SciPhysicsForums.com

The ultimate local realistic simulation that shoots Bell down for good! 2 hidden variables.



Not perfect yet but good enough to beat CHSH. I'm still tweaking it and trying to do some simplifications. Plus the code is a mess right now.
.

.
Great work, Fred. But I will have to see the analytical functions A(a, h1, h2) and B(b, h1, h2) written down explicitly to believe it. Here h1 and h2 are your two hidden variables.
.

Joy Christian wrote:.
Great work, Fred. But I will have to see the analytical functions A(a, h1, h2) and B(b, h1, h2) written down explicitly to believe it. Here h1 and h2 are your two hidden variables.

Me too! Keep on trying, Fred.

Joy Christian wrote:.
Great work, Fred. But I will have to see the analytical functions A(a, h1, h2) and B(b, h1, h2) written down explicitly to believe it. Here h1 and h2 are your two hidden variables.
.

Thanks. Easy! One is your HV +/-1, the other one is from Michel's epr-simple.
.

I'm still tweaking on it but here we go! 3 million trials; 98 percent of particles detected.





It's only going to get better from here. No more Bell junk physics!
.

Already tweaked it better. 2 million trials at essentially 100 percent particle detection.



And..., they said it was impossible. Nuts I say.
.

.
Brilliant, Fred. I suggest you ask Jay to have it patented, immediately. You know there is that person who goes around first trashing and then stealing other people's work.

But first check that << A >> = 0 and << B >> = 0.
.

Joy Christian wrote:.
Brilliant, Fred. I suggest you ask Jay to have it patented, immediately. You know there is that person who goes around first trashing and then stealing other people's work.

But first check that << A >> = 0 and << B >> = 0.
.

Jay already told me that you can't really patent stuff like this but it is time stamped on the forum here. But I'm going to try to simplify some more anyways.
.

FrediFizzx wrote:

Joy Christian wrote:.
Brilliant, Fred. I suggest you ask Jay to have it patented, immediately. You know there is that person who goes around first trashing and then stealing other people's work.

But first check that << A >> = 0 and << B >> = 0.
.

Jay already told me that you can't really patent stuff like this but it is time stamped on the forum here. But I'm going to try to simplify some more anyways.
.

Fred, could you post the Mathematica code when you get a final version so I see how it works?

jreed wrote:Fred, could you post the Mathematica code when you get a final version so I see how it works?

Hi John,

I'm still tinkering with it but I had to make sure it works with CHSH. So here is the notebook file for CHSH.

CHSH = 2.74565 using the Delft angles.

EPRsims/bingo2CHSH.nb

Enjoy!
.

Here is a better CHSH example Mathematic notebook file with more standard angles.

EPRsims/bingo2CHSH2.nb

CHSH = 2.89823! In fact we have exceeded the QM bound but of course this is not a QM model.

Enjoy! Bell is shot down all to pieces now! Finished! Kaput!
.

FrediFizzx wrote:Here is a better CHSH example Mathematic notebook file with more standard angles.

EPRsims/bingo2CHSH2.nb

CHSH = 2.89823! In fact we have exceeded the QM bound but of course this is not a QM model.

Enjoy! Bell is shot down all to pieces now! Finished! Kaput!

Are << A >> = 0 and << B >> = 0 satisfied for all A and B results?
.

Well, I'm still tweaking it but I can get with 10,000 trials,

AveA = -0.0630784
AveB = 0.0981491

There are 9 different parameters that I can adjust. It is just a matter of finding the sweet spot. Plus I still need to try to simplify. Got sidetrack by the CHSH stuff.
.

Here is a version of the CHSH example with 3 Do loops. IOW, A and B stations are completely independent and in fact run at different times. 5,000 trials.

CHSH = 2.53455!

Here are the PDF and Mathematica notebook files,

EPRsims/bingoCHSH_forum2.pdf
EPRsims/bingoCHSH_forum2.nb

Enjoy the fact that Bell's theory and Gill's theory are completely shot down now.

Here's an even better shoot down of CHSH!

CHSH = 3.01331!

If you aren't yet convinced that CHSH is a pile of junk, then you're lost.

EPRsims/bingoCHSH_forum4.pdf
EPRsims/bingoCHSH_forum4.nb
.

Here is a new version where I have expanded out a-b to cover the whole range of data to 720 degrees.



I've now got 25 adjustable parameters that I can use to tweak it even more. Here are the PDF and Mathematica notebook files.

EPRsims/posneg_forum2.pdf
EPRsims/posneg_forum.nb

Enjoy! Next I will be doing this with 3 separate Do loops for the particle and HV generation and then independent ones for A and B.
.

FrediFizzx wrote:Here is a new version where I have expanded out a-b to cover the whole range of data to 720 degrees.

…

I've now got 25 adjustable parameters that I can use to tweak it even more. Here are the PDF and Mathematica notebook files.

EPRsims/posneg_forum2.pdf
EPRsims/posneg_forum.nb

Enjoy! Next I will be doing this with 3 separate Do loops for the particle and HV generation and then independent ones for A and B.

I am greatly enjoying this. You sure have enough parameters to tweak. Have you thought of trying modern numerical optimisation methods? Lots of people got very close indeed with this sort of trial and error. I think of Michel Fodje and Caroline Thompson. Like Aristotle with his epicycles, who certainly beat Copernicus. But Kepler and then Newton cinched it. Circles became ellipses (more parameters!, but less conic sections) and then Newton’s laws explained Kepler’s, with much less parameters. In the meantime, Galileo saw the moons of Saturn.

Pearle showed 40 years ago how to get it exact.

Meanwhile, the standard model sure did fine for a long time, but I understand that it does have an awful lot of parameters.The Higgs’ boson fitted in that picture, but it is not clear if that was for real or not.

@gill1109 Of course..., more rambling nonsense. Do you have something more specific to say about this simulation?
.

As promised here is the 3 Do loop version. All independent and run at different times. One for particle and hidden variable generation and then separate A and B loops for detection. Note that I now have both hidden variables as functions of the singlet vector so all you need is the singlet vector and the a and b vectors to predict the A and B outcomes.

I could only run 200,000 trials at one degree resolution as the trial number matching function is very time consuming. I tried 300,000 and aborted after 12 hours of running because I had other things to do.



But easy to see that it is converging to -a.b. Here are the PDF and Mathematica notebook files. I did do some minor tweaking of the adjustable parameters.

EPRsims/posneg_forum3.pdf
EPRsims/posneg_forum3.nb

Enjoy! And ask questions if you don't understand something.
,

.
Congratulations, Fred. Someone is converging to eating their hat in public, as they boasted some years ago!
.