Website founded by
Milan Velimirović
in 2006

18:04 UTC
ISC 2021



Remember me

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

Rating lists


MatPlus.Net Forum General Terminology Corner
You can only view this page!
(1) Posted by Andrew Buchanan [Saturday, Feb 20, 2021 12:09]

Terminology Corner

Hi there folks.

Would like a little area where occasional question of terms and definitions can be aired, and agreed without overlong debate. I will keep this base note as a table of contents and summary of conclusions:

(1) anti-Phoenix
There are 118 PDB compositions with the unexplained keyword 'Antiphönix'. So what does this term mean exactly, please?
Conclusion: deprecate

(2) retro promotion terms for forward problems?

(Read Only)pid=20690
(2) Posted by Neal Turner [Saturday, Feb 20, 2021 12:37]

It seems to be that a pawn promotes and the newly promoted piece is subsequently captured.
(Read Only)pid=20691
(3) Posted by Andrew Buchanan [Saturday, Feb 20, 2021 12:43]

Thanks Neal... how is that different from Ceriani-Frolkin capture?

According to a doc by Ian Shanahan, Anti-Phoenix is "capture of a man on its game-array square by its opposite counterpart" Trying to wrap my head around that. Why would that be termed "Anti-Phoenix"? Can the captured unit be Pronkin, i.e. gone back to an apparent game-array square. Is the opposite counterpart e.g. bS from b8 capturing wSb1, or will any similar unit including promoted suffice?

See, which is a completely different meaning from Ian's idea. "A promoted unit captures a similar unit (not promoted), but nothing to do with starting square. The promoted unit does not get captured at any point." Page 29 - it's the theme of a tournament. Or maybe I am misunderstanding what Jurmala means. Maybe it's something else and they assume we all know.

This term is not in V&V Encyclopedia at all.
(Read Only)pid=20692
(4) Posted by Jakob Leck [Saturday, Feb 20, 2021 22:18]

I noticed this term also a few weeks ago when searching for some helpmates and had the same understanding of it as Neal.
Seems to be some sort of time reversal of the Phoenix and in this sense the "anti-"Phoenix, but to my mind this is not such a good idea. And yes, Andrew, I don't see a difference to Ceriani-Frolkin either (except that the latter term is usually associated to the retro genre).

Btw. if I were to nitpick - and I do that occasionally :) - then in an anti-Phoenix, wouldn't there be a promotion followed by the capture of a piece of the same kind, but not the promoted piece?
I definitely wouldn't recommend using the term anti-Phoenix as it is done in many of those pdb entries you mention.
(Read Only)pid=20694
(5) Posted by Viktoras Paliulionis [Tuesday, Feb 23, 2021 22:03]

Some helpmate composers use this term for Phoenix in reverse order, but I agree with Jakob Leck that this is a bad idea and this term is not needed.
(Read Only)pid=20708
(6) Posted by Andrew Buchanan [Tuesday, Feb 23, 2021 22:33]

Thanks all. I also solicited feedback from Andrey Frolkin who gave yet another definition:

The name “anti-phoenix” has only been used a couple of times and then it was replaced with “anti-Pronkin.”

I agree with all the comments, and will mark the term as deprecated. Can then try to look at the theme within each problem and classify it appropriately.

How appropriate is it to generalise terms like Ceriani-Frolkin, Pronkin, Phoenix & Schnoebelen from the retro space to the world of forward stipulations?
(Read Only)pid=20709
(7) Posted by Viktoras Paliulionis [Wednesday, Feb 24, 2021 00:04]

The terms Phoenix and Schnoebelen are already used in the helpmate genre. The term Pronkin, as I understand it, is a special case of Phoenix. The term Ceriani-Frolkin is useless for helpmates, because its definition is too broad: about 25% of helpmates-moremovers with black promotions satisfy this theme.
(Read Only)pid=20710
(8) Posted by Andrew Buchanan [Wednesday, Feb 24, 2021 04:03]

Thanks Viktoras,

I agree with almost all of what you say.

