No.1278 (JR)

Jacques Rotenberg


Original Fairy problems
October’2017 – March’2018

Definitions: (click to show/hide)

No.1278 Jacques Rotenberg

original – 25.02.2018
Dedicated to Peter Harris

Solutions: (click to show/hide)

Black Re8 Pd5 Neutral Kh7

h#2                                         (0+2+1)
b) +bBe8 ; c) +bSe8 ; d) Pd5→h6
Sentinelles PionAdvers
Transmuted Kings

29 Responses to No.1278 (JR)

  1. peter harris says:

    The plain fact is that with transmutedkings wPe7 bKd8 is mate and with wPe6 wRa6 bKd7 is not mate. In the first case the bK has to move as though it was a P and a bP on its first rank cannot move. In the second case, being on its second rank the K could move two squares to d5.

    Programmers after making a programming error sometimes do not admit it but instead try to justify a “solution” by some specious argument.

    Such would be the case here with WinChloe’s “ïnterpretation”.

  2. Jacques Rotenberg says:

    Peter, you seem to know the truth, it is wonderful!
    The fact is that I agree 100% here with winchloe : a pawn on its 1st rank moves like a pawn. And I don’t think it is an error.
    At the opposite, to say that pawn on its 1st rank cannot move seems strange : why to introduce new rules, when the existing are enough ?

    At least let’s say that there are two ways of understanding this, the one selected by Christian Poisson being by far the most natural :
    a pawn moves always the same way. On its second rank it has the double step move also. On its 8th rank it cannot move. Unless other rules are given explicitly.

  3. peter harris says:

    It is GENERALLY ACCEPTED AND ASSUMED that a pawn on its first rank is dead and cannot move. This is the basis used by composers, solvers and by magazines including ie Schwalbe, feenschach and Probleemblad and their editors. This dead pawn convention is used by Popeye.

    [As a matter of interest: the foremost fairy composer says pawns on their first rank should not appear on a board – unless the problem has a condition whereby they could have been reborn there – for example by supercirce].

    The foregoing means that when a problem is published with a pawn on its first rank there is no need to advise everyone that it cannot move.

    If however a composer composes on the basis that a first-rank pawn CAN move he should advise the publisher who in turn should advise readers when publishing. It is totally insufficient only to say “Solved by using WinChloe” This does not help a solver at all. It does not inform him that first-rank pawns can move. He more likely than not knows nothing about WinChloe.

    Writing a program in which pawns on their first rank can move is an example of SOWING CONFUSION in the Fairy problem world. Whatever argument can be made to justify an interpretation contrary to that generally accepted, the contrary interpretation should be foregone and forgotten unless there is some extraordinary reason not to do so.

    [A pawn on its first rank is unnatural and so the concept of naturalness should not enter into deliberations regarding its movement].

    Thank you for dedicating your problem to me!

  4. Andrew Buchanan says:

    Peter: where is this apparently key “generally stated and assumed” fact written down and (forums don’t count)? Without this, how are newcomers magically expected to come to the same conclusion as you? You can’t really accuse someone of “sowing conclusion” for simply not being a telepath. In this vacuum, my own approach is to take the existing Laws as literally as possible, with all the amusing paradoxes that result. Thanks Jacques, Christian.

  5. Luce Sebastien says:

    Peter : you will probably like a lot the following problem 🙂

    White : Pd1f1
    Black : Kh1 Pe1g1
    Sebastien LUCE
    Problemskak 2016
    sh‡24 (2+3) C+ Winchloe
    Circe symetrique horizontal

    4.Kxf1(Sf8) 8.Kxd1(Bd8) 15.Kxf8(Sf1) 17.Kxd8(Bd1) 24.Kh1 Bf3‡

    Horizontal symetrical Circe :
    When a piece is captured (King excepted, unless contrary indication), it must be placed on the horizontal symetrical square of the square of capture if it is empty : if not the captured piece vanishes.

