Daniel S. Riley...
>> Interesting point. My source only mentions 100 billion above 4 > snr. But where is this data? Its not in your paper. What is the total > chirp detection amount at or above snr 13 for Livingston in the entire > 01 run?
>See figure 7, p. 20 in
https://arxiv.org/pdf/1602.03844v3.pdf That shows the single-arm trigger rate as a function of the SNR. Unfortunately, it doesn't go >down to 4; a naive extrapolation down to 4 would put the rate at around 100 million, not billion, but that's just naive extrapolation.
Ive been trying to find an html version of this paper unsuccessfully but
just realized its the same paper as the LIGO paper on the website which
is available in html. And no reference to total detection *amount* at snr 13.
And it definitely is 100 billion total greater than snr 4 per detector.
I got that directly from the LIGO template matching team. In fact other info
I have indicates no such specific data is collated regarding total detections
at anything above 4. Because they seem to start excluding any data at any
level which isnt coincident in the 10 ms or so time parameter. That can
exclude potentially millions at snr 13 just because they didnt have a coincident
detection in the other detector within the narrow time parameter dictated by theory.
And as I explain below, you only need in the tens of thousands over 3 months at snr13
in each detector to get at least one random coincidence.
>> Dont forget its not the coincidence rate for chirps above 13 > within the millisecond time parameter thats important. Its the total > amount of similar >>Strength snr detections for each detector that > decides whether or not one can rule out random coincidence. If the > gravitational wave theory were to be >>the only viable explanation, it > would have to be the ONLY chirp detection in the total Livingston > detection run above 13.
>it's very rare that we can completely rule out random coincidence, and there's almost never only one conceivable explanation. Whether an explanation is "viable" depends on your threshold of viability. LIGO does what every science experiment does, they give a likelihood for the alternative explanations they can quantify, and that likelihood is very small. It isn't, IMO, unreasonable to wait for more events (and more detectors) before calling it certain, but GW150914 is >very strong evidence.
>> My guess is its probably still in the millions of chirps above 13 for > each detector.
>I don't think that's possible, given their definition of SNR. A distribution with a rate (as you claim) of 100 billion Hz at SNR 4 and a million Hz at SNR 13 would be very weird--for a normal distribution, we'd expect 5 orders of magnitude just going from 4 to 6 sigma. I can't imagine a physically realizable distribution >with only 5 orders of magnitude reduction from 4 to 13. Anyway, fig. 7 says less than one chirp every million seconds at Livingston has an SNR over 13.
I think the calculations may prove otherwise. To get 1 coincidence in 10 ms over 3 months
one only needs something like 15,000 total detections at 13 snr for each detector. Dont
forget only one gw was at that snr, the second was seven and ten snr. Thats a rough calculation
but its close to fulfilling your supposed order of magnitude restriction. I think the numbers
support 3 random coincidences from two detectors easily. This was no gravitational wave.
Its more likely random coincidence. Especially so considering we know each detector is
detecting 100 billion potential events over 3 months above snr 4.