[Retros] Illegal moves by grandmasters

[Guus Rol]

My book on the retrograde type is purely mathematical. Even though exactness
of definition has never caught on in this field, I have allways felt the
motivation for starting retro-analysis was the same as for starting science
in general. If we rate forward chess as "law based, sometimes intelligent
design" then the retro type serves to extract certain or justifiable
knowledge from those designs - with limited or no direct information on the
design process. In an abstract domain like chess, the applicable scientific
approach can only be "mathematics".

The rules of chess do not define the legality of positions. The definition
of a legal position is therefore a mathematical derivative: "A position is
legal iff a legal proof game for it exists". The legality of the proof game
can only be established by the "context rules". Mathematicians do
not discuss those in terms of "right" or "wrong". A position may be legal in
one context (one set of postulates) and illegal in another (another set of
postulates). Who cares? the discussion about the grandmaster game is only
valid in the context of the FIDE game rules and I think it has been properly
conducted in that context. The retrograde field will fundamentally inherit
all FIDE rules since it is set up as a slave system to that forward master
system. Exceptions can be made as explicitly deviant postulates at the risk
of having the disobedient servant fired by the master. In 1997, all
retro-designs based on automatic 50 move termination were fired by the PCCC.
The allmighty one had decided that its postulates strayed to far from the
"right path" in the preceding decades. As of this day, all composers suffer
PTSS from this event and have nu clue of what happened at the time.

Andrew Buchanan has very cautiously copied his postulates from the FIDE and
can therefore be considered of the "right belief". Anyone bothered by them
should say so in his "problem stipulations" (divine prerogative) or he
should try to get his preference in the retro codex.

Mathematics goes a long way if you only try a little.


Guus Rol.




On Tue, May 10, 2011 at 8:09 PM, <tregeryefim at aol.com> wrote:


>

>  The (composed) dead positions published by Andrew Buchanan are indeed

> illegal as no legal game can be construed to arrive at them. The "law" does

> not always look the same in forward and backward direction.

>

> What does "legal game" mean? This term is also may have several meanings.

> Upon Fide rules, indeed, a game cannot contain two or more dead positions.

> But we may consider a game as a sequence of any legal positions, regardless

> of fide rules (of dead positions, 3-times, 50 moves, etc.). Then there is a

> set of all such legal games (even infinite ones!) as a mathematical

> object... Interesting that this set is uncountable (why - I may explain it

> separately), yet the subset of all finite games (finishing by checkmate or

> stalemate positions) is countable.

>

> Yefim Treger

>

> _______________________________________________

> Retros mailing list

> Retros at janko.at

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

>

>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pairlist.net/pipermail/retros/attachments/20110512/555bf766/attachment.htm>