Ghenadie Mardari's recent works in a nutshell

[Richard Gill]

Dear Ghenadie, dear all,

I think I got it figured out at last.

Ghenadie’s basic spinning disk model
============================

8 equal segments of a disk are marked "A1+", "B1+", "A2+", "B2+", “A1–", “B1–", “A2–", “B2–“.

The disk is spun and comes to a stop at one of the 8 boundaries between two segments.

For example, between the 4th and the 5th, is between “B2+“ and “A1–“.

This gives us one line of a data file: Alice chose setting 1 and got outcome “–“, Bob chose setting 2 and got outcome “+“.

The pair of outcomes is always the same, when the setting pair is 11, 21, or 22, but always opposite when the setting pair is 12.

Repeat independently again and again.

The result is three correlations of +1 and one of –1; S = 4

For Alice and Bob separately, + and – are equally likely

This is actually a non-local hidden variable model with a hidden variable taking 8 different values all with equal probability.


(almost) Maximally simple LHV model
=============================

Alternatively, let both Alice and Bob choose a setting at random, let an outcome “=/–“ be chosen at random.

Both measurements are given that same outcome.

The result is four correlations of +1; S = 2.

For Alice and Bob separately, + and – are equally likely


This is a local hidden variable model with a hidden variable taking two equally likely values.


Mixed model
==========

Pick the above model 1 or model 2 at random, again and again.

Three correlations of +0.5, one of –0.5. S = 3.

However, the final computer programs which he has had written for him are far more complicated than necessary and don’t accurately match the description in the latest paper.

Instead, his programmer simulates the basic spinning disk model (with a very elaborate program) but randomly flips a sign somewhere to reduce S from 4 to 3.




Ghenardie has written a lot of papers on this theme in recent years, I found the first one very clear but as time went by they got more difficult to read.

I should have read it earlier.

Richard