solutions” via 2 different rules – en passant and dead reckoning. I can understand this option but the result looks a bit disharmonious to my mind – it is well-known and immediate that en passant capture is impossible in this setting, while the dead reckoning rule allowing a Schnoebelen Queen is the strong fact proved by the problem.

On another hand, it is generally admitted that “a Schnoebelen Queen is impossible to construct via an orthodox proof game”. Is it really true or is there some chance to break this old hat using the dead reckoning rule?

PS: it was so cool to meet you at Lille a couple of days ago!

5. Andrew Buchanan пишет:

Hi Nicolas,

I didn’t think of those two rules as harmonizing one another: just trying to get a functionally correct version with an ordinary kind of stipulation, and to provoke discussion on what is a retro-problem.

An alternative approach may allow for a double pawn hop to be provable, allowing e.p. and then DR. However the current matrix does not seem to accommodate that: wR is basically a loose cannon.

Proof games are already retro, so in one, is no need for the artificial en passant nonsense 🙂 Most of the pieces must be captured or blocked, and then for the rest you need to set up something like PDB’s P1261606. It’s a matter of generating a proof game to reach this position: the matrix is accurate and there are no philosophical issues.

6. Dmitri Turevski пишет:

I love the underlying logic of this problem.

But “ep / no ep” solved in favor of “no ep” is too cheap – it’s like having a thematic try in #2 with multifold non-thematic refutations.

Imagine, however, if there were several (except the pawn double-move) seemingly possible previous white moves all failing due to legality considerations, but one non-obvious, like, say, Kh4-h3 was ok. Nobody would doubt that’s a retro then.

7. Andrew Buchanan пишет:

Thanks for everyone’s comments – I really appreciate the feedback as I try to puzzle my way through the conventions. I agree with the consensus that the existing e.p. play is too cheap. Fortunately, I have found another version today with proper retro logic (not involving en passant), which I have sent to Julia, and I hope she can publish it soon 🙂

8. andrew buchanan пишет:

A typo has appeared in the stipulation of 1201.1, which should be h=3.0 not h=2.5. Hope Julia will have a chance to fix it soon, thanks 🙂

9. dupont пишет:

Really? The given solution is in 2.5…

10. Nikola Predrag пишет:

h=2.5 is incorrect, having 4 B3 moves/promotions.
Stipulation h=3 would make a joke problem, as though a helpstalemate could become a so called “retro problem”.

Dead position applies to retro-analysis and not to a “forward” play. That’s the pretty clear meaning of 17A.

But for a joke (or perhaps scorn?) we might interpret the rules as that the legality of ANY position should be proved by retro-analysis, which would make it a so called “retro problem”. And then to claim that any forward play is a part of such a “retro problem”, which might allow a distorted meaning of 17A.

Queen Schnoebelen requires applying the “expressly stipulated” part of 17A, just as Juraj had mentioned.

11. andrew buchanan пишет:

Hi Nicolas, Nikola!

Thanks for your responses!

This problem uses Codex Article 15: “First move. If the first move does not lie with the conventional party (examples see Footnote 9), this should either be indicated in the stipulation or deducible from retroanalysis.” So the stipulation must *not* indicate that White is actually the first player, so that the simple retroanalysis is necessary.

Article 15 formalizes a principle in frequent use since the 19th century, which has long since stopped being a joke or even a cliche, but is simply a tool.

As Kjell has left two articles which depend upon an undefined notion retro-problem, it’s normal that we should now enjoy exploring the question of what “counts” as retro.

Ultimately, the corner cases must surely be subjective. In 1201, I posed one of those corner-cases. The en passant convention is necessary in this problem hinging upon the retro-question. So I considered it to be retro, but I agree with you and Dmitri that at best it’s trivial and unharmonious. Juraj went further, as he feels it’s not even retro.

