New clocked EPR Simulation with 100% detection.

Foundations of physics and/or philosophy of physics, and in particular, posts on unresolved or controversial issues

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 5:54 am

minkwe wrote:
minkwe wrote: I would suggest that you make the same assumption that all particles are paired in any real experiment of your choice, including the recent Giustina experiment which claims to have "closed the coincidence loophole". I suggest you use the same data analysis procedure and assumptions you just introduced, and post the results here. It will be revealing.

No response.

My response is that I am not assuming anywhere that all particles are paired in any real experiment. Nor am I assuming they are paired in your simulations, nor in my data-analyses. I am analysing data using "your" algorithm (which is Jan-Ake's, which Jan and I discussed together long ago).
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 5:56 am

Michel, I am still waiting for your response to this one (minor edits by me to make it, hopefully, clearer still):

gill1109 wrote:Let me remind you of the Larsson-Gill claim: it is only possible to get CHSH approx equal to 2 sqrt 2 with a simulation like yours when the coincidence rate is below 87.87%.

For instance: it is not possible to get the coincidence rate to 90% (i.e. a bit above 87.87%) and at the same time CHSH above 2.82 (i.e. pretty close to 2 sqrt 2, or even better)

Both numbers refering to the limit as the experiment gets indefinitely large. There can obviously be chance fluctuations in both numbers in a finite experiment - you do know the 1 over square root of N law, I am sure.

Your own simulation results confirm our predictions.


Why don't you spend some time trying empirically to prove that we are wrong? Tweak your simulation so that it has coincidence rate 90% or better and, at the same time, CHSH equal to 2.8 or better (up to statistical error). Can you do it?
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 7:42 am

gill1109 wrote:Why don't you spend some time trying empirically to prove that we are wrong? Tweak your simulation so that it has coincidence rate 90% or better and, at the same time, CHSH equal to 2.8 or better (up to statistical error). Can you do it?

Why should I spend any time on that, when analytically, logically and mathematically, your paper is clearly wrong. Besides, my ability or inability to do anything does not change the fact that your paper is wrong. No doubt you won't answer my question:

In EPR test experiments, do you deny the fact that the sets of particles used to measure each correlation are disjoint?????
By what magic of mathematics or logic, do you claim that an intersection of 4 disjoint sets is not a null set?????


These are the two simple questions, whose simple answers are devastating to your paper. By the way you have no answers to the following ones either:

<a1b1> + <a2b2′> + <a3′b3′> − <a4′b4> Each paired-product term, has an upper bound of 1 and a lower bound of -1. Which clearly means the upper bound is 4. By what magic of mathematics or logic, do you claim that 4 disjoint sets of particles impose restrictions on each other such that the upper bound is less than 4????
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 8:13 am

minkwe wrote:
gill1109 wrote:Why don't you spend some time trying empirically to prove that we are wrong? Tweak your simulation so that it has coincidence rate 90% or better and, at the same time, CHSH equal to 2.8 or better (up to statistical error). Can you do it?

Why should I spend any time on that, when analytically, logically and mathematically, your paper is clearly wrong. Besides, my ability or inability to do anything does not change the fact that your paper is wrong. No doubt you won't answer my question:

In EPR test experiments, do you deny the fact that the sets of particles used to measure each correlation are disjoint?????
By what magic of mathematics or logic, do you claim that an intersection of 4 disjoint sets is not a null set?????


These are the two simple questions, whose simple answers are devastating to your paper. By the way you have no answers to the following ones either:

<a1b1> + <a2b2′> + <a3′b3′> − <a4′b4> Each paired-product term, has an upper bound of 1 and a lower bound of -1. Which clearly means the upper bound is 4. By what magic of mathematics or logic, do you claim that 4 disjoint sets of particles impose restrictions on each other such that the upper bound is less than 4????


Question 0 "Why should I spend any time on that, when analytically, logically and mathematically, your paper is clearly wrong".

Answer: I don't mind what you do with your time. But I will mention that your prior assumption here is false. Maybe if you might admit to yourself that maybe there are some things which you don't quite understand yet, you might be able to learn some useful new stuff. On the other hand, if you are always certain you're right, you'll not learn much new, ever.


Question 1 a "Do I deny the fact that...":

Answer: No.


Question 1 b: "By what magic, do you claim that an intersection of 4 disjoint sets is not a null set?"

Answer: I do not claim what you say that I claim.


Question 2: "By what magic...":

Answer: Here I did use what apparently appears to you to be magic. ("Any sufficiently advanced technology is indistinguishable from magic", Arthur C. Clarke). Actually, I assumed random sampling of the settings, and I used probability theory, which is part of modern mathematics. Moreover, I only made probabilistic assertions (my assertions involved probabilities and error quantities).

PS. I am beginning to get a bit worried that you maybe never learnt any probability theory, am I right? (There are lots of things I never learnt about, either!). For instance, do you know the Chernoff bound? https://en.wikipedia.org/wiki/Chernoff_bound?

