Julia's Fairies

No.680 (AA)

Alberto Armeni 

Warm welcome to Alberto in Original Problems section of JF!


Original Problems, Julia’s Fairies – 2015 (I): January – June

   →Previous ; →Next ; →List 2015(I)

Please send your original fairy problems to: julia@juliasfairies.com

No.680 by Alberto Armeni – Interpretation of Valladao theme with Imitator! (JV)


Imitator(I): Every time a piece moves an Imitator (or a set of Imitators) moves simultaneously in an identical manner. An Imitator cannot move of itself. If an Imitator cannot imitate the move of a piece, the move is illegal. An Imitator may only pass through or enter an unoccupied square and cannot move off the board. Castling is imitated by decomposing into a King move followed by a Rook move. 

No.680 Alberto Armeni

original – 04.01.2015

Solution: (click to show/hide)

White Kh1 Pb4 Black Ke8 Ra8 Sd5 Pa7 Pc7 Pb6 Neutral Ib1

h#4                                         (2+6)
Imitator b1

Notify of

Newest Most Voted
Inline Feedbacks
View all comments
January 5, 2015 01:17

cooked Winchloe:
1.0-0-0{Iç1} b5{Iç2} 2.a6{Iç1} b×a6{Ib2} 3.Cé7{Iç4} a7{Iç5} 4.Td7{Iç4} a8=D{Iç5}‡
1.0-0-0{Iç1} b5{Iç2} 2.a6{Iç1} b×a6{Ib2} 3.Td7{Ib1} a7{Ib2} 4.Cé7{Iç4} …
1.Cf6{Id2} b5{Id3} 2.a5{Id1} b×a6 e.p.{Iç2} 3.0-0-0{Id2} a7{Id3} 4.Cd7{Ib4} a8=D{Ib5}‡
1.Cf6{Id2} b5{Id3} 2.a5{Id1} b×a6 e.p.{Iç2} 3.Cd7{Ia3} a7{Ia4} 4.0-0-0{Ib4} …
1.Cé7{Iç3} b5{Iç4} 2.a5{Iç2} b×a6 e.p.{Ib3} 3.0-0-0{Iç3} a7{Iç4} 4.Td7{Iç3} a8=D{Iç4}‡
1.Cé7{Iç3} b5{Iç4} 2.a6{Iç3} b×a6{Ib4} 3.0-0-0{Iç4} …

Diyan Kostadinov
Diyan Kostadinov
January 5, 2015 02:39
Reply to  luce

Actually the problem is correct. In the Award of Christmas Blitz tourney 2014 the judge Petko Petkov explain that in case of castle the Imitator should follow first the King’s move and after that the Rook’s move.
So for example in the first solution of your comment the castle 1.0-0-0 is not possible because the Ib1 can not follow the King playing from e8 to c8. In other solutions is the same.

January 5, 2015 18:35

In the state of things, the problem is only C+ Popeye.
But I think the programmers should determine the most logical common rule for many fairy conditions because it is a little desorder actually !…

January 5, 2015 18:54

Yes, it seems dubious to claim that a problem is “correct” when it depends on some unclear rule interpretation. In particular in this setting, where Popeye’s interpretation of the 0-0 with Imitator is breaking some old problems, for example:

Gerhard W. JENSCH
Schachmatt 1947

White : Kb1 Qf5
Black : Ke8 Rh8 Pa6g5
Imitator d4


1.Dç2{Ia1}! 0-0{Ia1} 2.Dh7{If6}‡

But no solution according to Popeye.

peter harris
peter harris
January 5, 2015 19:31

I suggest the Definition appearing above the problem be read before saying that the problem is cooked. [It is not cooked]. Popeye follows this definition – which is endorsed by Petko Petkov as stated in his preamble to the A Section awards.