Phoenix is unfortunately narrow - it implies that the promotion happens after the capture of an earlier unit. In the retro world Prenix is a slightly strained neologism to cover the case where the two events occur in the other order. Nix is a happy word in English to cover both of these situations, suggesting “zeroing the balance”. In retros, Prenix/Phoenix is just one way to regularise promotions, the others being C-F and (as defects) non-standard material or obtrusive material.

Pronkin implies Nix, however a single capture of an original unit might allow multiple Pronkins. The balance might then be restored by C-F of all but one promoted unit. A promoted pawn itself might be captured, but I suppose that if that is captured then it doesn’t open a square on the second rank as a Pronkin target square?

Do you have a clear algorithm in mind for translating a proof game or forward solution into such achievements?

Yes, Ceriani-Frolkin does not mean so much in forward compositions, and I suspect nor would Prentos, the case where an officer captured the promoted unit.

Thanks again
(Read Only)pid=20711
(9) Posted by Viktoras Paliulionis [Wednesday, Feb 24, 2021 11:34]

In different sources there are different definitions of the Pronkin theme.

Pronkin: A promoted piece comes to the home square of a piece of its type and colour.

Pronkin: A promoted piece moves on the initial square of a captured unit of the same kind.

Pronkin: A promoted piece occupies the original square of an already captured piece of the same kind and colour.

I don't know which definition is correct. The definitions in WinChloe and in PDB are almost the same, but according to the Encyclopedia, the piece should move not to the starting square of the captured piece in the diagram, but to the home square (= game-array square).

Helpmate examples showing Phoenix:

If the second definition is correct, then most examples also demonstrate Pronkin theme. The problem of C.M.Fox (1930) demonstrate Ceriani-Frolkin + Phoenix. In the last example (Gavriliv, Zalokockij, Soroka, 1988) bB promotion-capture is Prenix, if I understood correctly.
(Read Only)pid=20712
(10) Posted by Joost de Heer [Wednesday, Feb 24, 2021 11:57]


It seems to be that a pawn promotes and the newly promoted piece is subsequently captured.

Then how do you explain Antiphönix in P0563547? In 2) none of the promoted pieces is captured, but it's labelled AP anyway.
(Read Only)pid=20713
(11) Posted by Jakob Leck [Wednesday, Feb 24, 2021 14:57]

The "Antiphönix" keyword in P0563547 must be an error.
(Read Only)pid=20714
(12) Posted by Andrew Buchanan [Friday, Feb 26, 2021 06:32]

It's natural that a word has nuances of meaning. However "Anti-Phoenix" has about 5 completely disparate senses. It's rare that a term is as broken as this. The question "what's the opposite of Phoenix?" is no easier to answer than "what's the opposite of a tree?" I have marked it in PDB as deprecated, and associated problems should be reclassified, before deleting the keyword. Some "Anti-" keywords are well established, and I don't want to disturb them, but this one isn't.

By contrast, different definitions of Pronkin share the central idea of a promoted unit visiting a start square, identified by the capture of its original kindred occupier. However, there are many details on which the definition can differ. A glossary definition should be inclusive, but not weak, and indicate the directions of divergence. If we want a system (as do WinChloe & Helpmate Analyzer) to identify automatically acrobatic feats conducted by the chess pieces, then we need precise definitions. There is a tension there, as we don't want to sideline problems based on variant definitions.

Vita brevis ars longa etc. Personally I think we should aim for greater automatic tagging in PDB, and use the time saved to classify the exceptions better, and support those with automatic detection too. It must be made clear how someone can modify the auto tagging in a way that won't be over-written automatically by an over eager function every time the composition details are modified.

Convergent or divergent thoughts welcome.
(Read Only)pid=20720
(13) Posted by Hauke Reddmann [Friday, Feb 26, 2021 08:45]

Convergent: Hey, it can't be THAT difficult to
write a program that has a FEN (and stipulation etc.)
as input and a theme index as output.
Divergent: Would make a nice theme for my
doctoral thesis in computer science - if only
anyone could supervise it. Sorry, I'm out :P
(Read Only)pid=20721

No more posts

MatPlus.Net Forum General Terminology Corner