  6. Jacques Rotenberg says:

    Dear Peter,
    Sorry, I don’t like this kind of discussion.
    There have been already many discussions on “what fairy rules should be”
    Block letters are not arguments
    Adjectives are not arguments
    “General acceptance” could be one, but here it is not. What you call “general acceptance” is the fact that some people use Popeye.
    In other words you say “what is the option in popeye must be the rule”
    Don’t you feel that it is somewhat excessive ?

  7. peter harris says:

    When Die Schwalbe, feenschach, Probleemblad and I would guess The Problemist, publish a problem with supercirce and pawns, they do so on the basis that if a pawn is reborn on its first rank it is dead, cannot move. This is assumed by solvers, composers, and all readers of their magazines This Jacques and Andrew is what I mean by GENERALLY ACCEPTED AND ASSUMED

    That Popeye uses this dead pawn rule is quite beside the point I am making.

    Christian Poisson is free to progam WinChloe on the basis that pawns on the first rank can move. This will confound and confuse everyone except the chosen few. It necessitates that such an interpretation be made known when such problems are published.

    This will be my last post on this problem. I have said all I wish to say.

  8. Jacques Rotenberg says:

    Due to the fact that there are different ways to understand things, it seems a good idea to add to the stipulation :
    “A pawn on its 1st rank can move as a pawn”
    Also perhaps :
    “When a neutral piece leaves a pawn on its vacated square, the pawn is white or black, not neutral”

  9. Luce Sebastien says:

    Peter : the “rules” in chess composition must allow the greatest creative possibilities, NOT TO LIMIT IT ! It is nonsense to deprive the Pawn of its mobility on its first rank. Better to use a “dummy Pawn” but without excess.

    You could be of the same opinion as you precisely like to explore the limits of chess composition !

  10. Nicolas Dupontdupont says:

    “When Die Schwalbe, feenschach, Probleemblad and I would guess The Problemist, publish a problem with supercirce and pawns, they do so on the basis that if a pawn is reborn on its first rank it is dead, cannot move.”

    And so is working WinChloe:

    Super-Circé :
    Lorsqu’une pièce est capturée (Roi excepté, sauf indication contraire), elle peut être replacée sur n’importe quelle case vide.
    Exception aux règles par défaut : Un Pion est immobile sur sa 1° rangée.

    The fact is that there are no unified laws about moving possibilities of a pawn standing on its first rank – it depends on the fairy condition you are working with. For example in Einstein, such a white pawn may play a triple step move from a1 to a4…

  11. Jacques Rotenberg says:

    I translate :

    Super-Circé :
    Lorsqu’une pièce est capturée (Roi excepté, sauf indication contraire), elle peut être replacée sur n’importe quelle case vide.
    Exception aux règles par défaut : Un Pion est immobile sur sa 1° rangée. :
    when a piece is captured (not the king – if not mentioned) it may be replaced on any empty square.
    As exception on the default rule, a pawn replaced on its 1st rank cannot move.

    Well…it appears that associations of fairy rules provoke sometimes unexpected complications, and the problem, although having only three pieces seems suddenly much heavier

  12. Andrew Buchanan says:

    Jacques: our understandable zest for lightness is at odds with the actual complexity of certain fairy rules, and the ways they interact. PDB rule definitions predictably miss important corner cases. As Nicolas points out, there is no “one-size-fits-all” across different rules, although some general orienting principles can be helpful.

    The newcomer may not revisit if all we do is buried in opaque terminology, and unwritten assumptions. This is the golden age of the puzzle, but many more puzzlers follow sudoko and its many kin than chess.

    Encouraging newcomers is much more important than promoting the elite, because if the hobby is bigger, a more skilled elite will emerge anyway. Explaining fairy rules properly is a central task here, and I think Julia through this site does a wonderful job in growing the hobby.

  13. peter harris says:

    Nicholas: I was interested to learn that with WinChloe, when there is the Supercirce condition, a pawn on its first rank is immovable. I would like to know what the position is when there is no Supercirce condition (or any other).

