Yeah, jzs didn't suggest I buy one.jzs said:
In any case, I don't see how getting my own RNG would help. I'd prefer to test the actual bitstream that is being used to produce the GCP results. Since that apparently is not possible, then that's it.
Global Consciousness Project
- Thread starter paintedbird
- Start date
-
- Tags
- consciousness global project
In any case, I don't see how getting my own RNG would help. I'd prefer to test the actual bitstream that is being used to produce the GCP results. Since that apparently is not possible, then that's it.
Pragmatist
Graduate Poster
- Joined
- May 12, 2004
- Messages
- 1,529
jzs said:
I was asking if that was what you were suggesting. If it wasn't, fine.
jzs said:
But as Donks has already pointed out - unless I misunderstood something - the GCP didn't keep the relevant data. So he's not going to get actual data. Since it is the hypothesis of GCP that the output of the RNG is anomalous under specific past, unique conditions, it doesn't help at all to look at the output of a similar RNG now, because that wouldn't clarify anything. The similar RNG may or may not be anomalous, but that wouldn't necessarily indicate in any way whether the original RNG was anomalous or not, would it? Therefore, your suggestion appears to be an exercise in futility - unless you meant something else. In which case please make yourself clearer.
Pragmatist said:
Yeah, the email said that they "do not record the bitstream." And I agree with your view. My original intent was to see if the data, the part that shows anomalies according to their analysis, can pass the DIEHARD tests or not.
Pragmatist
Graduate Poster
- Joined
- May 12, 2004
- Messages
- 1,529
Donks said:
And that's what I thought you meant, thanks.
The issue here is real simple - and it appears obvious to all but Justin. But Justin appears to have trouble understanding real simple things...
The output of an RNG is either "random" or it isn't. If it isn't, it shouldn't rightly be called a "Random Number Generator".
Sampling the output of an RNG for a short period of time (even if it passes DIEHARD tests) does not establish whether the output is always "random" in perpetuity. Unless the output is continually sampled and recorded at all times, there is no baseline for comparison. It is meaningless to compare the actual output at some limited point in time with a generic ideal of "randomness" - because true randomness can be "random overall", without (apparently) being "random in particular" over a short period.
The hypothesis of GCP is that external events constitute "anomalous" output. But without the full record there is no way to establish whether they really are anomalous or not (compared to the rest of the bit stream - which is the only meaningful comparison possible). The fact that a small sample window may appear anomalous compared with some statistical ideal of "randomness", does not imply that the overall output was ever anything but random. If such a comparison (between a small data window and statistical ideal) was valid, then a run of say 10 heads in a thousand coin flips which gives overall 50/50 statistics would imply that the coin wasn't fair.
There is reason to suspect that the RNG's are not necessarily truly random over a long period anyway. That reason is because the manufacturer says so: http://www.randomnumbergenerator.nl/rng/home.html
To put that in context - the output may be guaranteed to be within certain parameters at the time of shipping, but there is no guarantee of long term stability in actual use - particularly not given that electronic components tend to change their parameters with age and use.
The RNG is supposed to be "shielded". Shielded against what exactly? No shielding is 100% effective for EM, let alone gamma radiation, cosmic rays, neutrinos etc. And what generates the "randomness" in the RNG? A Zener diode generating noise due to quantum effects - a quantum source that by definition has to be susceptible to just about all of the above. It makes little difference whether you use two Zeners or one or ten.
But the hypothesis of GCP is that they are not shielded against GC. And even the manufacturers believe that the "psi phenomenon" may be a problem:
But of course, in the case of GCP, no, "no subject condition" is possible by definition. If the device is not shielded against GC then there is no control condition to compare with and the data is meaningless - and the device can never be shielded against all possible extraneous influences anyway. If the device is shielded against GC, then the "results" have nothing to do with GC, and in the context of GC, the data is meaningless. The only reasonable conclusion overall, regardless of whether GC exists or not is that the data is meaningless!
Pragmatist
Graduate Poster
- Joined
- May 12, 2004
- Messages
- 1,529
CFLarsen said:
Indeed, he most certainly appears to be!
Pragmatist said:
Well thats a tautology, I agree.
I'm asking if the RNG output is statistically different than what one would expect by chance. I'm not asking if the RNG is random.
