Julia's Fairies

No.848 (PH)

Peter Harris (South Africa)


Original Problems, Julia’s Fairies – 2015 (II): July – December

   →Previous ; →Next→List 2015(II)

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

No.848 by Peter Harris – Complicated solutions, six fairy conditions, and a question about change of color with combination of Anti-Circe and Andernach. (JV)


Sentinels Pion Advers: When a piece (not a Pawn) moves, a Pawn of the colour of the opposite side appears on the vacated square if it is not on the first or the last rank, and if there are less than 8 Pawns of that colour on the board.

Anti-Circe: After a capture the capturing piece (Ks included) must immediately be removed to its game array square (necessarily vacant, else the capture is illegal). Captures on the rebirth square are allowed. Game array squares are determined as in Circe.

Super-Circe: When captured, a piece is reborn on any free field on the chess board without causing self-check or selfmate. Possible is also removal of captured piece from the board. The Pawns (white, black, neutrals, half- neutrals) can be reborn on the first or eight row also. When reborn on the first row (for Black) or on the eight row (for White) the promotion is obligatory. When reborn on the first row (for White) or on the eight row (for Black) the Pawns are immovable.

Circe: Captured units (not Ks) reappear on their game-array squares, of the same colour in the case of pieces, on the file of capture in the case of pawns, and on the promotion square of the file of capture in the case of fairy pieces. If the rebirth square is occupied the capture is normal.

Andernach Chess: A piece (excluding King) changes its color after any capturing move. Rooks on a1, h1, a8 and h8 can be used for castling, provided the usual other rules for that move are satisfied.

Anti-Andernach Chess: A piece (excluding King) changes its color after any non-capturing move. After capture, the piece retains its color. Rooks on a1, h1, a8 and h8 can be used for castling, provided the usual other rules for that move are satisfied. After castling, Rooks do not change color, If White makes a non-capturing move with neutral or halfneutral piece, that piece becomes black and vice versa.

Isardam: Any move, including capture of the King, is Isardam illegal if a Madrasi-type paralysis would result from it.

Madrasi: Units, other than Kings, are paralysed when they attack each other. Paralysed units cannot move, capture or give check, their only power being that of causing paralysis.

No.848 Peter Harris
South Africa

original – 03.07.2015
Dedicated to Hans Gruber

Solutions: (click to show/hide)

white ka1 qb1 ra2 bb2 black kh8 qg8 rg7 bh7 sh1

h#1.5          2 solutions           (4+5)
Sentinels Pion Advers
Andernach Chess

NOTE: with Anticirce and Andernach the colour changes after rebirth so for example when wQ captures it is reborn bQd1. (PH)

Notify of

Newest Most Voted
Inline Feedbacks
View all comments
Luce Sebastien
Luce Sebastien
July 3, 2015 23:23

Winchloe gives all these solutions:
1…Q×h1(Qd1N;Sf7) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8)‡
1…Q×h1(Qd1N;Sd8) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8)‡
1…B×g7(Bç1N;Rd2;+b2) 2.Q×a2(Qd8B;Rg8)+ K×b2(Ké1;é3)‡
1…B×g7(Bç1N;Rd2;+b2) 2.Q×a2(Qd8B;Rg8)+ K×b2(Ké1;ç2)‡
1…B×g7(Bç1N;Rd2;+b2) 2.Q×a2(Qd8B;Rg8)+ K×b2(Ké1;a8)‡
1…Qg6(N) 2.Qh6(B;+g6) g×h7(h2N;Bd8)‡
1…Qg6(N) 2.Qh5(B;+g6) g×h7(h2N;Bd8)‡


Luce Sebastien
Luce Sebastien
July 3, 2015 23:29

Sorry, with Anticirce Calvet, it is only three solutions:
1…Q×h1(Qd1N;Sf7) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8)‡
1…Q×h1(Qd1N;Sd8) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8)‡
1…B×g7(Bç1N;Rd2;+b2) 2.Q×a2(Qd8B;Rg8)+ K×b2(Ké1;é3)‡


Eric Huber
Eric Huber
July 4, 2015 00:43

After 1…Q×h1(Qd1N;Sd8) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8) Popeye considers it is not mate because of 3.Qg6=wQ ! using condition Isardam.

In the second solution 1…Bb2*g7[wBg7->c1=b][+bRd2][+bPb2] 2.Qg8*a2[bQa2->d8=w][+wRg8] + Ka1*b2[wKb2->e1][+bPe3] # it is also very interesting to discover why simply 2…Ka2 is not mate, and then why the bPb2 captured by wKa1 must be reborn on e3 and not on c3.

This problem is quite incredible.

Nikola Predrag
Nikola Predrag
July 5, 2015 01:24

5 diverse fairy threads are beautifully interwoven into one magical fabric.
(Andernach+AntiAndernach is a single thread).

A hypothetical “capture of a King” seems dubiously treated by Popeye, but at least consistently.
Winchloe’s treatment looks inconsistent:
if 3.Qg6=wQ doesn’t parry the check after 1…Q×h1(Qd1N;Sd8) 2.Rg1(B;+g7) R×d1(Rh1N;Qé8), then why 1…B×g7(Bç1N;Rd2;+b2) is not a selfcheck?