  14. Nicolas Dupontdupont says:

    With position wPa7 bTKb8, the black transmuted king is checkmated for Popeye but not for WinChloe – it allows TKxa7 and TKb7.

    The reason seems that the TK definition doesn’t cover such a situation, hence Popeye and WinChloe are using their own default rules for a black pawn standing on the eight row, which are not the same!

    Indeed it looks like this default rule is immobility for Popeye, but a move to the seven row for WinChloe (diagonally in case of a capture).

  15. peter harris says:

    Nicolas,you are not answering the question I asked. The question could not be simpler.

    With WinChloe, when a problem has NO CONDITIONS what is the position regarding pawns on the first ranks, are they immovable or not?

    [I would mention that it is generally accepted that such pawns are immovable. As I mentioned in an earlier post by generally accepted I mean that it is the basis used by Die Schwalbe, feenschach and Probleemblad when they publish problems and is assumed to be so by solvers, composers and others].

  16. andrew buchanan says:

    FIDE Law 3.7 is quite clear about a pawn on the first rank. If it’s not capturing, it can move one square forwards, or two if it’s the pawn’s first move. Clearly such a position is illegal, but it’s *only* illegal, and the rules of chess can adapt perfectly well to illegal positions. If we are looking for a fun, consistent foundation for default fairy conditions, then this is the right place to begin, rather than some possibly “generally accepted” notions from the remote past that no one bothered to write down at the time. Specific fairy conditions may of course have terms which override this default. Also, I like the idea of a brave little pawn starting on the first rank and heroically managing a super-excelsior run to the eighth. That has much more creative potential than just sad deadwood stuck on the first.

  17. Nicolas Dupontdupont says:

    Peter, with position wPa1 bPb2 and without any condition, WinChloe considers there are 2 legal moves: a1-a2 and a1xb2. Moreover, after such a move, the resulting pawn is considered as a “normal” one.

    With Popeye 4.79:

    pieces white Pa1 black Pb2
    stip h~0.5

    provides no legal move.

  18. Nicolas Dupontdupont says:

    To my mind neither program is faulty, they just made a different choice regarding the above particular situation. This is the transmuted king definition which is incomplete, as it doesn’t tell which choice has been made when such a white king on the first row is checked by a pawn.

  19. François Labelle says:

    Popeye has the condition “NormalPawn” which seems to exist to obtain WinChloe’s pawn behavior. With it the problem is C+ by Popeye, except for part (b) where Popeye disagrees with the move 2…nKf7-g8 (+Pf7)#. This seems to be a disagreement about TransmutedKings, though. A white king in check by a black pawn will move like a white pawn, but can it move diagonally without capturing? Popeye says no and WinChloe says yes.

  20. peter harris says:

    Thank you for this information Nicolas.

    Two points arise:

    First, there is an inconsistency with regard to how WinChloe treats pawns on their first ranks. Given a problem with pawns on first ranks: if the diagram of a problem has Supercirce under it, the pawns are immovable; remove Supercirce and the pawns become mobile. This not only inconsistent, it is ridiculous.

    Second, more serious: WinChloe is going its own way against (widespread) accepted understanding. Why does WinChloe not follow the accepted practice as it does with Supercirce – by providing that first rank pawns are always immovable? It is an example of sowing confusion quite gratuitously. Solvers and everyone else become confused. It is serving no good purpose to be contrary.

    NOTE: It also causes difficulty for editors when a problem is presented (to solvers) when published. I do not think WinChloe always realizes the full consequences of its decisions and in this regard can become irresponsible.

    As I wrote previously: however much making pawns mobile can be justified WinChloe should at times forgo the alternative interpretation for the sake of avoiding confusion – unless there is some compelling reason not to.

  21. Nicolas Dupontdupont says:

    For part b), my own WinChloe doesn’t give the expected result…

    • François Labelle says:

      Strange. An easy fix is to change part (b) to +bBe8. Maybe that was the position checked by Julia using WinChloe?