They make their data available.
Show me their exact claim from that page that supports your claim that they say "RNG's are not necessarily truly random over a long period anyway".
That's why you look at the data they produce, Claus Lite, to see if they are out of wack or not.
What the GCP people do for a 'control' is compare the data on a day that is in the formal hypothesis registry to a neighboring day that is not in the formal hypothesis registry. For the day of the formal hypothesis there should be something going on. For the control day, nothing should be going on.
That is lame copout, even for you. Nothing in any field can ever be shielded against "all possible" influences. Therefore, one can always say 'it wasn't shielded against X!', thus attempting to debunk it. But one can say this about any thing in any area of study, so who cares.
Well, that might be the conclusion that you want to perpetuate, but it is certainly not the "only reasonable conclusion". How about examining the data? That seems reasonable.
So if there is no GC, you are saying that RNG data is still meaningless? Gee, try telling that to the statistical industries that thrive on using data generated from RNG and PRNGs.
I tried. They do not record the raw data. They can't currently make it available. It can't currently be tested. And unless someone else kept a record of the eggs' raw data for Sept. 11, etc., it is no longer possible to confirm those results.jzs said:
Donks said:
Why would they need to keep recording the 0's and 1's? They already know the Orion, for example, has passed the DIEHARD. There is just no reason to.
They do make raw count data available, and I guess we should be thankful they make any data easily available (unlike JREF, for example) over the internet. I downloaded a very small chunk of one day's worht of data and it was over 10 megs!
You could buy a RNG, right? Maybe a team of skeptics could all pitch in?
Because that's the output from the RNGs, and that's what's supposedly being affected.jzs said:
What does the JREF have to do with anything?
And yes, I suppose an RNG can produce quite a bit of data in a day.
On a student's salary? Ha! Either I buy one or I eat
In any case, it wouldn't matter. I wanted to check their results, and to do that I'd like to use the input they use, not my own.
Donks said:
Well, from their page on RNG's
"The final output of the physical REG unit is a sequence of bytes presented to the computer's serial port, which are then formed by the acquisition software into a sequence of trials (sums of 200 bits), generated at 1 per second."
so there may not be a way to just get the 1's and 0's if the acquisition software is summing them automatically. I don't know.
On a student's salary? Ha! Either I buy one or I eat
Which is why I suggested getting several people to split the cost.
Using the same RNG would be suitable I'd imagine. Or are you suggesting you want to use the exact same RNGs taht the GCP people and volunteers are using? Like you want to borrow one of theirs?
If you keep this up, Justin, you will become simply another Kumar on this forum. Really. You now have at least four people have told you the same thing about those RNGs in a number of different ways, and yet you want to persist in arguing for them. Plus we now know that the original source data, from which all these amazing PEAR claims are derived, is not available. So your whole line of argument now looks a lot like this...
The software the Orion people offer doesn't seem to have this limitation. If GCP has it, it's because they programmed it that way.jzs said:
No, me having one of theirs accomplishes nothing. I'd like to use the data they used, pass it through DIEHARD. I don't care to test a different RNG. I want to see if DIEHARD detects an anomaly in the same period they detect an anomaly.
That's great Zep, you figured out how to post images.
Anyway, we see that
1) people assume the GC exists, but only for the purpose of saying the data is worthless
2) no one has actually analyzed their data, because from 1) they say it is worthless
3) the maker of the Orion says it passed the DIEHARD test, but people will still claim the RNG has not passed it
4) Claud asked for the number of times it was done in a graph I posted (I couldn't find it exactly for that graph and told him so) so he could comment on my question. I graphed some GCP data and told him it was done 10,000 times. Still, no substantial comments from him. He avoids the main question entirely.
Are these skeptical approaches?
Anyway, we see that
1) people assume the GC exists, but only for the purpose of saying the data is worthless
2) no one has actually analyzed their data, because from 1) they say it is worthless
3) the maker of the Orion says it passed the DIEHARD test, but people will still claim the RNG has not passed it
4) Claud asked for the number of times it was done in a graph I posted (I couldn't find it exactly for that graph and told him so) so he could comment on my question. I graphed some GCP data and told him it was done 10,000 times. Still, no substantial comments from him. He avoids the main question entirely.