Georgy Evseev
Georgy Evseev
July 7, 2015 15:43
Reply to  Nikola Predrag

I am not sure, but it looks like after capture on a1 and any promotion promoted piece is rebirthed and square a1 is vacated. And then Rd2 is under attack from Ra2 (rebirth square is free) – so all this is Isardam-illegal and that is why the king is not under the check.

Nikola Predrag
Nikola Predrag
July 7, 2015 18:37
Reply to  Georgy Evseev

Yes, but in principle, SuperCirce rebirth might occur on b2/c2, closing the Isardam-line.
So, square a1 is attacked in principle due to SuperCirce.
wSa1 could be captured by bPb2.
A King can not be captured but could be attacked. So, a precise definition of “check” would determine what’s happening.

Georgy Evseev
Georgy Evseev
July 8, 2015 21:12
Reply to  Nikola Predrag

This is a thin ice, as there is no strict definition of “capture of king”. But in my opinion “check” means that a “virtual” capture of king is possible. And no rebirth rules are applied to the king (unless the opposite is stated directly).

Nikola Predrag
Nikola Predrag
July 9, 2015 05:21
Reply to  Georgy Evseev

Some orthodox concepts should be altered or at least reformulated if we want the precise and unambiguous fairy rules.
King is in check if it is attacked by some opponent’s piece(s), even if such piece(s) can’t move to that King’s square.
The attack is in general defined through attacking a certain square.
If a piece X could make a capture on some square, then an opponent’s piece Y on that square is attacked.
It is NOT said “if X could capture Y”.
Such definition is necessary because it’s explicitly said that a King can not be captured. King is in check when its square is attacked, although the King himself can’t be captured.

There’s no “hypothetical/virtual” capture of a King in chess rules.
The concept of attack seems to require a “hypothetical/virtual” capture of some “hypothetical/virtual” opponent’s piece, other than King.

I don’t know the full original definition of Isardam, but I remember the point that the “spike” is “stronger” than the orthodox pin.
A pinned piece can’t vacate the pinning line but nevertheless, it does attack if it could make a capture.
A spiked piece does not attack if it can’t vacate the spiking line without leaving it open after the move is completed.

SuperCice allows that a hypothetical capture of some hypothetical piece could end with a hypothetical rebirth. If such hypothetical rebirth could close the spiking line, the spiked piece could make a capture on a certain square and so, it attacks that square.

A “hypothetical/virtual capture of a King” might be applied in fairy chess if a clear definition is provided.
The essential distinction is the “royalness”.
“IF” a Royal piece was NOT royal, it could be captured.
That appears to me as a logical ground for definitions of “hypothetical/virtual”.

Georgy Evseev
Georgy Evseev
July 9, 2015 13:13
Reply to  Nikola Predrag

>It’s explicitly said that a King can not be captured.
This is a very unfortunate rule, because in orthodox chess the move, capturing the king, is never possible anyway.

From my point of view this rule is intended to avoid the situations when in game the king is unintentionally left under attack.

Generally, the definition of “check” may be made very simple, if it based on “capture” concept and not on “attack” or “observe” concepts.

But there are some difficulties, concerning rebirth effects.

Nikola Predrag
Nikola Predrag
July 9, 2015 20:11
Reply to  Georgy Evseev

The game of chess has come out of, and was founded on the most essential idea which requires the uncapturable King.
That’s the reason for such a successful development and survival of chess through so many centuries.
HUMAN understanding intuitively recognizes the uncapturable King as an axiomatic premiss, even if that’s not even mentioned.
MACHINE reasoning (“understanding”) can’t recognize that, so it must be mentioned in the rules. That doesn’t affect the machine’s “understanding” of chess (since there’s no such thing), but gives what a machine can “understand” in practice – an explicit rule!

Now, orthodox chess can be played by mere “machine reasoning” but the point is that a machine is expected to SIMPLY APPLY the rules, without questioning their purpose.

Human understanding will ask for a purpose but it’s absurd to search for it through a machine-reasoning.
We should either THINK as the human beings or, simply apply the rules without questioning them if we so much adore the COMPUTING (machine-reasoning).

Thus, the fairy chess may develop in two major directions, based on:
1. Human understanding of chess (the concept of uncapturable Royalty can’t even be questioned)
2. Machine “understanding” of chess (any dehumanized systems)

peter harris
peter harris
July 6, 2015 09:11

Thank you for your comment Nikola.

I have been waiting for someone else to write about the Winchloe solution with the bS on d8 – and the matter of inconsistency.

I do not think WinChloe is being inconsistent. It is to do with their interpretation of conditions.

I think the explanation is as follows:

With the scheme:

W: Ka5 Rc3
B: Kf3 Rc5 Sh8
Conditions: madrasi anticirce

Winchloe would say that the bK is in check because the wR is not paralysed because the bR could not capture it. Only the bR is paralysed – so wKa5 is not in check.

[Popeye does not work this way – saying that the wR is paralysed].