Hence I was happy to find 1201.1, where the simple retro-problem element is unquestionable. There is a real richness in problems which combine both forward and retro logic, sometimes called “retro-active” problems in another forum, and I think they have a big “future” (as well as a big past, haha).

So now I feel there is no need to inhabit the dubious areas. We might as well build our constructs on solid ground, one way of the other.

Similarly, verification of legality might be most of the point of a problem: that would make it retro. However, the rote checking that a position contains e.g. no illegal pawn mob is not sufficient to make a problem retro. So there is some subjectivity there, which means the area should be avoided for purposes of triggering 17 or 17A.

I like the new term “detective fiction” for problems which deliberately introduce a genuine element of retro in order to trigger Articles 17 or 17A. An example for 50M is PDB’s P1292681. This might be through retro-analysis or retro-stipulation.

Thanks again & all the best,
Andrew

12. dupont пишет:

Ah ok I get it now…

To show a Schoebelen Queen, it is necessary to use the DR rule, and to use the DR rule, it is necessary to have a retro problem, and to have a retro problem, it is necessary to have a “non-conventional” stipulation h=3.

Of course this is the last point which looks “suspicious”… You are using here your own interpretation of article 15, something like: “A help n.0 problem where white has no previous move is in fact a retro help n-1/2 problem”…

I’m not sure this is conform to the spirit of article 15. At least this interpretation would imply that each help n.0 problem where the author doesn’t care of an existing previous white move (and de-facto there is no such move) is cooked when the intended “solution” is beginning normally by black.

13. Nikola Predrag пишет:

Hi Andrew, I wasn’t clear enough.
I first answered to Nicolas that h=3 stip. indeed introduces the retro-analysis (as you say, through Article 15).

The joke is in the attempt to “trigger” a funny interpretation of the Article 17, as if the forward play should be considered as a retro-analysis?!

Retro stands for backward in my understanding so, where the Codex suggests that retro should mean forward?
(There is actually tremendous “subjectivity there” on the retro-forums)

Partial Retrograde Analysis (PRA) convention might allow the retro-analysis after some forward play, that’s where might be the “big future”.

14. Joost пишет:

I think it should be h=2.5. With the current stipulation, black must deliver stalemate on his 3rd move (halfduplex).

15. Andrew Buchanan пишет:

Hi Folks,

Thanks for your continuing interest!

@Nicolas,Joost: Werner Keym has a large number of Article 15 problems (often combined with castling & e.p. as “Drei Bosheiten”) in my favourite chess book, “Eigenartige Schachprobleme”). (“Bosheit” is apparently an archaic German word meaning “mean trick” as it might be used in a fairy story.)

The way that I have Article 15 working is conventional, and unsuspicious: the relevant footnotes in Codex which make this even clearer. Footnote 19 for example reads: “..the preliminary move by the unconventional party is not counted, except in help-play problems. The number of moves to be expressed in the stipulation should be the number of moves to be made by White.” Note that it says “help-play” not just “help-mate”.

There are about 100 such problems in PDB labeled as “No legal last move for White/Black” and I am sure that, as is usual in PDB, there are many more but they have not been identified as such.

There is no Codex back-up for Joost’s suggested interpretation as kind of “duplex”, although there are certainly retro problems for which this is the answer. Like rotation of the board, I would interpret it as still on the joke side of the fence.

@Nikola: I think you are stuck on a false dichotomy between “forward” and “retro”. A problem may have either or both of these keywords. A very few problems (e.g. 8 queens) may have neither. But there are great creative possibilities that come from combining the two. A new name for these is “retro-active” problems. Once again, Werner Keym is a master here, with e.g. many problems where forward mates depend upon the retroanalytic interaction between up to 4 multiple castlings.

PRA is one part of a bigger picture of how to conventionally resolve multiple “Bosheiten”. The central missed opportunity with Keym’s recent convention that prioritized PRA over RS is that it is too wedded to orthodox chess, and cuts off fairy chess, as the great theoreticians Guus Rol & Ronald Turnbull have pointed out.