Are these skeptical approaches?
rwguinn
Penultimate Amazing
AH-I see...jzs said:
Ignorance is curable. Your condition, however....
Pragmatist
Graduate Poster
- Joined
- May 12, 2004
- Messages
- 1,529
jzs said:
You know perfectly well what I meant. Assume "random" to mean "statistically what one would expect by chance". The RNG manufacturers and the GCP use the word "random" in exactly that sense.
jzs said:
Apparently not. They only make a small pre-massaged, cherry-picked subset of their data available - there's a difference you know.
jzs said:
I did. I'm afraid I am unable to help you with your reading comprehension problem though, sorry.
And here's a related quote from the GCP web site:
See also "Rotten Eggs"
jzs said:
Since you've now resorted to name calling, I guess you won't mind if I call you "L'ie Che'at" - it seems appropriate somehow, don't you think?
And believe it or not, I agree with you! You are absolutely right! All we need to do is look at the long term output of the actual RNG to see whether there are variations. So where is that data? Oh, I forgot, they discarded it...oops...
For what it's worth, if I was expected to analyse such data I wouldn't just be running simple stats on it. I'd be running Fourier analyses on it to see if there were persistent periodic patterns, and autocorrelations across various windows as well as crosscorrelations with redundant generators (with different degrees of shielding) during the same time period - and in particular I'd be looking for aliasing across different windows. And I'd be doing similar analyses of the power line state and other measurable variables like the background radiation flux etc. And of course the DIEHARD as well (I don't know whether the DIEHARD tests include any of these analyses).
GCP mention that Fourier analyses can be done - but they don't say that they are done routinely. And I see no evidence that data from potential influence sources is acquired at all.
jzs said:
Yeah, nothing should be going on. I guess they just "know" somehow what everyone on earth just happens to be doing every day. A lot of major events don't get reported. And where are the external influence measurements? Have they bothered to try correlating with solar activity for example? I note in their FAQ they simply claim their equipment is "shielded" and external factors don't count - yeah, right!
jzs said:
Crap. You know perfectly well what I mean. I gave you examples. Power line conditions. Ambient radio activity. Solar EM and other astrophysical major EM sources. Atomic tests and EMP. Gamma and Cosmic Flux. Neutrino flux, things like that. And your argument is a total straw man in any event. Nearly everything can be potentially influenced, but it's a matter of degree since in this case they are essentially measuring quantum level activity it is vastly more susceptible to miniscule influences than the majority of experiments in most areas. And, here is the killer for your argument, if they had actually bothered to monitor a wide range of ambiental variables and record them along with the data then it would give a much higher basis for confidence in rejecting alternative correlations with external events other than "global consciousness". Why do they just write off external influences and assume "global consciousness" as the primary hypothesis? Are you seriously going to argue that is objectve science?
Where does the Orion egg get its power from BTW? It takes power from the RS232 port of the computer it's attached to. Any halfway decent engineer will tell you that the RS232 lines vary enormously from computer to computer - I forget the allowable voltage variation but it's something of the order of +/- 3 volts to 18 volts! And anyone who thinks that a PC power supply (which feeds those lines) is somehow perfectly stable and clean would have to be a complete moron. And no, there is no reasonable way of regulating and filtering that particular source to make it extremely clean, there will always be residual noise that feeds all the way down to the Zener under test. But that noise isn't just "random" noise. It's often synced to things like the circuit clocks, switching transients etc. The fact that they can't even be bothered to design a proper, stable power source for the eggs says a lot about the whole methodology and integrity of the project.
jzs said:
I repeat. What data? The pre-selected, filtered, massaged data they actually kept is useless. Doubly so in the absence of correlate data for possible influence sources.
jzs said:
Some would consider it a sad indictment of society that a "statistical industry" could even exist!
And in any event it's a straw man. The data is useless for the purposes of determining whether there is "global consciousness" or not. There is insufficient data to allow proper examination of key alternate correlates, as I have repeatedly said. In short, the experiment is badly designed. The data is useless for the purpose for which it was allegedly obtained.If you really, honestly believe that there is some useful data and significance in this project then feel free to actually address the issues. I suspect you won't - as usual.