And this WinChloe interpretation carries over to Isardam.

In my problem: WinChloe says because d8 the rebirth square of a bQ is occupied, the wPg7 can capture the bKh8 without violating Isardam even though the bQg8 and a wQg6 would then be looking at each other.

Nikola Predrag
Nikola Predrag
July 6, 2015 21:09
Reply to  peter harris

Thanks for the explanation, Peter. I didn’t know the Winchloe’s interpretation, since I don’t have it.
So, there’s a crucial difference between the two programs about the mutual “observing”/attacking.

I wonder what is a definition of the “check” (attack on a King) which makes the 2nd solution legal.
And with no check defined, there can be no mate.

Actually, I enjoy the incredible beauty of that fantasy, pretending as though there was a satisfying definition.

July 9, 2015 02:00

It seems the difference comes from how Madrasi is handled when combined with another condition X such as Anticirce.

Popeye: a piece is paralyzed when threatened, in the orthodox sense, by a piece of same nature.

WinChloe: a piece is paralyzed when threatened, in the Anticirce sense, by a piece of same nature.

I have not checked whether or not Popeye is always considering paralysis in the orthodox sense – whatever is the additional condition X.

On the other way WinChloe is (unfortunately to my mind) fluctuant… For example when X = T&M, then the threat is considered in the orthodox sense.

As an example consider the scheme wPb6 bPa7 bSb7. Then the wP is paralyzed although it can’t be T&M-captured by the bB.

Anyway, providing definitions for condition X and condition Y is not enough for a full definition of the mixed condition X&Y – the way X and Y are interfering should be specified too – as the problem under consideration clearly demonstrates.

peter harris
peter harris
July 9, 2015 11:40

About Isardam:

POSITION I: W: Ba1 B: Kd4 Bh8

Conditions: Could the K be captured?

(i) None Yes
(ii) Isardam No
(iii) Isardam+Anticirce Yes
(iv) Isardam+Andernach Yes
(v) Isardam+Chameleonch Yes [ the bB cannot
(vi) Isardam+AntiAndernach No [the bB cannot move]

POSITION II: Position I + bSc1
As per POSITION I except (iii) No

POSITION III: Position I + bSc1 + wQd3
As per Position I except (ii)(vi) Yes
(iii) No

POSITION IV: W: Ba1 Qd4 B: Kc5 Bh8
All No except (i) in Position I

It would make no difference to the answers if EITHER Supercirce was added to any of the above conditions OR a bS replaced the bK. There would a difference if BOTH happened. In POSITIONS 1 and IV all answers would then be Yes because the bS could always be reborn on the a1h8 diagonal.

[Definitions of conditions should provide for the different concepts of attacking/observing and could capture].

peter harris
peter harris
July 9, 2015 12:35


You are not right in your analysis.

Never mind about condition X.

The matter is to do with the definition of Madrasi – alone.

The Definition of Madrasi used by Popeye [and most magazines] state that paralysis occurs with OBSERVATION.

A wRa4 and a bRc4 OBSERVE and paralyse each other.

WinChloe on the other hand says paralysis occurs only when a piece can capture.

This means that AntiCirce [and no doubt other conditions] can enter the picture – by preventing a capture. This serves to confuse.

July 9, 2015 15:11


Consider the scheme wPb5 bPa6b6b7, white on move.
The wP cannot be captured, nor in T&M neither in Anticirce, but according to WinChloe:

– wP is paralyzed under Madrasi + T&M
– wP is not paralyzed under Madrasi + Anticirce

So the additional condition really matter to determine if a piece is Madrasi-paralyzed or not, at least for WinChloe.

Now what about Popeye? You claim that a piece is Madrasi-paralyzed when observed by a piece of same type. The point is that “observation” also depends on the additional condition!

As an example wRa1 and bRb1 don’t observe each other in Monochrom – and indeed Popeye considers them as not Madrasi-paralyzed…

peter harris
peter harris
July 9, 2015 17:13


I am not writing anything more on this subject.

NB: I say this in a friendly and genial way!

adrian storisteanu
adrian storisteanu
July 9, 2015 18:13
Reply to  peter harris

Not sure how related — but in a domain of art as well, and just as genial: 🙂

I must tell you with regard to this matter that the many studies I made having given only negative results, and dreading the critics who are only too justified, I had resolved to work in silence until the day when I should feel myself able to defend in theory the results of my attempts.
— Paul Cézanne,
letter to Octave Maus, Paris, 27/xi/1889

July 9, 2015 17:22

In fact I’m in trouble with the “observation” concept, and its link with Madrasi-paralysis is unclear to my eyes. For example wPa2 is certainly observing square a3 while a bPa3 is clearly not Madrasi-paralyzed by the wPa2…

May someone remember me what is the precise definition of “observation” ?

Btw it is worth to mention that the Madrasi definition provided at the beginning of this thread doesn’t fit Popeye’s approach: in the already given Anticirce scheme wRf7 bRf8 bSh8, the wR is Madrasi-paralyzed although not attacked by the bR…

Would love your thoughts, please comment.x