There’s a fantastically creative fairy retro space available, but we need to clear up our thinking about the conventions in order to do so. The two versions of this problem here are just a couple of steps in this broader campaign.

All the best,
Andrew

16. Georgy Evseev пишет:

Let me add my 2 cents.

There are two different approaches in “classic” retro problems.

The first one is “hidden duplex” (typical for #1 stipulation): we are really to decide which side is able to fulfill stipulation.

The second one is “hidden halfmove”. The side to play is _not on move_ and we really start solving after the first move of opposite side, which goes unnumbered.

It looks like we have the last case in this problem. So, the stipulation should be h=2, shouldn’t it?
And the solution should be written:
0…Rxd1! 1.Se1 and so on…

• Andrew Buchanan пишет:

Hi George,

Here it’s sort of the second case, but whereas in direct play, there would indeed be an extra un-numbered move for Black, in help play there is a move *less* for Black, i.e. effectively the “set play”.

Footnote: 19: “For the purpose of Article 9, the preliminary move by the unconventional party is not counted, except in help-play problems. The number of moves to be expressed in the stipulation should be the number of moves to be made by White.” A good example is pdb.dieschwalbe.de/search.jsp?expression=PROBID=’P1068173′.

Hi Nikola,
Stop being rude.

• Nikola Predrag пишет:

Hi Andrew, I’m sorry if my words or some of them sound rude to you.
If I knew what in my approach appears as rude, I might try to “soften” it or just apologize and stop discussing.

Some words may have a rude meaning if you want to “hear” them as rude, regardless of the context. If their meaning is indeed always rude in any context, then I must apologize for my wrong understanding of English.

By nonsense I mean no-sense, having no (relevant) meaning, opposite to meaningful…

“…insisting on misuse…for misunderstanding” might sound rude if perceived as an imputation. I’m sorry and I do apologize if the effect of that isolated sentence was rude.
I just wasn’t aware of a possible meaning outside the context which includes the very next sentence.
Misused interpretation and misunderstanding is the main tool for jokes, and you’re playing with it in an interesting way (I said it’s interesting).
No rude intentions in it.

Anyway, tell me what else you see as rude.
I try to speak directly in order to avoid misunderstanding but this is often (mis)understood as rudeness. I discuss the topic and not the personalities.

17. Nikola Predrag пишет:

Andrew, I’m stuck on the dichotomy between meaningful concepts/words and nonsense.
Until you provide a different but convincing consistent meaning of retro-analysis, the Codex should remain stuck on the meaningful dichotomy between forward and retro, just to avoid the nonsense caused by meaningless words.

A “forum-definition” of “a retro problem” is not given in Codex but hypothetically it could be included one day if it would have a convincing meaning.

-“where the legality of the position is a key element”-
That would apply to any problem (except for the fairies) and would ruin the meaningful completeness of Codex, particularly the very point of 17&17A.

ou should expressly stipulate the dead position rule and everyone should be able to understand what to look for in your h=3, applying 15&17A.

The most essential purpose of Codex is guiding to our mutual understanding of our ideas.
The most fundamental for a “solution” of a “problem” is the “solver’s” clear understanding of the various rules that make the problem.

Apparently, you insist on MISusing the Codex for MISunderstanding.
This might be occasionally interesting if taken as a joke (and it is).

“Fantastic possibilities” are really fantastic when they create a world of deeply convincing consistency.
Trivial misinterpretations of the Codex may be funny but certainly not fantastic, since Codex clearly allows showing the “forward Schnoebelen Queen” by expressly stipulated DP.

It would be interesting and probably fantastic, if the retro-analytical part would already require “triggering” 17A. Then, the completeness of the problem would require the consistent rules for the whole play, applying DP rule from the beginning of the game to the very end.

If Codex doesn’t effectively apply DP rule from the beginning of the game to the diagram, it doesn’t actually trigger that rule.

