Global Consciousness Project

[Zep]

No, that is nonsense, because, again, you are hypothesizing what the hypothetical global consciousness hypothetically is capable of doing or not doing, and all calibration means in the case of RNGs is that the RNG did not fail the tests. That is all. [/B]

The problem is (for the hundredth time) that those TESTS were a load of hogwash in the first place. It certainly seems if they weren't incompetent then they were deliberate misrepresentations. Therefore the "calibration" phase was invalid. Therefore the data gathering phase accumulated meaningless numbers. Therefore the analysis of those numbers was meaningless graphs. Therefore the excited claims were based on those graphs were...bovine excreta.

And the people who trumpeted those claims, knowing all this was indeed BS and who are refusing to respond to the questions about their methodologies??? What do we think of them?

 

[T'ai Chi]

The problem is (for the hundredth time) that those TESTS were a load of hogwash in the first place.

The DIEHARD tests?

 

[CFLarsen]

The DIEHARD tests?

The Orion RNG was tested this way:

Each RNG passes a 256 run random ness test before being shipped. Each run consists of 8192 8-bit samples.

That gives us 2,097,152 bytes.

However, the DIEHARD tests require more:

The minimum length of a tested file is 10 000 000 bytes.

The Orion RNG is not tested for as long as the DIEHARD test requires. Therefore, the Orion RNG could not have passed the DIEHARD test.

 

[jj]

It is a RNG, so we know its mean, its spread; we basically know everything about what it should be producing. These are RNG's that have been used in other applications for a while. No odd results.

What are their frequency analysis characteristics in both short and long-term?

 

[jj]

What are their frequency analysis characteristics in both short and long-term?

And the only sound we hear is the sound of the wind
As it blows through the town, weave and spin, weave and spin.

 

[Zep]

The DIEHARD tests? [/B]

Yes, THOSE tests. The statistical analysis of the DIEHARD tests are fine on a RNG that is NOT being influenced by the testing factor. It's just that in this case they were run on data that was generated with "global conciousness" turned ON. And the whole point of the exercise with the EGGs is that they measure the CHANGES in the RNG output when "global conciousness" is turned ON.

So you now have a completely contradictory situation.

a) The EGGs produced a statistically perfect random bit stream that passed the DIEHARD tests, with GC turned ON.

b) The EGGs then produced a supposed non-random bit stream, with no change in operation and again with GC turned ON.

There are two logical reasons to explain this situation without having to resort to woo-woo ones:

a) The DIEHARD tests are NOT a good test of statistical randomness. I find that difficult to sustain, but others may well know better than I.

b) The "supposed non-random bit stream" was no such thing. The "peaks in the data at critical times" were merely wishful thinking on the part of the people running the EGG experiments based on random fluctuations in the data.

Given the previous history of the people involved and their reluctance to address a lot of issues to do with their objectivity (see above ), I find this a more likely scenario. I gave you a reference to a previous effort of a similar situation, plus my commentary as to why I think that happened. I'm sure you will see there are parallels with this situation.

 

[Paul C. Anagnostopoulos]

jzs said:
These are RNG's that have been used in other applications for a while. No odd results.

No odd results? Why not? Are they somehow isolated from global consciousness?

Wouldn't it be cool if RNGs really generate streams of all 1s, and only generate random streams when influenced by global consciousness?

~~ Paul

 

[rwguinn]

No odd results? Why not? Are they somehow isolated from global consciousness?

Wouldn't it be cool if RNGs really generate streams of all 1s, and only generate random streams when influenced by global consciousness?

~~ Paul

good point
I liken the whole argument to calibrating a flow meter while it is immersed in the flow it is to measure. A fluctuating flow. .

Can't be done.

 

[CFLarsen]

Wouldn't it be cool if RNGs really generate streams of all 1s, and only generate random streams when influenced by global consciousness?

That's all you need to do: Have the eggs spit out all 1s and look for 0s.

In fact, all that is needed is one 0. Kinda like a white fowl of the Corvus family...

 

[CFLarsen]

I liken the whole argument to calibrating a flow meter while it is immersed in the flow it is to measure. A fluctuating flow. .

Good analogy.

 

[jj]

What are their frequency analysis characteristics in both short and long-term?

Well? I'm still waiting. What's the long-term and short-term autocorrelations look like, and on what scales?

 

[T'ai Chi]

What are their frequency analysis characteristics in both short and long-term?

That is extremely vague. Could you be more specific?

 

[CFLarsen]

Bumpety-bump.

 

[T'ai Chi]

The statistical analysis of the DIEHARD tests are fine on a RNG that is NOT being influenced by the testing factor. It's just that in this case they were run on data that was generated with "global conciousness" turned ON.