      Earlier Jacques wrote that it seems a good idea to add to the stipulation “A pawn on its 1st rank can move as a pawn”, so maybe one option is to add “Normal Pawn” as a condition, with Jacques’s sentence under the show/hide “Definitions”? With all 3 conditions and the change to part (b) the problem is C+ Popeye.

      • Nicolas Dupontdupont says:

        “An easy fix is to change part (b) to +bBe8”. Yes indeed – there is probably a misprint in the b) stipulation.

        “maybe one option is to add “Normal Pawn” as a condition”. Yes, even if providing a full definition of transmuted kings sounds another valid option! Two different versions might also be quite coherent – transmuted kings with normal pawns and transmuted kings with immovable pawns.

  22. Nicolas Dupontdupont says:

    Peter, it seems you are thinking that Popeye always considers first rank white pawns as immovable. This is not true:

    pieces white Pa1
    stip h~0.5
    cond einstein


    It appears that both Popeye and WinChloe correctly follow the way first rank white pawns should move according to a given condition, providing the definition of this condition includes this rule (supercirce or einstein). But when the definition tells nothing about this rule (transmuted kings), then programs follow their own default rules, which are different.

  23. peter harris says:


    Of course with Einstein pawns can move from their first ranks.

    This is totally irrelevant to what I wrote in my previous post.

    In particular it is totally irrelevant to the crucial issue at hand.

    The crucial issue is how a pawn can move when on its first rank when there no Conditions.

    The fairy world as a whole believes / assumes that the pawn is immobile.

    You have not addressed the points I raised in my last post.

    It’s no use Nicolas.

    I am out of here now.

  24. Nicolas Dupontdupont says:

    Concerning WinChloe, you wrote: if the diagram of a problem has Supercirce under it, the pawns are immovable; remove Supercirce and the pawns become mobile. This not only inconsistent, it is ridiculous.

    Why don’t you add, concerning Popeye: if the diagram of a problem has Einstein under it, the pawns are mobile; remove Einstein and the pawns become immovable. This not only inconsistent, it is ridiculous.

  25. Jacques Rotenberg says:

    We arrive now to 26 comments for a poor problem with only 2 “like”
    Even so, I’ll add a 27th one.

    First of all there is a mistake in the written twins : b) is with a bishop on e8! (not a pawn!)

    now, about the discussion :
    Our situation is a bit special: we speak of pawns on their 1st rank…when in fact there is none.
    The difficult thing comes from the condition “Transmuted kings”
    The definition given here (Julia’s Fairies – Fairy terms) is :
    “When they are threatened, the Kings move only like the threatening unit(s).”
    In winchloe in French, it is the same.

    That definition needs some explanations, and specially with pawns attacks
    1. Everybody understands that in case of a pawn the attacked king moves as a pawn of its one colour – and not as a pawn of the attacking pawn colour.
    2. Everybody understands that if the attacking piece has some technical limitation (for example being pinned or blocked) the king will not have these limitations

    these two points, not because of the given definition, but because of the name: “transmuted” meaning more “transformed in”. A checked king is transformed in a piece of the same nature as the checking one but does not change its colour. when checked by more than one piece, it cumulates the powers of the checking pieces.

    About checking pawns it works like this in winchloe – and I feel it is quite “natural”:
    a. a king checked by a pawn moves as a pawn of its own colour: one step ahead without capturing and diagonal-ahead with capturing.
    b. a king on its 2nd rank, checked by a pawn, may play one or two steps ahead without capturing it may be blocked or intercepted
    c. a king on its 1st rank, checked by a pawn moves like in a.
    this is the controversial point.

  26. Jacques Rotenberg says:

    Also to answer the question of Peter:
    In winchloe under orthodox rules a pawn placed on its 1st rank play as a “normal” pawn (1 step ahead without capturing, 1 step diagonal-ahead with capture)

Leave a Reply

Your email address will not be published. Required fields are marked *

You can add images to your comment by clicking here.