18. andrew buchanan пишет:

I think the word “nonsense” is exaggerated and pejorative here. We may need to reason approximately for now, but that’s true of most human debate.

I asked Kjell in 2015 if he intended to supply a definition of “retro-problem”, and he said he would prefer to leave it open. In the absence of a precise definition, I still feel confident to say that most problems are definitely *not* retros, and most of the remainder definitely *are* retros, and neither category is particularly impacted by the existence of corner-cases which are harder to classify. Certainly there is no reason to particularly back off the fruitful area of retro + forward = retro-active compositions.

A personal rule of thumb is to ask: would the content of the problem be the same if we were playing a fairy version where game has just started at the current diagram? I.e. suppose the game has zero history?

I hope that eventually the conventions will be more mathematical, and as part of that we will enjoy a definition of ‘retro-problem’. Until that time, I am consoled by Wittgenstein’s demonstration that the term “game” cannot be defined precisely, but can only be approached through a series of cases. (Wittgenstein’s purpose here, by the way, was to give one argument why we cannot hope to define the concept of “love”.)

In the mean time, unless one is fascinated by legal reasoning, maybe best to stick the the clear cases of retros including (1) retractors of all kinds (2) proof games & other problems where the stipulation directly asks about the history of the game (3) illegal clusters (4) the retro-active case: problems where non-trivial reasoning usefully determines the validity of conditional moves, or who has the move.

There’s nothing special about Article 15: it’s just one way to introduce retro content. It’s helpful that it’s simple, clear and explicitly mentions retroanalysis. But there’s nothing special about it: there are now help-retractor versions of all 4 h=1 Schnoebelens. And there is no fundamental reason why Q Schnoebelen should not occur in a proof game, as Nicolas suggested.

19. Per Olin пишет:

The discussion has centered around important fundamental and technical matters, that I will not confuse with further comments. Somehow seems to have got into the background the fact that we here have a historical problem. As far as I understand, this is for the first time a queen Schnoebelen in an orthodox problem. A brilliant invention, we can only image the thrill the composer gets when the flash turns into a diagram. Congratulations!

20. Nikola Predrag пишет:

Yes Per, a good point, congratulations are well deserved for Schnoebelen Queen by Fide rules (which is Codex + DP expressly stipulated). That’s more than enough for appreciation.

Retro-analysis is irrelevant for that and thus superfluous & non-economical.

Anrew, a whole system/concept without a relevant distinctive meaning makes a NONSENSE (a merely technical term for such a system). Suggest a better term without that “pejorative tone”.

It’s not up to me to “prove the “meaninglessness”, it’s the claimer who has to define a hypothetical relevant distinctive meaning.

21. Andrew Buchanan пишет:

@Nikola,

Yes, I thought that apology wouldn’t last long 🙂 Our humble activity is perhaps as much art as science, but your responses made me think to look for some quotes by philosopher Thomas Kuhn:

“In science novelty emerges only with difficulty, manifested by resistance, against a background provided by expectation.”

“No part of the aim of normal science is to call forth new sorts of phenomena; indeed those that will not fit the box are often not seen at all. Nor do scientists normally aim to invent new theories, and they are often intolerant of those invented by others.”

22. Nikola Predrag пишет:

Andrew, I asked you to suggest another English word to replace the technical term which is so rude to you.
For something that doesn’t make sense, I use the term nonsense. I don’t see how a solution of h=3, which is “played forward”, could be considered as “retro-analysis”.
If the clear-cut meaning of “retro” is lost, the consequences make no sense.

Thinking “out of the box” requires some deeper and more general principles – which do make sense. But who knows what you mean by “science and new theories”, since you don’t show any curiosity about the critical points in your own theories. And that’s certainly not scientific.

Now I apologize again for the lack of proper expressions in my English. Without your cooperation and will to understand the essence of my words, I’m clearly disqualified.

