Website founded by
Milan Velimirović
in 2006

17:56 UTC
ISC 2024



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum Fairies Discussion around P.Petkov's fairy problem at
You can only view this page!
(1) Posted by Julia Vysotska [Thursday, May 24, 2012 08:59]

Discussion around P.Petkov's fairy problem at

For those who is interested in discussions around fairy problems - see Original problems section at Famous composers like Petko A.Petkov, Juraj Lörinc and others are participating. The direct link is:
The last discussion is even more, than that. P.Petkov's comments can be considered as a mini-lectures about the technical and aesthetical principles in composition - see the comments below the original problems at
(Read Only)pid=8606
(2) Posted by Kevin Begley [Friday, May 25, 2012 00:37]; edited by Kevin Begley [12-05-25]

This is a very interesting discussion, indeed; but, to profit from it, we first need to dispel some overreaching claims...

Petko goes too far when he claims that Juraj's goal, with his "obviously experimental position" (read: Juraj's scheme), was to "prove" that "in every case a problem without [Nightriders] is better [than] a problem [which does employ] Nightriders..."

Such an absolute straw-man fallacy is so easily toppled, that Petko's further analysis tends towards the preposterous.

The attentive reader will notice that Juraj already states, quite clearly, that his scheme is "mainly trying to show" that "the use of Nightriders is not necessary."

The key word here is: trying (Juraj concedes that his quickly realized scheme is not fully equivalent).
The wrong word here is: is (what Juraj calls a "proof of concept," is really only a confirmation of good intuition, which suggests that Nightriders might not be necessary).

This scheme may not (yet) fully prove that, in this specific case, Nightriders are unnecessary; but, let's not burden Juraj with a need to provide a mathematical proof that "fairy economy" represents, in all cases, a superior form.

I'd also like to hear an elaboration from Petko, regarding his further claim:

"Therefore it is wrong to say abstractly – as general conclusion, that at the identical contents, a problem which has 2 kind of fairy pieces (for example Grasshoppers and Locusts) is better than a problem which has 3 kinds of pieces (for example – Grasshoppers, Locusts and Nightrieders)."

I suspect many would disagree with the sentiment of this claim, as it is (rather loosely) stated.
It seemingly neglects entirely that "economy of fairy elements" is a very important goal.

We should all be aware that the principle of "fairy economy" dictates only a preference to avoid fairy elements, when they are both thematically unnecessary, and not adequately demonstrated.


James S. Hall
South African Chess Magazine, 1938
(= 3+1 )


1.Kg2 Sgh3 2.Kf1 Se3#

If we replace the Knights with Nightriders, we would obtain a completely identical solution (and no other).
I can't fathom that Petko would ever intentionally imply that this would be considered a valid use of Nightriders.
Which is why, I believe, he needs to elaborate on -- more likely: qualify -- his claim.

The burning question: do the Nightriders constitute a fairy excess, in Petko's problem?

Unless (until?) it can be shown that a fully equivalent solution is possible, Petko may have an adequate justification for this additional fairy element.
On the other hand, without a careful investigation, no judge could easily discount Juraj's intuition: the Nightriders might well prove to be an excess.

I suspect Juraj was only suggesting that Petko might want to reconsider whether his expression is optimal.
If I were him, I'd take this as good, friendly advice.

Incidentally, I should mention that the notion of "fairy economy" is a very complicated issue, which is not yet fully developed.
The underlying theme of this convention *seems* to be a preference towards greater orthodoxy.
But, there are times when that may deviate completely from what we might consider a true economy.

For example: if memory serves, Peter Gvozdjak showed some ortho-impossible cyclone theme on a 17x17 board.
Let's label that diagram A.
But, strictly hypothetically, suppose he had managed a diagram B, with an identical solution on a 16x17 board.
Finally, suppose he had managed a diagram C, with an identical solution on a 15x15 board, with some holes.
[note: I haven't actually found that squares can be saved in such a fashion -- but, just suppose it were true.]

Which, of these three versions, should we consider the more economical?
Whereas C may exhibit the best economy of squares, one may argue that the board is non-rectangular (perhaps even non-contiguous).
Furthermore, if you consider a hole to be a fairy unit (this is how it is often implemented), then it becomes clear that the addition of holes represents an infraction on the economy of fairy units employed.
Similarly, it may be argued that B is not square-shaped.

