Graft's arXiv:1404.4329 paper

[Guest]

What is nonresponsive? You claim to know what nonmaximal states are, so you must accept that inserting 1/2 at will into the CH inequality is nonsense. Let's see your new analysis that proceeds without that error.

I guess Don is not smarter than a 5th grader so I speak the truth.

Is this what passes for logic in your narcissistic world, Fred?

Keep it coming, Fred, you're making a total fool of yourself.

[FrediFizzx]

What is nonresponsive? You claim to know what nonmaximal states are, so you must accept that inserting 1/2 at will into the CH inequality is nonsense. Let's see your new analysis that proceeds without that error.

I guess Don is not smarter than a 5th grader so I speak the truth.

Is this what passes for logic in your narcissistic world, Fred?

Keep it coming, Fred, you're making a total fool of yourself.

Amazing! Don lost the debate already and doesn't even know it as one can easily tell from his statement above. Do you know what "simple inspection" means?

[Guest]

Do you know what "simple inspection" means?

I asked you to re-state your analysis correcting the error of omitting the possibility of nonmaximal states. But you didn't do so, because you are a coward, a charlatan, and a bully. So tell us, Fred, what "simple inspection" leads you to your conclusion that the bound is 1/2?

[FrediFizzx]

Do you know what "simple inspection" means?

I asked you to re-state your analysis correcting the error of omitting the possibility of nonmaximal states. But you didn't do so, because you are a coward, a charlatan, and a bully. So tell us, Fred, what "simple inspection" leads you to your conclusion that the bound is 1/2?

I didn't do it because I don't need to. Something you would understand if you knew what "simple inspection" is. Here we go again. If you claim the terms in the CH74 inequality can range from 0 to 1 then we could have as a possibility the following for independent terms.

1 - 0 + 1 +1 -1 -1 = 1

If the range is 0 to 1/2, as anyone can simply see, then we could have as a possibility,

1/2 - 0 + 1/2 +1/2 - 1/2 -1/2 = 1/2

That is the truth by simple inspection for independent terms. Of course if the terms are dependent like shown in the derivation of CH74, then the upper bound is surely 0. But that is not what QM nor the experiments nor your demonstations do. They use independent terms. You lost the debate. Get over it!

[Guest]

The problem is that not everybody is smart enough to see things so deeply as you. You can't just call them all stupid if you want your theory to be accepted. So this "simple inspection" stuff is dubious. Anyone can say that everybody else is stupid or mathematically insane. You gave us two equations that simplify to:

1 = 1

and

1/2 = 1/2

but for stupid people, that's not enough. Stupid people want to know how you know when terms are dependent or independent in an experiment. Maybe you could explain that to the stupid people. TFH (Thanks For Helping).

[FrediFizzx]

The problem is that not everybody is smart enough to see things so deeply as you. You can't just call them all stupid if you want your theory to be accepted. So this "simple inspection" stuff is dubious. Anyone can say that everybody else is stupid or mathematically insane. You gave us two equations that simplify to:

1 = 1

and

1/2 = 1/2

but for stupid people, that's not enough. Stupid people want to know how you know when terms are dependent or independent in an experiment. Maybe you could explain that to the stupid people. TFH (Thanks For Helping).

Is that all you got? Of course it is because of the truth of the matter. BTW, we do enjoy teasing Bell fanatics since they don't really have a leg to stand on. It is only a matter of time before Bell's theory is relegated to the dust bin of other failed no-go theorems. You should save yourself while there is still time.

As far as knowing if the inequality terms are dependent or independent in an experiment, it is quite simple. They are always independent because it is impossible to do the dependency in an experiment. Now... as far as QM goes, it is possible to formulate the probabilities to conform with the dependency as seen in this paper. Then there is no "violation" of CH by QM. Otherwise QM shifts to the inequality with the higher upper bound and still no violation of that inequality.
****

[Guest]

Is that all you got?

Yeah, it must seem pretty lame to you. Thing is, you said "it is impossible to do the dependency in an experiment". Why is it impossible? Dont' forget, you have to convince stupid people. You can't just make bald assertions.

[FrediFizzx]

Is that all you got?

Yeah, it must seem pretty lame to you. Thing is, you said "it is impossible to do the dependency in an experiment". Why is it impossible? Dont' forget, you have to convince stupid people. You can't just make bald assertions.

You should be able to answer that yourself since you did so much work analyzing the experiments. Think about how the data is measured, collected and analyzed for an experiment. You end up with independent sub-experiments for each term. IOW, for the following expression,

p(a, b) - p(a, b') + p(a', b) + p(a', b') - p(a') - p(b),

you can't do all terms all at once. You will have an iteration that will be b - a then perhaps an iteration that will be b' - a' and so forth. So you end up with A and B outcome results for each term independently. Pretty common sense.

[Guest]

What if two experiments both gave 1 (or some other value)? They would be dependent in your sense. You say that an experiment can never do that? Why not?

[FrediFizzx]

What if two experiments both gave 1 (or some other value)? They would be dependent in your sense. You say that an experiment can never do that? Why not?

That question does not parse at all. Why the heck do you think it has anything to do with what I am saying? Please don't ask boring questions. You were doing OK before this.