See also http://www.cs.berkeley.edu/~sinclair/cs271/n13.pdf, Lecture notes CS271 Randomness & Computation Fall 2011, Lecture 13: October 6, by Alistair Sinclair.
Last edited by gill1109 on Mon Mar 03, 2014 8:46 am, edited 5 times in total.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 8:25 am

Dear Michel

Can you tweak your simulation so that it has coincidence rate 90% or better and, at the same time, CHSH equal to 2.8 or better? (both statements: up to statistical error. Obviously in a small simulation you might just happen to be lucky. What happens in really big simulations?).

Richard

PS ... and a similar question regarding your earlier simulation model.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 8:46 am

gill1109 wrote:Obviously in a small simulation you might just happen to be lucky.

I suppose you want a simulation as big as the one you yourself are using in your arguments to discredit my simulation? Run for 1 second with ~20000 particle pairs?

Code: Select all
No. of detected particles, non-zero outcomes only
   Alice:           26848
     Bob:           26848
   CHSH: <= 2.0, Sim: 3.228, QM: 2.828
   Coincidence Efficiency:  91.2 %


Your theory says that is impossible, 1 counter-example is enough. You do know the meaning of "UPPER BOUND" don't you. It means it is impossible to violate it even in one case. Luck has nothing to do with it and your paper is clearly wrong. Your ad-hominem attacks won't change that fact.

EDIT: Oh I know what you will say next, that it doesn't match QM. Moving the goal-post yet again.
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 8:51 am

minkwe wrote:
gill1109 wrote:Obviously in a small simulation you might just happen to be lucky.

I suppose you want a simulation as big as the one you yourself are using in your arguments to discredit my simulation? Run for 1 second with ~20000 particle pairs?

Code: Select all
No. of detected particles, non-zero outcomes only
   Alice:           26848
     Bob:           26848
   CHSH: <= 2.0, Sim: 3.228, QM: 2.828
   Coincidence Efficiency:  91.2 %


Your theory says that is impossible, 1 counter-example is enough. You do know the meaning of "UPPER BOUND" don't you. It means it is impossible to violate it even in one case. Luck has nothing to do with it and your paper is clearly wrong. Your ad-hominem attacks won't change that fact.

EDIT: Oh I know what you will say next, that it doesn't match QM. Moving the goal-post yet again.


Splendid! Please tell me exactly the sequence of commands which you gave so that I can reproduce this myself. "Now we have a contradiction, now we can make some progress".

By the way, *one* simulation is not enough, you *might* just have been lucky!
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 8:54 am

gill1109 wrote:Answer: ... Actually, I assumed random sampling of the settings, and I used probability theory, which is part of modern mathematics. Moreover, I only made probabilistic assertions (my assertions involved probabilities and error quantities).

May the record show that Richard Gill has claimed that by randomly sampling "settings", it is possible to cause 4 disjoint sets of "hidden variables" to have a non-null intersection.

You didn't truly expect me not to catch the slight of hand, substituting "settings" where your sets are dealing with "hidden variables" did you? You are right, I did not learn this brand of "Probability theory".

BTW, I have given you more than enough already, I'll follow the advice of some very wise men, and write a paper about this instead, then you will find out how I did it in the paper.
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 8:59 am

minkwe wrote:
gill1109 wrote:Answer: ... Actually, I assumed random sampling of the settings, and I used probability theory, which is part of modern mathematics. Moreover, I only made probabilistic assertions (my assertions involved probabilities and error quantities).

May the record show that Richard Gill has claimed that by randomly sampling "settings", it is possible to cause 4 disjoint sets of "hidden variables" to have a non-null intersection.

You didn't truly expect me not to catch the slight of hand, substituting "settings" where your sets are dealing with "hidden variables" did you? You are right, I did not learn this brand of "Probability theory".


Sorry Michel, you repeatedly do not read what I write.

I did not say that the four disjoint sets thereby had a non-empty intersection.

And why do you say "sleight of hand" when I refer to random selection of settings?

You are assuming that I am a cheat (you may believe it, but you must behave as though you trust my good intentions). I will report this breach of the rules of the forum, immediately.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 9:06 am

minkwe wrote:BTW, I have given you more than enough already, I'll follow the advice of some very wise men, and write a paper about this instead, then you will find out how I did it in the paper.

Now that's a very good idea.

But please, before you do that, tell me exactly the sequence of commands which you gave to generate the numbers you just showed us, so that I can reproduce them myself. If you're right, Jan-Ake and I will have to withdraw our paper. But I suspect a misunderstanding.

"Now we have a contradiction, now we can make some progress".

I would say that this is a win-win situation.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 9:11 am

gill1109 wrote:But please, before you do that, tell me exactly the sequence of commands which you gave to generate the numbers you just showed us, so that I can reproduce them myself.

I simply introduced non-stationarity. That's all I'm going to say and if you understand probability theory as you claim, then you should be worried.
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 9:30 am

Very good Michel, that's exactly how I would do it too! So you do know more about probability theory than you let on.

What about the settings and the times of the setting switches - were they chosen at random *outside* of your program?
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 9:40 am

I take it you still believe that QM violates the CHSH and your Larsson and Gill, "upper bound"? And you believe that, based on using the same QM prediction calculated from distinct experimental arrangements in a single expression?
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 9:47 am

gill1109 wrote:And why do you say "sleight of hand" when I refer to random selection of settings? You are assuming that I am a cheat ...

