http://volterra.mat.uniroma2.it
This is only the most recent contribution to a nearly fourty year old controversy, as to what the elementary probabilistic inequality, concerning four 0/1-valued random variables X1, X2, Y1, Y2 :
Pr(X1=Y2) <= Pr(X1=Y1) + Pr(X2=Y1) + Pr(X2=Y2)
has to say about the experimentally confirmed prediction of quantum mechanics, that certain counts in an experiment with polarized photons can satisfy:
N12 >> N11 + N21 + N22 .
I have bet Luigi Accardi 1500 Euro, that his computer programme cannot be written according to certain specifications, reflecting the spatial structure of the experiment with the photons. I will explain the background to this bet, and use some martingale theory to show why even time-dependence cannot be used to escape Bell's inequality.
Draft paper:
http://www.math.uu.nl/people/gill/Temp/constance.pdf
Last Updated: September 17, 2001