[Retros] 50 moves rule

[Rol, Guus]

I see I have made some typo's and other errors in my original message.
Below I edited out 'the big one'. Guus Rol.



> -----Oorspronkelijk bericht-----

> Van: retros-bounces at janko.at [mailto:retros-bounces at janko.at]

> Namens Rol, Guus

> Verzonden: dinsdag 9 januari 2007 10:28

> Aan: The Retrograde Analysis Mailing List

> Onderwerp: Re: [Retros] 50 moves rule

> 

> 

> Yes, that is an important question. To undercut this

> discussion I have used "positional properties" rather than 

> "position" in my comments. Surely castling and e.p. rights 

> are properties of positions. For me "positions" are synonym 

> to "states". States contain all present time information 

> of a system pertaining to possible future developments. 

> Besides the usual suspects "diagram", "castling/e.p. rights" 

> there are also more obscure factors involved like "how far we 

> advanced in the 50 moves count" and "all positions eligible 

> for repetition + repetition counts". In my comment to Tom 

> Volet I argued that it would be unwise to include the latter 

> two in the evaluation cycle of the the 50-moves/repetition 

> rules since that would lead to oscillating evaluations - the 

> evaluation process affects its own outcome.

> 

> Going into specifics, one might think that the analytical

> impossibility of future castling conforms to the state 

> requirement of being "incapable of affecting the future 

> developments in the game". If that were the case, then the 

> castling right property in such positions would have to be 

> designated as "false". However, a repetition query of that 

> nature was raised (I think somewhere in the 1960s)in actual 

> tournament practice with the familiar move sequence

> Sd5-c7+,Ke8-e7,Sd5+,Ke8,Sc7+,Ke7,Sd5+,Ke8,Sc7+; draw? The resounding

> FIDE arbitration was NO! I tricked you slightly by adding the last

> move Sc7+ which was redundant in the original issue. It was 

> included here to conform to your wish to look at a position in 

> which (a) castling right was initially present (b) castling 

> has become an impossible future option

> (c) a termination criterium (triple repetition) was 

> apparently met.


Replace (a) to (c) by:
(a) black castling right was still present after the first Sd5-c7+ (b)
but since the knight could not be captured, it was also certain that
black would never be able to castle in the future (c) a termination
criterium (double repetition) was seemingly met during the play sequence


> The verdict shows that the FIDE has chosen to

> look at castling right as statutory (I hope this the 

> appropriate english designation) chess law. In other words 

> "the condition of castling right is not dictated by the 

> possibility to execute castling at some point"; I assume the 

> same is true for en passant. This is probably another wise 

> decision as it keeps the need to analyze positions away from 

> the rules. Also, it is easy to create ambiguous situations 

> between castling rule and 50-moves rule if you choose the 

> alternative option. Even when I adopt a highly provocative approach 

> to the FIDE/Codex building, it is not my intention to destroy 

> the premises. Just to improve the living conditions.

> 

> Guus Rol.

> 

> 

> -----Oorspronkelijk bericht-----

> Van: retros-bounces at janko.at [mailto:retros-bounces at janko.at]

> Namens Seth Breidbart

> Verzonden: vrijdag 5 januari 2007 20:40

> Aan: retros at janko.at

> Onderwerp: Re: [Retros] 50 moves rule

> 

> 

> It seems to me that the underlying issue is the definition of

> "position".  Is it a photograph of the pieces on the board?  

> That plus information as to whose move it is?  That plus 

> information as to who might be allowed to castle (and on 

> which side) in the future? (Consider a position in which 

> White has not castled, his King and Rooks have not moved, but 

> it can be proven that there is no future play which involves 

> White castling; how does that count?)

> 

> Seth

> _______________________________________________

> Retros mailing list

> Retros at janko.at http://www.pairlist.net/mailman/listinfo/retros

> _______________________________________________

> Retros mailing list

> Retros at janko.at

> http://www.pairlist.net/mailman/listinfo/retros

>