Bell theorem without hypotheses?

[jk22]

i found the following proof of Bell's theorem :

we measure spin in 3 different directions a b c we can note the counting of events

N1=n(a+,b+,c+)
N2. + + -
N3. + - +
N4. + - -
N5. - + +
N6. - + -
N7. - - +

We have N3+n4<=n7+n3+n4+n2

With n3+n4=p(+a,-b)
N7+n3=p(-b,+c)
N4+n2=p(+a,-c)

It is violated by quantum mechanics but i don't see where the hypothesis of locality and reality comes into play.

 

[DrChinese]

The answer is that you can replicate the results with a classical system if either non-locality or non-realism are allowed. Surely, you could imagine non-local mechanisms that would allow 2 particles to mimic each other in just the right amount to give quantum results.

Harder to picture non-realism. But essentially you are assuming this when you have N1=a+, b+, c+ because you only ever measure 2 of these at a time at most.