Sci.Physics.Foundations

FrediFizzx wrote: ↑Fri Dec 03, 2021 2:44 am

FrediFizzx wrote: ↑Thu Dec 02, 2021 5:41 pm

jreed wrote: ↑Thu Dec 02, 2021 2:19 pm I have looked at your latest version, and here are my comments:

You did a great job of getting things to fit the cosine curve. The fitting worked really well.
There are now two detectors for each observer. They generate two records for each experiment for each observer.
The detection loophole is still there. This is easy to see by adding the line of code:
Total[nPP + nPN + nNP + nNN]
after the statistical analysis. This will display how many records were saved out of the total of twice the number of trials.
I get about 50%. The rest were discarded.
The discarded records were those where 0.001 was added to the derived sign value, giving 1.001 or -0.999.
When these go through the statistical analysis, the program drops them since 1.001 is not equal to 1, and -0.999 is not equal to -1.

Well, that would be a partial thanks, but here is an update for you. Took out the 0.001's as they aren't needed after all.

Cloud File

https://www.wolframcloud.com/obj/fredif ... c-forum.nb

Direct files.

download/CS-40-S3quat-3D-new-prodcalc-forum.pdf
download/CS-40-S3quat-3D-new-prodcalc-forum.nb

Enjoy!

Here is a 5 million trial plot to go with the update.



Baby is still screamin' -a.b!
.

jreed wrote: ↑Fri Dec 03, 2021 12:27 pm The detection loophole is still alive and well.

FrediFizzx wrote: ↑Fri Dec 03, 2021 12:37 pm

jreed wrote: ↑Fri Dec 03, 2021 12:27 pm The detection loophole is still alive and well.

Nope! Total Events = 4993634. So, we are only missing just a few events out of 5 million trials. Not enough for any detection loophole.

jreed wrote: ↑Sat Dec 04, 2021 12:53 pm

FrediFizzx wrote: ↑Fri Dec 03, 2021 12:37 pm

jreed wrote: ↑Fri Dec 03, 2021 12:27 pm The detection loophole is still alive and well.

Nope! Total Events = 4993634. So, we are only missing just a few events out of 5 million trials. Not enough for any detection loophole.

Let's discuss how your program works. You set up an array of random angles on a unit sphere (s), then two random variables, lambda1 and lambda2, then go into a do loop for the samples generated by Alice and Bob. For Alice, a detector angle is found, and two comparisons are made with the quaternion product of that detector angle and the random angle s. These comparisons are with lambda1 and lambda2. For each comparison, if the absolute value of the real part of the quaternion product is greater then lambda, this is output as the detector angle, followed by the detector value, (+ or - 1). If the value is less than lambda, the detector angle, followed by a quaternion is output. This is done for both values of lambda, lambda1 and lambda2. That gives twice the number of sample points the program starts with. If you do 1 million tests, you get 2 million samples. This is also done for Bob's part.

These samples then go to the statistical analysis part of the program. Samples with +/- 1 are saved, but samples with quaternions are discarded. The program typically discards about 50% of the samples. This is why it appears that the program is analyzing all the samples, when actually it is only using about 50%. This is the origin of the detection loophole.

jreed wrote: ↑Sat Dec 04, 2021 12:53 pm Let's discuss how your program works. You set up an array of random angles on a unit sphere (s), then two random variables, lambda1 and lambda2, then go into a do loop for the samples generated by Alice and Bob. For Alice, a detector angle is found, and two comparisons are made with the quaternion product of that detector angle and the random angle s. These comparisons are with lambda1 and lambda2. For each comparison, if the absolute value of the real part of the quaternion product is greater then lambda, this is output as the detector angle, followed by the detector value, (+ or - 1). If the value is less than lambda, the detector angle, followed by a quaternion is output. This is done for both values of lambda, lambda1 and lambda2. That gives twice the number of sample points the program starts with. If you do 1 million tests, you get 2 million samples. This is also done for Bob's part.

These samples then go to the statistical analysis part of the program. Samples with +/- 1 are saved, but samples with quaternions are discarded. The program typically discards about 50% of the samples. This is why it appears that the program is analyzing all the samples, when actually it is only using about 50%. This is the origin of the detection loophole.

gill1109 wrote: ↑Sat Dec 04, 2021 9:54 pm
Two lambdas to make it more complicated and in order to disguise data rejection.

Joy Christian wrote: ↑Sun Dec 05, 2021 2:07 am

gill1109 wrote: ↑Sat Dec 04, 2021 9:54 pm
Two lambdas to make it more complicated and in order to disguise data rejection.

There is no data rejection in Fred's code.

