[Retros] 50-moves draw revisited

[Rol, Guus]

I wish to revisit the discussion about the role of castling inside - or may be outside depending on the verdict - a 50 moves series. The basis for the discussion is apparently article 5 in the Piran Codex:

Art. 5: Wenn die Lösung des Schachproblems mehr als 50 Züge erfordert [16 <http://www.janko.at/Retros/Glossary/Codex1958.htm#16> ], ohne das während dieser Züge ein Stein geschlagen, ein Bauer gezogen oder die Rochade ausgeführt wird, dann gilt die Stellung nicht als remis [17 <http://www.janko.at/Retros/Glossary/Codex1958.htm#17> ]. Wenn nachgewiesen wird, das in der Retro-Analyse und in der Lösung beiderseits insgesamt 50 oder mehr als 50 Züge - unter den erwähnten Bedingungen - ausgeführt worden sind, so wird die entstandene Stellung automatisch als remis gewertet [18 <http://www.janko.at/Retros/Glossary/Codex1958.htm#18> ]. Die Art. 3, 4 und 5 dürfen nur auf legale Stellungen angewendet werden.

The handling of rules is always endangered by this corruption scheme: (a) creative concept (b) rule translation (c) query (d) rule extension (e) concept corruption. This has clearly occurred in the case of article 5.
 
The first thing that can be observed about the 50-moves rule is that it is a Sudoku-tool, by which I mean that it is for "bookkeeping", not for "understanding", not for "definition". It actually conceals the concept that the rule is based on. What that concept is, becomes clear if you look at the integral result of a 50-move drawing series. What it begets us is (a) an initial position (b) the information that 50 time ticks (moves) have passed (c) an end position with all the pawns on identical spots as in the initial position and the same amount of pieces on the board. This tells us that nothing needs to be known whatsoever about the moves that were played in the process. The fact that a witness can observe the game, then at some point leave his observation post for the duration of 50 moves, return, and then tell us whether or not the position is a draw, proves that the concept has nothing to do with the properties of the moves played. Nevertheless, the properties of individual "moves" are the only thing that the 50-moves rule talks about in setting criteria. So the 50-moves rule is for the bookkeeping of game players. Still, it worked fine in its original form as a practical translation of the 50-moves concept.
 
Then came along "the castling query". It was decided that "something about castling" needed to be conceived as "essential progress" in the game and therefore should be incorporated in the 50-moves rule. Rather than revisiting the original concept of the rule - the required action - attention was focused on the adaptation of the existing rule. The objects available in that rule were "individual moves" and their standard operators like "a pawn move ", "a capturing move". What could be more fitting than to insert "the castling move" (legality implied) in that series? Why would anyone be so silly to add to the list an ugly duck like "a castling right losing move"? After all, "losing castling right " is not a property of the move but of the position! And who is interested in positions in a rule that only talks about moves?
 
So that is how the 50-moves rule extension corrupted the 50-moves concept. The measurement of progress is the "comparison of states" and not the volume or nature of activity generated. By that distinction the castling query should have resulted in a rule adaptation somewhat like this: (a) an automatic draw occurs if the essential properties of a position have not changed in 50 moved (b) essential property changes are pawn moves, captures and change in (=loss of) castling rights.
 
By the way, the en passant query is of a different nature which I will write about at another time
 
Guus Rol.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pairlist.net/pipermail/retros/attachments/20070104/a22f8877/attachment.htm>