I think you are trying to confuse things here. No one is talking about "turning off" and "turning on" but you. They look at the RNG output when certain events occur.

a) The EGGs produced a statistically perfect random bit stream that passed the DIEHARD tests, with GC turned ON.

b) The EGGs then produced a supposed non-random bit stream, with no change in operation and again with GC turned ON.

There are two logical reasons to explain this situation without having to resort to woo-woo ones:

a) The DIEHARD tests are NOT a good test of statistical randomness. I find that difficult to sustain, but others may well know better than I.

b) The "supposed non-random bit stream" was no such thing. The "peaks in the data at critical times" were merely wishful thinking on the part of the people running the EGG experiments based on random fluctuations in the data.

I don't agree with a) at all. It could be b), sure. You also leave out a c) possibility, that there may be something going on that has unlikely RNG output occur on days with events.

 

[T'ai Chi]

The Orion RNG is not tested for as long as the DIEHARD test requires. Therefore, the Orion RNG could not have passed the DIEHARD test.

I'm going from the Orion webpage which states:

"The Random Number Generator parallel port dongle(Mac/DOS/Win) is the first true RNG to pass Marsaglia's famous DIEHARD randomness test."

Read what they wrote again:

"Each RNG passes a 256 run random ness test before being shipped. Each run consists of 8192 8-bit samples.

You wrote:

"That gives us 2,097,152 bytes."

That number, times 256.

 

[CFLarsen]

I'm going from the Orion webpage which states:

"The Random Number Generator parallel port dongle(Mac/DOS/Win) is the first true RNG to pass Marsaglia's famous DIEHARD randomness test."

Read what they wrote again:

"Each RNG passes a 256 run random ness test before being shipped. Each run consists of 8192 8-bit samples.

You wrote:

"That gives us 2,097,152 bytes."

That number, times 256.

No, Justin. Please learn the difference between BITS and BYTES.

1 Byte is (usually) 8 bits. With me so far?

Each run consists of 8192 bytes. Still with me?

They do 256 runs. That's (8 x 256) bytes = 2,097,152 bytes. Are you there, Justin?

BYTES, Justin.

The DIEHARD test requires 10,000,000 bytes.

BYTES, Justin.

 

[T'ai Chi]

No, Justin. Please learn the difference between BITS and BYTES.

Yes, I see my error, thanks.

I fail to see how the page you cited says that a minimun of 10-whatever bytes is required to test a RNG. Can you find something like from the creator of the DIEHARD test or something more official?

If you can, you should contact the Orion people and see what they have to say.

Also, on the Orion page under "The test data" I see they didn't specifically mention the DIEHARD tests here.

 

[CFLarsen]

Yes, I see my error, thanks.

You're welcome.

I fail to see how the page you cited says that a minimun of 10-whatever bytes is required to test a RNG. Can you find something like from the creator of the DIEHARD test or something more official?

If you can, you should contact the Orion people and see what they have to say.

Also, on the Orion page under "The test data" I see they didn't specifically mention the DIEHARD tests here.

Justin, what part of "The minimum length of a tested file is 10 000 000 bytes" don't you understand?

That's 10 million bytes. The Orion RNG is tested with 2 million bytes.

You are the one who brings this up. You are the one who argues that the Orion RNG passed the DIEHARD test.

And now, suddenly, you don't understand what it's about?!?

 

[T'ai Chi]

Justin, what part of "The minimum length of a tested file is 10 000 000 bytes" don't you understand?

That is understood, but you are not understanding me. I am asking you for an official site that says that. An official DIEHARD site, for example, or some writing from the creator of the DIEHARD tests. You know, some official verification that that is indeed the case.

That's 10 million bytes. The Orion RNG is tested with 2 million bytes.

Will you contact the Orion people and let them know your findings?

You are the one who brings this up. You are the one who argues that the Orion RNG passed the DIEHARD test.

I don't argue anything. I am just repeating what is on the Orion webpage.

 

[CFLarsen]

That is understood, but you are not understanding me. I am asking you for an official site that says that. An official DIEHARD site, for example, or some writing from the creator of the DIEHARD tests. You know, some official verification that that is indeed the case.

Justin, you bring up the DIEHARD test as an argument that the ORION RNGs are reliable.

Now, you suddenly want to sow doubt about them?

Will you contact the Orion people and let them know your findings?

Sure, no problem.

I don't argue anything. I am just repeating what is on the Orion webpage.

You argue that the RNGs are calibrated. They are not.

You argue that the RNGs produce numbers that pass the DIEHARD test for non-randomness. They do not pass this test.

You argue that calibration means that they pass these tests. It does not.

Justin, how do you think you are doing?