Bell's detectors only detect plus or minus results: A = +/-1, B = +/-1. Bell's detectors do not detect quaternions. Anyone who claims that quaternions must be considered data does not know what Bell's theorem is all about. If they claim that they are "experts" in Bell's theorem, then they are kidding themselves.

Look for a flaw in the code other than data rejection. If you are truly an expert in Bell's theorem, then you should be able to find one --- a genuine one.

jreed wrote: ↑Sun Dec 05, 2021 2:13 pm If you want to see the effects of data rejection, remove or comment lambda1 and lambda2 in the first do-loop. Then replace these with:

Code: Select all

lambda1[[i]]=lambda2[[i]]=0. 

This will do away with data rejection. You will now get all events in the calculation. The output plot will be a perfect triangle. No samples will be discarded. No artifacts baby.

FrediFizzx wrote: ↑Sun Dec 05, 2021 2:17 pm

jreed wrote: ↑Sun Dec 05, 2021 2:13 pm If you want to see the effects of data rejection, remove or comment lambda1 and lambda2 in the first do-loop. Then replace these with:

Code: Select all

lambda1[[i]]=lambda2[[i]]=0. 

This will do away with data rejection. You will now get all events in the calculation. The output plot will be a perfect triangle. No samples will be discarded. No artifacts baby.

Triple LOL! STRAWMAN! What did I tell you about changing the code? Well, I will tell you again since it hasn't seemed to sunk into your fanatic brain yet. If you change the code, it is not the same model and is in fact a strawman model!

.

jreed wrote: ↑Mon Dec 06, 2021 6:42 am I thought you might be interested in trying this experiment yourself. I assumed you want to understand how the cosine curve is generated. Maybe you aren't.

jreed wrote: ↑Mon Dec 06, 2021 9:09 am If you add the Total statement in that run of the program, you would find all samples are used. What do you conclude from that?

jreed wrote: ↑Mon Dec 06, 2021 4:06 pm After the statistical analysis of events loop, add the following statement in both your normal calculation and lambda = 0 calculation:

Total[nPP + nPN + nNP + nNN]

this will show how many samples were used in each calculation. In the example with lambdas not equal zero, you''ll see about 50% samples used and the cosine curve calculated. In the example with lambdas equal zero you'll see 100% samples used and the triangle curve calculated. What do you conclude from that?

FrediFizzx wrote: ↑Mon Dec 06, 2021 5:23 pm

jreed wrote: ↑Mon Dec 06, 2021 4:06 pm After the statistical analysis of events loop, add the following statement in both your normal calculation and lambda = 0 calculation:

Total[nPP + nPN + nNP + nNN]

this will show how many samples were used in each calculation. In the example with lambdas not equal zero, you''ll see about 50% samples used and the cosine curve calculated. In the example with lambdas equal zero you'll see 100% samples used and the triangle curve calculated. What do you conclude from that?

Absolutely nothing! As I said before, we don't count mathematical artifacts as valid detections. The valid detections for 5 million trials, Total Events Detected = 4 998 238. You really should admit that Bell's theory is junk physics.

gill1109 wrote: ↑Sat Dec 04, 2021 9:54 pm

jreed wrote: ↑Sat Dec 04, 2021 12:53 pm Let's discuss how your program works. You set up an array of random angles on a unit sphere (s), then two random variables, lambda1 and lambda2, then go into a do loop for the samples generated by Alice and Bob. For Alice, a detector angle is found, and two comparisons are made with the quaternion product of that detector angle and the random angle s. These comparisons are with lambda1 and lambda2. For each comparison, if the absolute value of the real part of the quaternion product is greater then lambda, this is output as the detector angle, followed by the detector value, (+ or - 1). If the value is less than lambda, the detector angle, followed by a quaternion is output. This is done for both values of lambda, lambda1 and lambda2. That gives twice the number of sample points the program starts with. If you do 1 million tests, you get 2 million samples. This is also done for Bob's part.

These samples then go to the statistical analysis part of the program. Samples with +/- 1 are saved, but samples with quaternions are discarded. The program typically discards about 50% of the samples. This is why it appears that the program is analyzing all the samples, when actually it is only using about 50%. This is the origin of the detection loophole.

Thanks, John! This is very helpful.

The program contains both elements of Joy’s original ideas and of Fred’s ideas with the three adjustable parameters. Two lambdas to make it more complicated and in order to disguise data rejection.

The last plot of the deviations shows that it does not exactly reproduce the negative cosine. I’m looking forward to see how Joy and Fred will try to sell this as physics.

The idea of duplicating each particle or particle pair has turned up many times before. Bryan Sanctuary did that. @Esail (Eugen Muchowski) does it too.