[Guest]

You said two terms of the CH inequality were "dependent" because they both contained a common term (one of the singles probabilities). Here the common term is 1.

Look, Fred, I know this is emotionally hard for you, but if you want to be taken seriously, you have to provide a clear mathematical definition of "dependent terms". Your whole claim to fame hangs on it. Why won't you make it explicit?

And don't forget that I can do one experiment and have three friends all do the other experiments somewhere else, all at the same time. So don't go on about "independent means done at different times". As I told you, that contextuality stuff was discredited long ago. CH does not involve contextuality.

[FrediFizzx]

You said two terms of the CH inequality were "dependent" because they both contained a common term (one of the singles probabilities). Here the common term is 1.

Look, Fred, I know this is emotionally hard for you, but if you want to be taken seriously, you have to provide a clear mathematical definition of "dependent terms". Your whole claim to fame hangs on it. Why won't you make it explicit?

And don't forget that I can do one experiment and have three friends all do the other experiments somewhere else, all at the same time. So don't go on about "independent means done at different times". As I told you, that contextuality stuff was discredited long ago. CH does not involve contextuality.

All of this has already been discussed to death on this forum so it is pretty easy to tease Bell fanatics like you. LOL! No difficulty at all least not anything emotional. We don't really care all that much if you want to continue with your misguided program. However, we do care about the truth in physics. But there is still time for you to save yourself from a bunch of embarrassment if you wish.

You never did present a reference other than your own papers for this so-called discrediting of contextuality "stuff". Put up or shut up. Time to get off the pot and explain how it is possible for an experiment to conform to dependent terms. Your crap above doesn't cut it.

[Guest]

You never did present a reference other than your own papers for this so-called discrediting of contextuality "stuff".

Sure I did. It's CH74. The CH inequality is there derived without requiring identical ensembles of alpha for each experiment, i.e., the results of the experiments commute. Combining the results from commuting experiments in an inequality is not the same as sampling from a single distribution.

Your crap above doesn't cut it.

Your rhetoric indicates you are not interested in being taken seriously, so I'll see you around. Let me know if you ever write or publish anything.

[FrediFizzx]

You never did present a reference other than your own papers for this so-called discrediting of contextuality "stuff".

Sure I did. It's CH74. The CH inequality is there derived without requiring identical ensembles of alpha for each experiment, i.e., the results of the experiments commute. Combining the results from commuting experiments in an inequality is not the same as sampling from a single distribution.

Your crap above doesn't cut it.

Your rhetoric indicates you are not interested in being taken seriously, so I'll see you around. Let me know if you ever write or publish anything.

Rubbish! More Bell fanatic gibberish. Withdraw your papers, man. They are pure junk physics.

[FrediFizzx]

1. It is mathematically insane to think that it is possible to violate an inequality of Bell's type. If that is true then the inequality was false to start with. The CH inequality is no different and this argument by minkwe (Michel) applies equally to the CH inequality. I have read your paper that is the topic of this thread and referring to your eq. (11) we can say that the paired terms in the inequality can range from 0 to +1 and the single terms can range from 0 to +1/2. So for independent terms we could have,

+1 - 1 +1 +1 - 1/2 - 1/2 = +1 not 0.

So the actual absolute bound on the CH inequality is 1 not 0 with independent terms. If you look carefully, you will find the experiments shift to an inequality with this bound of 1 and they don't violate it.

So we need to fix this based on a new development from the CH Inequality thread. Since the CH inequality only deals with + and ++ counts, the range for all probability terms is 0 to 1/2. Plus the maximum result of the CH string for QM is about 0.207. So I believe the absolute upper bound of CH74 for independent terms is 1/2. We could have,

1/2 - 0 + 1/2 + 1/2 - 1/2 - 1/2 = 1/2

*****

So continuing on with this now that we have the upper absolute bound of 1/2 for independent terms, it is quite amazing that the Bell fanatics want to continue to restrict local hidden variable theories with dependent terms of an inequality whilst QM and experiments are not restricted. Fortunately, Joy Christian's classical local-realistic model "blows them out of the water". Anyways, we have presented here the most profound objection to Bell's inequalities ever. Graft should fix his paper or withdraw it that is the subject of this thread. He has a few other papers that contain misguided content also that should be withdrawn. For sure they will never survive scrutiny against the experiments. For some reason, Graft must think that quantum theory is wrong. Well, mostly it was his nonsense about marginal probabilities vs. joint probabilities. We think and expect quantum theory is correct. It is only the interpretation about entanglement that is wrong.

[minkwe]

I agree with Fred that the CH inequality can never be violated. Thanks for the "CERN" article, it very clearly shows that.

Adenier's paper which I've given in the past shows the same, as well as Rosinger's paper and deRaedt's paper, and Vorob'ev's paper. I will try to find the links and reference them here.

A paradigm shift is needed to see the problem, which originates from improper use of the same mathematical symbols to represent different numbers, with different statistical properties. A common problem in discussions of QM.

