The simulation does not contradict anything Bell ever wrote nor anything I ever wrote. It seems to me to have no physical relevance. But I’m just a statistician.
In particular, it seems to have no connection whatsoever with Oxford physicist J. Christian’s theories. But I earlier pointed out how F. Diether could make a connection. So far Christian and Diether have ignored this suggestion.
One can convert the model into a Joy Christian type model by randomly choosing the channel 1 or the channel 2 outcomes, using an extra random and if you like spinorial or quaternionic coin toss! There are then *two* hidden variables: the random (uniformly distributed) spin orientation *and* an independent fair coin toss, corresponding to Dirac’s belt trick, a kind of Möbius strip idea.
Of course, the two channels are not actually two real channels. Real experiments don’t have two sets of two detectors, both for Alice and for Bob! But your physical picture could be that only one of the two sets of events is real. Just as you say that a set of events with a zero value of A times B was no particle pair at all.
I’ll draw the plot of the particle-pair production rate in a couple of weeks, and further illustrate the mathematical connection with Pearle’s 1970 model. I still wonder if you would not get a better fit to the negative cosine using Pearle’s rather than Michel Fodje’s formulas. Right now you are playing with three or more arbitrary numerical parameters in an attempt to nudge the curve to the right place. Pearle has it exactly right, no arbitrary parameters. In fact, Pearle has a one-parameter family of models which all get it exactly right, he just took the most simple one as an example to make the point he was making: namely, the experiments of his day had a big loophole due to low detector efficiencies.