21:21 CET

Remember me

 Rating ListJanuary 1st2013

MatPlus.Net Forum General What is Fairy Chess?

Page: [Previous] [Next] 1 2 3

Err, I take back what I said about #n in the "Black moves only to Parry Check (otherwise idles)" condition.
Since black is only moving when checked, there is never a need apply threat-notation.
There is never a threat when a player is in check (we simply apply all full-length variations).

However, a more intelligent threat-notation is required in some other conditions within the idle-mover family.
For example, in the "Black moves only to check (otherwise idles)" condition, threat-notation should apply to any white move which permits black to play.
I don't expect zugzwang threats would require any special treatment here -- the notation should still show all defenses (against the ZZ threat), save those which are not full-length (which are already dropped in standard threat-notation).

This brings up another good point about a recursive stipulation form...
Win Chloe implements pser-h#n by applying a rather convoluted condition: "Les Blancs ne jouent que s'ils sont en échec ou pour atteindre le but".
Roughly translated, this condition reads: "White moves only to parry check, or to reach the aim."

This is not the most clear expression.
Simpler would be to declare that the primary goal is to achieve a position in which a stipulation proves true -- in this case, black's primary goal is to reach a (h#0.5) position -or simply- a (#1) position.
The ultimate objective is mate <#>.

Thus, we could better express pser-h#n as: H*(<#>1)n + "* = White moves only to parry check".
H* indicates that the condition only applies up until reaching the primary objective (whereas, we normally would presume that a fairy condition applies universally).
So, upon reaching the primary objective (a position where white has #1), we drop the fairy condition, and recursively resolve the inner parenthesis (solve white's #1, down to <#>).

This also shows how a problem can be layered.
If you have a desired s#1 position [read: O(s<#>0.5)1], you can easily add an outer layer of parenthesis, to transform it into a hs#n -- thus, H ( O(s<#>0.5)1 ) n.
[note: give or take a 0.5 at the end, which results from the poorly formed hs#n stipulation]

And, if you want to add another direct-play layer onto this hs#n (generally, we presume n=1, but it need not be), you can make a direct-hs#m, by simply adding yet another layer -- thus, O(H(O(s<#>0.5)1)n)m.

HOHOHO. :)

...However, without a threat-notation, you can not define a dual
...Zugzwang is already well covered by threat-notation -- all legal moves are considered a defense to a ZZ threat.

There is the trouble - dual in any variarion in ZZ is seriously taken, but in threat-problems - in variations where black doesn't defend against a supposed threat, duals are forgiven. The concept of a threat as it is now, is illegal and absurd according to the chess rules. It well serves for virtual reasons but should not decide about the correctness of a problem.

A stupid example: 1.Sg6! -> 2.h8Q(R)#; 1...Be5 2.g4# - dualistic promotion is not a dual and 1...B~ 2.h8Q(R)/g4# is not dual and what about 1...Bh4 2.h8Q(R)/g4/Sf4# - not a triple mate. All according to a seriesmover rules 1.Sg6 2.h8Q(R)# - illegal and absurd in orthodox chess! Without a legal threat there are no 'defences'. But remove wPh7 - there is ZZ with a serious dual.

Nikola,

Hmm, I think I understand what you're saying, but I still don't believe you're actually challenging threat-notation.

In your example, if white is threatening >2.h8Q/R#, there is simply no use to consider black moves which do not prevent the threat -- right?
White has a threat, and anything that doesn't account for it should be discarded.
This seems straightforward, and logical -- so, why do you consider this absurd?

But, I'm open-minded, and very curious to know where this leads...
So, if threat-notation is inadequate, what would you suggest we replace it with?

Remember, if we ignore threats entirely, then every single legal black move would constitute a valid defense.
We would be swimming in duals.

Or, are you saying that threat-notation only breaks down in situations where the threat is zugzwang?
If so, then... given that white has no threat, what else can you do, except consider every legal move?
How do you propose to resolve that?

And, what would you suggest we do about variations which lead to mates of less than full-length?

(44) Posted by Nikola Predrag [Thursday, May 10, 2012 17:07]; edited by Nikola Predrag [12-05-10]

Kevin,
As I said I accept the traditional concept of a threat, although I believe it is absurd. But you mentioned threat several times while speking about fundamental and pefectly clear axioms in chess (composition) and I only tried to warn you that some traditional absurdities are not a good basis for new definitions.

According to chess rules, White can only prepare a particular continuation which he will play AFTER Black plays particular move(s). That can be called a threat. The same prepared continuations happen in ZZ and we can also call them (prepared) threats. Nothing in chess rules allows a difference between a 'threat' and 'ZZ'. A virtual threat is absurdity best shown when Black is stalemated, what is the real meaning of a threat before Black plays his move?

The results from FIDE Olympic Tournament 2012, co-organized with WFCC, can be seen in another thread of this Forum. Here a comment concerning the groupings of the problems, therefore it is in this thread. What is said here is also valid for other tourneys co-organized by WFCC.

The above mentioned competitions have separate groups for on one hand Fairies and the other hand Retros & proofgames. However, in the Retros & proofgames section have been competing problems with fairy conditions (circe, anticirce etc.). If one composes a direct twomover with circe condition, it is considered to be a fairy composition. If one composes a retro with circe condition, it competes in the retro section. Of course, a tourney organizer can group the entries in whatever way; in this case the titles of the groupings are misleading. Accepting fairy compositions among retros is also the practice of many renowned magazines.

As there is no definition for fairy chess, we can not state that for retro problems this definition is different. We only happen to have the situation, that fairy compositions are accepted within the Retros & proogames. For all other problem genres, the fairy compositions are shuffled into Fairies. Is this irregular sitution documented anywhere? Does anybody know the origin of the acceptance of fairies among retro problems? Is this partly the reason for the scarce participation in the retros and proofgames section, i.e. composers of orthodox retros do not want to compete with fairy problems, where a composer can invent a new circe condition in order to get his problem correct?

We, makers and solvers of problems, have a problem to solve. There is one solution and two thematic tries. Tries: 1) continue as presently and hope that the issue gets as little attention as possible 2) rename the groups to be on one hand Fairy problems excluding Fairy retros & Fairy proofgames and on the other hand Retros & Proofgames & Fairy retros & Fairy proofgames. Solution: for this issue, but mostly for more important issues, rewrite Codex Chapter II. - No other solutions found.

In reply to Joshua at http://matplus.net/pub/start.php?px=1345073725&app=forum&act=posts&fid=gen&tid=1050&pid=8418#n8418

Siegfried Hornecker
MatPlus 32, no. 1175
White wins
No Capture

1.a5 b5 2.Ka3! b4+ 3.Kb3zz Kb7! 4.a6+ Ka7zz 5.Ka2! b3+ 6.Kb2zz Kb8! 7.a7+ Ka8zz 8.Ka1! b2+ 9.Kb1zz and white wins

IIRC, Vaclav Kotesovec had a Köko study in the FIDE album (or in the Annexe). And on his website there are some articles about fairy EGTB research (http://web.telecom.cz/vaclav.kotesovec/ckpo25l/paovaoleo.htm, http://web.telecom.cz/vaclav.kotesovec/ckpo25l/ckpo25l.htm, http://web.telecom.cz/vaclav.kotesovec/ckpo25l/vc41ge6.htm, http://problem64.beda.cz/silo/kotesovec_fairy_endgames_2008.pdf)