It is a phrase I borrowed from a certain Englishman, who himself used it to describe the work of a certain I. Pitowsky. I didn't know this englishman actually meant Pitowsky was a cheat!?
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 9:52 am

minkwe wrote:I take it you still believe that QM violates the CHSH and your Larsson and Gill, "upper bound"?


This first question makes sense to me, but is incomplete. Are you talking about QM theory or QM experiment?

So far there has been no loophole free experiment. Some experimenters believe one may be done in the next five years (and they believe it will be successful too).


minkwe wrote:And you believe that, based on using the same QM prediction calculated from distinct experimental arrangements in a single expression?


This second question makes no sense to me.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 9:55 am

minkwe wrote:
gill1109 wrote:And why do you say "sleight of hand" when I refer to random selection of settings? You are assuming that I am a cheat ...

It is a phrase I borrowed from a certain Englishman, who himself used it to describe the work of a certain I. Pitowsky. I didn't know this englishman actually meant Pitowsky was a cheat!?

Cool! You had better remind me the whole context to the phrase of that Englishman. (By the way, where are you from? I believe you studied in Lund, Sweden, but I guess that you or your parents came from Africa. I am very ashamed indeed of what the British have done in Africa over the centuries).

I don't think that Pitowsky was deliberately trying to mislead. But his "theorem" is technically far too advanced for any ordinary person to understand, and the result he got does not actually say what he wants it to say. So I think the phrase "sleight of hand" is appropriate. I did not say that he was deliberately trying to mislead.
Last edited by gill1109 on Mon Mar 03, 2014 10:11 am, edited 1 time in total.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby gill1109 » Mon Mar 03, 2014 10:05 am

Let's take a look at the context of your use of the phrase "sleight of hand":
minkwe wrote:You didn't truly expect me not to catch the slight of hand, substituting "settings" where your sets are dealing with "hidden variables" did you?

You are saying that I deliberately substituted one word for another, intending that you wouldn't notice, with the intention that you would fall into a trap. You accuse me straight out of deliberately trying to trick you. That is discourteous, impolite. It violates the rules of this forum.

I expect you to read carefully what I write, before you reply to anything I write. I expect you to carefully read what you have written, before you hit the "submit" button. Your Python coding is very careful, and very intelligent. So there should not be a problem.
gill1109
Mathematical Statistician
 
Posts: 2049
Joined: Tue Feb 04, 2014 10:39 pm
Location: Leiden

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 10:16 am

gill1109 wrote:Cool! You had better remind me the whole context to the phrase of that Englishman. (By the way, where are you from? I believe you studied in Lund, Sweden, but I guess that you or your parents came from Africa. I am very ashamed indeed of what the British have done in Africa over the centuries).

It matters not a hoot, where I'm from or where I studied, or what the British have done anywhere. This is not a history or personal history forum.

I don't think that Pitowsky was deliberately trying to mislead. But his "theorem" is technically far too advanced for any ordinary person to understand, and the result he got does not actually say what he wants it to say.

I was just pointing out that you chose to selectively misinterpret the use of phrases, contrary to the way you use them yourself in scientific discourse. As concerns the context of your use, let me remind you of it:

http://arxiv.org/pdf/quant-ph/0301059v1.pdf, pg 21:

In the case of Pitowsky (1989), exotic probability does not “explain” at all;
what is called an explanation is sleight-of-hand hidden under impressive
(but very specialistic) mathematics.


I'm also happy that in the same article (talk), you agreed with me that no "loophole"-free experiment will ever be done.

I'll leave that up to others to interpret whether you implied Pitowsky was deliberately trying to hide something. In anycase, my use of the term was in exactly similar context. You can replace Pitowsky (1989) with Gill and obtain my statement almost exactly. Were you being discourteous, impolite to Pitowsky?
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby minkwe » Mon Mar 03, 2014 10:26 am

gill1109 wrote:
minkwe wrote:I take it you still believe that QM violates the CHSH and your Larsson and Gill, "upper bound"?

This first question makes sense to me, but is incomplete. Are you talking about QM theory or QM experiment?

QM is a theory. Do you believe the predictions of QM violate the CHSH or not? If yes, how do you arrive at that violation? What QM calculations lead you to believe that the predictions of QM violate the CHSH?

This is what I'm asking?
minkwe
 
Posts: 1151
Joined: Sat Feb 08, 2014 10:22 am

Re: New clocked EPR Simulation with 100% detection.

Postby Heinera » Mon Mar 03, 2014 11:17 am

Couple of quick questions for Michel:

Are the latest developments of your model still updated in the github sources? (https://github.com/minkwe/epr-clocked)

And, if you are familiar with the papers of De Raedt et al., what is your opinion on your model vs theirs?
Heinera
 
Posts: 788
Joined: Thu Feb 06, 2014 1:50 am

PreviousNext

Return to Sci.Physics.Foundations

Who is online

Users browsing this forum: No registered users and 3 guests

CodeCogs - An Open Source Scientific Library