WinChloe uses a different definition. It would be interesting to read the wording of the definition.

petko petkov
petko petkov
January 5, 2015 20:14

Luce says:
January 5, 2015 at 18:35

In the state of things, the problem is only C+ Popeye.
But I think the programmers should determine the most logical common rule for many fairy conditions because it is a little desorder actually !…

But….the problem is C+ also according to Alybadix -2010!
On this reason I have some doubts that the “Cooks” of WinChloe are simply bugs!
But in such cases it is useful to learn and the arguments of programmers, although
they can not be “the last instance of truth” in chess composition.
However, I will not change my opinion!

January 5, 2015 21:11

Indeed I already asked Christian Poisson on this stuff – it is not a bug from WinChloe, the rules followed by this program are those which govern the older published problems (as the one I mentioned in my previous post).

peter harris
peter harris
January 5, 2015 21:24

With regard to Castling, Popeye uses the Definition as stated above the diagram.

WinChloe uses some other Definition.

The WinChloe Definition should be made known so that Editors can show the Definition when publishing a problem based on that Definition.

The different Definitions could then be compared

There will be different opinions on the merits of the two Definitions.

[From the “Imitator moving simultaneously with a moving piece” point of view I think any alternative to King move followed by Rook move would not make sense].

Kostas Prentos
Kostas Prentos
January 6, 2015 05:02

WinChloe’s definition of the Imitator seems quite obvious to me. The only difference with the one used by Popeye is regarding castling. Popeye moves the Imitator back and forth during the execution of the King and Rook moves, while WinChloe takes the effect of the castling move after the two-step move is completed and leaves the Imitator on the same square for 0-0, or moves it one square to the right for 0-0-0.

I will not attempt to give preference to one version, or the other. Like many other examples of fairy conditions, with discrepancy between the solving programs, I will accept that this problem is sound according to Popeye (and Alybadix). I will also accept older examples like the one mentioned by Nicolas as sound per WinChloe. Until such matters are resolved by a definitive ruling from the WFCC, one way or another, anyone is entitled to an arbitrary opinion on how to interpret the rules and we will keep discussing the same things again and again, without agreeing. It is inevitable!

peter harris
peter harris
January 6, 2015 08:07


It is not as simple as that.

AT THE VERY HEART of the Imitator concept is MOVEMENT by the imitator in imitation of pieces – as per the Definition above the diagram.

With WinChloe there is no movement with 0-0. If an Imitator was on f1 or g1 those squares will be occupied by a piece and the Imitator together. This would not be in contravention the basic Imitator definition – because there is no movement by the Imitator. If it was to be said that there was a contravention then the basic Imitator definition would have to be altered.

Bearing the above in mind: I would be interested to see [your idea of] the WORDING of the basic definition of the Imitator and as well as the WORDING for castling.

[In my opinion, it is just commonsense to have two separate Imitator MOVEMENTS when castling].

N.B. My concern however is not on the merits of the different Definitions.

My MAIN POINT is that a [full] Definition of the Imitator HAS to be given [by Editors] when publishing an Imitator problem.

So: you or anybody else would be doing a service if you gave the WinChloe definition to Julia – and to all other Editors – so that they could use the Definition when an author composed a problem using WinChloe.

Finally: it is not the end of the world if two Definitions coexist PROVIDED THAT the Definitions are made known.

peter harris
peter harris
January 6, 2015 10:59

Yes Julia it is about Definitions.

Quite simply: when an Editor publishes a problem a Fairy element has to be defined. It is impossible for things to be otherwise.

Taking the rules for Castling in an Imitator problem as an example, an Editor must state the Definition.

If an author submits a problem based on WinChloe you could not use the Definition you are currently using. What definition WOULD you use?

It is unacceptable nonsense that any program is not “keen” to publish its Definitions. How, I ask, can a problem be published without a Definition?

Please reread my last two comments.

Otherwise Julia, remember that I am on your side in this matter!

January 6, 2015 11:01

For sure Kostas, as far as we are discussing with interpretation of some rule, it is difficult, or even impossible, to reach a general agreement. Nevertheless it is not forbidden to exchange some ideas!

The beginning of the Imitator definition “every time a piece moves […]” clearly indicates that an Imitator imitates moves.
So Popeye’s approach would be fully convincing if castling were a combination of 2 moves – one by the King followed by one by the Rook.