While it is clear that Peter had a justification to go beyond the 8x8 board, it is not immediately clear which version would exemplify the most economical form.
We don't worry too much about such issues, but if we're going to speak about economy of fairy elements, such issues can't be ignored.

It may be argued that our preference for orthodoxy leads us away from "true" economy (in this case, economy of squares are sacrificed).
In fact, it may be argued that special-case rules (castling, en passant, double-step pawn moves, promotion) actually constitute an infraction on an ideal economy.

The issue here, in my view, always centers around orthocentricity -- our general tendency to favor the FIDE variant, creates a false notion of economy in fairy chess problems.
Another example of a strange orthocentricity: some actually prefer to use a Grasshopper (which was invented first, and is more well known) -- when a simpler Rookhopper or Bishopper would suffice (if that's our notion of economy, it requires an explanation that is sure to strain logic).
In many ways, I think the simplicity of Shantraj/Chaturanga provided a superior model for orthodoxy.
But, I often think that we would be well served by engineering an orthodox "ground state" for chess, upon which all things fairy had a solid foundation, such that the economy of a problem might be more easily ascertained.

Finally, it's worth noting that there are, effectively, two versions of Circe Equipollents in use today.
A) pawns reborn onto the 1st rank behave normally (as pawns not on the 2nd/7th ranks), and
B) pawns reborn onto the 1st rank are dummies.

I believe that the former version was the original intent (and a popeye bug has spawned the latter version).
There is good "orthocentric" support for this: consider that there is no orthodox case where pawns behave as dummies.

Now, there's two ways to correct this issue:
1) you may require an additional fairy constraint, which alters the default behavior of pawns on the 1st rank, or
2) you may fracture one fairy condition into two different names.

It should be immediately clear that neither interpretation is more economical than the other (so you might prefer the second solution).
However, the first solution might better provide for similar issues in other conditions (Circe Exchange, T&M, Circe Parrain, Circe Antipoden, Einstein Chess, etc).

Fairy economy can not be surmised by simply summing the fairy elements listed.
Without a meaningful classification system, an holistic approach is required.
But, none of these complexities provide a sufficient justification to ignore the good principle which suggests that fairy elements should be employed only for good purpose.
(Read Only)pid=8609
(3) Posted by Julia Vysotska [Friday, May 25, 2012 10:52]

Dear Kevin, thanks for your comment! But in case if you're expecting Mr.Petkov's answer -
I believe, you'd better write it under his comment at the site.
I won't retranslate it from here to him...
Myself I'm not going to take someone's side - I'm just learning.
I appreciate Mr.Petkov's theoretical articles a lot!
And the explanations I've got from him about fairy pieces are simply great.
I'm happy to learn from him.
But the same time I'm interested in the points of view of other composers, as I believe
that everyone has something unique to show and everyone has his own preferences,
and, after all, if we speak about aesthetics in composition - there're lots of very
subjective things. I think that everyone can take something from the comments of
experienced composers. Maybe not all.
Well, of course, I can tell only about myself or how I see this.
Like, listening to different people, I feel what I accept or not for myself.
And, probably, this way is formed my feeling or understanding of what is aesthetically
or technically good looking problem.
I'd like to say Thank you! to Juraj and also to the others who's written at my site,
who also commented my problems.. as otherwise - how to learn how to become better?..
(Read Only)pid=8610
(4) Posted by Kevin Begley [Sunday, May 27, 2012 02:05]

@ Julia,

>I appreciate Mr.Petkov's theoretical articles a lot!
>And the explanations I've got from him about fairy pieces are simply great.
>I'm happy to learn from him.

A appreciate Petko's perspective, as well; and, I too have learned a great deal from him.
It's not really necessary to translate my comments from here to there.
I'm confident that Petko will investigate whether Nightriders are necessary in his problem; just as I'm confident that he will find a way to avoid them, if at all possible.
(Read Only)pid=8612

No more posts

MatPlus.Net Forum Fairies Discussion around P.Petkov's fairy problem at