I will strongly encourage anyone in doubt to review Rosinger's paper.
http://vixra.org/pdf/1305.0129v1.pdf

Also, study the difference between strongly objective and weakly objective.
http://arxiv.org/pdf/quant-ph/0006014v3.pdf

Also, study Vorob'ev's cyclicities
http://www.panix.com/~jays/vorob.pdf

The last article predates Bell, and presents mathematical evidence against the idea that Bell-like inequalities can be applied to independent experiments. I have summarised such arguments here viewtopic.php?f=6&t=181#p4912

In short, applying Bell and CH inequalities to independent experiments involves a contradiction, as explained in my post linked above.

I hope the above is enough to provide the paradigm shift required to see what by now appears so obvious to us - No Bell-like inequalities can ever be violated, apparent "violations" can only arise by erroneously substituting independent (weakly objective)terms into an inequality requiring dependent (strongly objective) ones.

[FrediFizzx]

I agree with Fred that the CH inequality can never be violated. Thanks for the "CERN" article, it very clearly shows that.
... I hope the above is enough to provide the paradigm shift required to see what by now appears so obvious to us - No Bell-like inequalities can ever be violated, apparent "violations" can only arise by erroneously substituting independent (weakly objective)terms into an inequality requiring dependent (strongly objective) ones.

You're welcome and thanks. You would think that my "simple inspection" method for getting the absolute upper bound for independent terms would be simple enough for the Bell fans to understand. But as you can see from the discussions on the forum here that they are still in denial when presented with the truth.

[minkwe]

Yes Fred. It appears very obvious to us because we've been discussing this for a long time. But I understand why some people may be having a mental block needing a paradigm shift. That's why I always try to provide the same argument from many different perspectives so that perhaps one of them would trigger the required epiphany.

For example, a few years ago, I presented evidence (viewtopic.php?f=6&t=75&start=180#p4398) showing how the data generated from a single simulation can both appear to "violate" and not violate the CHSH based on our understanding of "strongly objective" vs "weakly objective". I took a local realistic simulation, calculated strongly objective expectations, showed that they satisfied the inequality, and then calculated "weakly objective" expectation values and showed that they appeared to "violate" the inequality. This was from the exact same simulation. Unfortunately, the detractors missed the point of the simulation entirely by focusing on "loopholes" in the simulation, forgetting one crucial point: you can't violate the inequality even with loopholes. In fact, there is only one loophole -- using the wrong inequality, comparing weakly objective expectation values with a strongly objective inequality. This loophole cannot be fixed.

So, for anyone in doubt I would simply advice the following two things:

1. Try to understand why the CHSH, CH, and all such inequalities are Strongly Objective inequalities (cf the dependencies Fred has been talking about), and note the number of degrees of freedom in the inequality
2. Try to understand why the expectation values calculated from QM and experiments are Weakly Objective and note the number of degrees of freedom in the experimental expectations, and the fact that they are different from those in the inequalities.

Apparent "violations" are simply telling us what we put in, "garbage in, garbage out".

[minkwe]

Sure I did. It's CH74. The CH inequality is there derived without requiring identical ensembles of alpha for each experiment

This is untrue, the factorization and algebra of the derivation implies identical ensembles. There is no other mathematical basis for combining probabilities into a single expression and performing factorization operations.

[minkwe]

As concerns Graft's paper, where the heck did equations (1) and (2) come from. They couldn't possibly be from Jaynes' or Bell's papers. He should fix the blatant error in equations (1) and (2). See Jayne's equation (12) and (13) in the article cited by Jaynes.

Secondly, it's quite astonishing how Jaynes' elegant argument has been misrepresented by the likes of Gill (who is quoted in the above article). Jaynes argument has 4 main points:

1. Bell says "It would be very remarkable if b proved to be a causal factor for A, or a for B; i.e., if P(A|aλ) depended on b or P(B|bλ) depended on a. But
according to quantum mechanics, such a dilemma can happen. Moreover, this peculiar long-range influence in question seems to go faster than light"

2. But Jaynes argues, according to QM, P(B|ab) = P(B|b) = 1/2 for all a,b. Therefore, Bell is dead wrong that according to QM P(B|bλ) can depend on a. That is, QM is not consistent with remote parameter dependence.

3. Jaynes also argues that according to QM P(B|Aab) = (1-cos θ)/2. Therefore in trying to complete QM with hidden variables, Bell introduced the schism with QM by assuming that P(B|Aab) = P(B|ab) = P(B|b). Therefore, according to Jaynes, there are many easier and more cogent tests that could be performed to test this experimentally.

Equation (14) is, therefore, the point where Bell introduces a conflict with QM. Recognizing this, it is evident that one could produce any number of experimental tests where the predictions of QM conflict with various predictions of (14). The particular set of inequalities given by Bell is only one example of this, and not even the most cogent one. We shall leave it as an exercise for the reader to show that, at this point, application of Bayesian principles would have yielded a
significance test for (14) that is more powerful than the Bell inequalities

4. Outcome dependence such as the case in QM P(B|Aab) =/= P(B|ab), cannot in general be interpreted as physical causation (see Bernouli's urn example in Jaynes paper).

Many of the discussions in Graft's paper would have benefited from a fuller understanding of Jaynes, than the caricature being spread around by Gill.