The problem is that, according to Fide Laws of Chess 3.8, this is not true. Indeed during the castling process King and Rook are not moving, they are just transferred to their destination squares… Saying otherwise castling is an “indivisible” move, and not the conjunction of 2 “semi-moves”!

January 6, 2015 11:11
Reply to  dupont

In NO. 680, Imitator appears to be used solely to force the en-passant with a7-a5 move in a H#5 format.
I find, using Madrasi instead, the Valladao theme can be achieved in H#3 moves s follows:-
W Kh1 Pb5 B Ke8 Ra8 Pa7c7 (2+4)
Even the prob. can be NoWk!

Geoff Foster
Geoff Foster
January 6, 2015 23:07

Perhaps castling should require the presence of two Imitators on the same row (3 squares apart for 0-0 and 4 squares apart for 0-0-0), with one Imitator doing the King move and the other Imitator doing the Rook move?

Kostas Prentos
Kostas Prentos
January 7, 2015 00:45

Peter, as I said, I am not taking sides or supporting one version or the other. I believe I already gave a brief wording of how WinChloe defines the Imitator and castling, the way I understand it from the given examples. I can refine it a bit, in the following way: During castling, the King and Rook move SIMULTANEOUSLY towards each other, and not in two separate steps. This results in a non-moving Imitator in one case (for 0-0: +2-2=0) and in one square to the right movement of the Imitator in the other (for 0-0-0: -2+3=1).

I am not saying that this is the best definition, or that WinChloe’s version of Imitator rules is correct, or better than Popeye’s. All I am saying, is that both definitions are reasonable, and unless one of them is decided to be the “correct” one, we can continue to argue (not you and me, but all of us) and support whatever version of rules we want. And since we are at it, if we decide that the castling must break in two steps, why not have the Rook move first and then the King? I can see the obvious argument coming, that castling is a King’s move and in otb games, we have to touch the King first, etc. But then, in otb games, it is allowed to move both pieces simultaneously, using both hands. Nicolas describes it well, in reference to the FIDE laws of chess. The problem is that chess composition rules do not (have to) necessarily follow the rules of tournament chess, especially when it comes to fairy chess.

Finally, Peter’s example of an Imitator standing between the King and Rook (f1 or g1) will most likely lead to an illegal castling with both computer program versions. I agree that this example demonstrates that more detailed definition is necessary if we want to rule out such obvious irregularities and make sure that the Imitator does not end up on an already occupied square. However, it appears that not moving the Imitator in the case of short castling, or moving it one square to the right for long castling, without caring if there was enough room towards either direction, was the way (some, all?) composers used and defined the Imitator in the past.

In any case, I support the view that any doubts on interpretation of the rules should benefit the composer. Therefore, the present problem should be considered sound per Popeye, and the older problems sound per WinChloe. I understand that for many of us, it is obvious that only one version is correct and I respect this view. Yet, I can still have my own different opinion, at least until one of these versions is ruled out by WFCC.

peter harris
peter harris
January 7, 2015 09:06

Good Kostas.

To say in a different way what I said in my previous posts:

Popeye, WinChloe and other programs should be perfectly free to program their programs according to their own Definitions PROVIDED that composers, solvers and editors are told and know what the Definitions are. [There is no need for WFCC involvement].

Composers would be free to compose according to their Definition of choice.

I would ask program makers: of what earthly use is it to make a program and then not inform composers, solvers and editors of the Definition they used, thus making their program unusable?

Without a Definition composers cannot compose, solvers cannot solve and editors cannot publish.

[Regarding the merits of the different rules pertaining to Castling in an Imitator problem:

In the context of the Imitator concept it is my opinion that Castling should be made TWO moves.

In the Imitator Definition an Imitator SIMULTANEOUSLY MOVES WITH A PIECE. This I think precludes the concept of an Imitator simultaneously moving according to TWO pieces at one and the same time].

Finally: perhaps WinChloe or one of its avid apologists could give Julia a Definition to replace the [Popeye] Definition she uses, when she publishes a problem based on WinChloe.

Would love your thoughts, please comment.x