23. Andrew buchanan пишет:

Hi Nikola,

To see problems which are both retro and forward, look at some of the over 1100 g=’retro’ and g=’h#’ compositions in PDB, some going back 100 years. Have a look at the list of composers whose work you are casually demeaning.

Anyway I’ve spent enough time responding to your naughtiness.

24. Nikola Predrag пишет:

Hi Andrew,
you didn’t respond to the question critical for the topic, the question which was my intention – the clear-cut meaning of “retro”.
Instead, it’s YOU who interprets and directs the discussion to the personalities.
Naughtiness and rudeness – these words pretty obviously have no intention to discuss the rules and logic, so take a look in a mirror.

25. Andrew Buchanan пишет:

Hi Nikola,

As I reported, there is deliberately no definition of retro in the conventions, and this isn’t considered s showstopper by the WFCC codex committee. I gave 4 examples of classes of problem, including the popular “retro-active” class, which by common accord fall into the retro category.

I am far from being a saint myself, but I do my best to learn. Recently I have made the friendship of a happy soul, Hilmar Ebert, who used to compose a lot, and who is now thriving in the Philippines, with a wonderful encyclopaedic site about Daoism (http://www.tao-te-king.org/). One beautiful quote from Zhuangzi is perhaps relevant here:

“In a river mist, if another boat knocks against yours, you might yell at the other fellow to stay clear. But if you notice then, that it’s an empty boat, adrift with nobody aboard, you stop yelling. When you discover that all the others are drifting boats, there’s no one to yell at. And when you find out you are an empty boat, there’s no one to yell.”

26. Nikola Predrag пишет:

The great majority of words in Codex have no “special definition” because they mean exactly what they usually mean. Retro-analysis means analyzing backwards from a certain moment. If there’s a chess diagram, it’s about the possible lines of chess play ending with the legal position(s) which the setting on the diagram possibly represents.
(Hypothetically, in some non-typical case, it might not be about the play, and something else could be retro-analyzed, providing it makes sense within the Codex.)

27. Andrew Buchanan пишет:

Hi Nikola,
I agree that retro-analysis means analyzing backwards. Kjell’s challenge to us is what to make of the subtly different term “retro-problem”: how much retro-analysis is necessary for a problem to become a retro-problem?
Maybe we should take this off-line if we are going further, rather than bother the good folks of JF. My email is at the bottom of the first page of my little website: http://anselan.com/chess.html.

28. Nikola Predrag пишет:

Hi Andrew,
yes, this is not a place for too wide and long talking. I hope I’ve sent you an e-mail (if I’ve solved the address-puzzle correctly).

But 🙂 , just a short post here.
“Kjell’s challenge” is primarily about what could be accepted for evaluation as a (sufficient/convincing) retro-content (how complex/deep is the retro-analysis). This is where subjectivity is expected from the public and, to some extent, from the judges.

However, the intrinsic RULES of a stipulation require a clear-cut distinction of what IS and what IS NOT a retro-problem, such that a solver knows what IS and what IS NOT possibly legal for the solution. No room for subjectivity about that.

Here are two similar simple examples with different stipulations – how the public should understand them:
1)
White Pa7 Ka6 Pc6 Pg6 Pb5 Pg5 Pg4
Black Ka8 Pc7 Pg7 Pb6 Ph5
#3 ??
2)
White Pb3 Pc2 Pg2 Ka1
Black Pg5 Pb4 Pg4 Ka3 Pc3 Pg3 Pa2
hs=5.5 ??

29. Andrew Buchanan пишет:

Hi Nikola,

I think your latest post is worth answering here.

First, I agree that there is a difference between aesthetic concerns and the determinate conventions, but by “Kjell’s challenge” I meant the latter: the fact that he is expressly *not* giving a definition of retro-problem. But if he isn’t worried about that, then neither am I.

