--- On Mon, 16/2/09, Michael Allan <m...@zelea.com> wrote: > OK, a time series Tx, Ty, ...
> Tx(p) ~= Ty(p) for all Tx, Ty [always] > > Tx(q) ~= Ty(q) for all Tx, Ty [usually] (p = practicality, q = probability) One can model the world in different ways. One could say that everything that will happen has probability 1 and everything else has probability 0. Same with practicalities. An alternative approach is to identify some area in the world that is assumed to vary. In this model p may set the space where the end result may vary, and q operates in the remaining space. The space of variation could be could depend on the interest group the discusses the alternatives (could be e.g. a country, a party, citizens, election reformists). One more observation on sequential transitions and IT based direct democracy style systems. If the political system of the USA was initially planned to operate without parties but then quickly changed to a system that has parties, then one could expect also the new system to have similar tendencies, i.e. interest groups would find ways to influence the system. Juho ---- Election-Methods mailing list - see http://electorama.com/em for list info