On aesthetics: the DP rule is powerful but very context-dependent. Compositions involving this already have “one foot in the grave”. Your two positions are typical with both players near stalemate and one in near retro-stalemate: this configuration is almost the standard for a DP position. So DP is never going to run roughshod over chess problems, it’s a whimsical feature which can occasionally achieve some neat effects, but only in some near-dead positions.

I think you’re missing wPh4 from your second problem.

With your problems I think you are raising two excellent known points:

(1) In both problems, in order to get the e.p. convention to trigger, we already need DR to exclude the possibility that the last move was a single move by the hP. So is there a circularity of reasoning here? Are we assuming what we are meant to prove? I am certainly looking for candidate definitions for “retro-problem”, but maybe the real answer will have to wait for Guus Rol’s eventual book on retro convention fundamentals. For now, both to my mind are clearly “retro-problems”, but OK one might not want to put either in a solving tourney.

(2) The other easier issue only occurs with the second problem. The intended “solution” ends with wPg7 bBh8, and Black would be forced to play 6. … hxg7=, but the game dies because of DR. Any “retro-problem” which is s= or hs= will not be sound under Article 17A, even ones which do not involve DR explicitly (as this one does in order to drive the ep). Article 17A is known to be a compromise, and there are bound to be a few problems on both sides which don’t fit the rule. The whole point of 1201 was to repair a problem which was on the wrong side of the article.

30. Per Olin пишет:

The first position in post by Nikola April, 25 is very close to the following problem:

Andrey Frolkin & Per Olin
feenschach 185 Band XXXII
Januar – März 2011, page 26, nr 10180

White: Kh6 pawns d2, d3, b5, d5, g5, b6, f6, h7
Black: Kh8 pawns a5, b7, d7 f7, g6

White to move. Game result?
a) diagram (9+6 pieces)
b) remove pawn d2 (8+6 pieces)

Solution: a) 1.d6 a4 2.d4 a3 3.d5 a2 4.d4 a1D stalemate b) 1.bxa6 e.p.! d6 2.d4 (2.a7? or 2.axb7? stalemate) bxa6 3.b7 a5 4.b8D mate – Black’s last move has been either 0. – a6-a5 or 0. – a7-a5. In a) it cannot be proven that the last move has been 0. – a7-a5, so there is no right to e.p.-capture. The a) position is not a dead one, as Black wins if White makes the move 4.d3?. The b)-position would be a dead one, if the black a5-pawn came from a6. However, the a5-pawn cannot have come from a6, as also this previous position is a dead one; this would have ended the game immediately without permitting Black to move his pawn to a5. As the pawn must come from a7, White has the right to e.p.-capture resulting in a win for White.

Comment: the readers / solvers were not as overwhelmed as the composers…

31. Nikola Predrag пишет:

Yes Per, my example 1) shows indeed the same retro-point as twin b) in the problem you gave.

I gave ex.1) just as an introduction to ex.2) with essentially the same position (with reversed colors), but related to a different stipulation.
Please notice the correct Andrew’s comment that I’ve dropped wPh4.

The point is that in ex.2), the dead position occurs in the forward play that itself was legalized by applying DP-rule to retro-analysis.

Andrew,
The term “retro-problem” in Codex refers to retro-analysis of the possible lines of play before the diagram. I see not even a hint (and certainly no logic) that “retro” should mean anything else (particularly not “forward”).

Every orthodox composition consists of a retro-problem which determines the legal properties of the diagram.
Stipulations mostly ALSO require achieving a certain goal/aim by the forward play (“ALSO” highlights that retro-analysis is always required).

Since Codex allows “expressly stipulated” DP-rule for any chess composition, there’s no need for imposing some artificial subjective “rules” which remain well hidden from a solver.
Or, the author should “expressly stipulate” RP, so the solver might try to apply the retro rules to h=3.

32. Andrew Buchanan пишет:

Hi Nikola, let’s keep this thread for new chess content. Theological debate is of narrower interest: please email me and I am happy to respond.