What Is The Statistical Probability That All Sightings Are False?

February 14, 2012

Cotter, welcome back,

had to say that, sorry.

So you have probably worked with frequentist probability almost exclusively, as did most of us, until we got into non random situations and unknown prior probabilities.

Long/short: I'm sure there are, though none will ever be able to rigorously attribute the findings to bigfoot, until a bigfoot is proven to exist.

But a background in probability theory is important in determining whether the commonly used frequentist statistics are appropriate.

Here is one incredibly naive example: on one cable television show, it was alleged that the figure shown in the PGF was 7ft -4 inches tall.* They then presented, with the dramatic "hey, look at THIS!" tone, the prevalence of humans of that height, as something like 1 in 100 million, with an implication that this was important to the issue of whether or not it was a man in a suit or a bigfoot. Well, I'm sure you can see why that is an inappropriate use of probability theory. But I'm sure it fooled a lot of people. The DNA study is another area where frequentist statistics have been misapplied, as judged by postings on the internet. Whether that will be corrected in any forthcoming paper is yet to be seen.

People enumerate and do measurements of all sorts of things, that might be amenable to frequentist statistical analysis; they compile statistics, characterize distributions, etc. But with bigfoot, there are problems with trying to use data gathered by a variety of unknown persons in unknown settings with unknown methods, of course, Take the "track measurement" stuff: Did they correctly identify the track? did they measure the inside dimensions at the bottom of the track, or the top? was it in partially melted snow? did they select the one that looked the biggest? was there sliding? These are questions that a single researcher can specify, and attach error "bars", but when the data are collected by a hundred different people...well, you get the idea (and of course there are other problems with that study). Serious researchers don't start collecting data until they are confident of the statistical tests that will be used to analyze it. Sometimes this requires involving a statistician in the planning phases.

of course, it will never be a statistical certainty that bigfoot doesn't exist.

"Zero"? does that mean rounded off? lol in that case, I'd guess that a Bayesian approach would indicate that some time ago we got to a "rounded off zero" probability that "the next reported sighting will result in a proven bigfoot".. But that's not much of a statement, and furthermore, I'm not a statistician.

p.

*later withdrawn.

Edited February 14, 2012 by parnassus

February 15, 2012

Giganto,

the 'balls in the box' is a "thought experiment" and was meant only to introduce the idea of evidential probability, and falsify the idea that "probability theory tells us" that with a lot of reports, some must be true, or that the chances of a true or verified report must increase with each report. If you would like to think of the white balls as unverified reports, and the red balls as verified reports, that might be helpful, but I don't represent that as a rigorous model. One could also view the total number of balls as the number of days that man has lived on this continent, with the white balls being days for which we have no proof of an encounter with a bigfoot (as popularly described.) and the red balls being the number of days for which there exists proof of a bigfoot encounter.

Or the number of vehicle-days driven with/without a roadkill being found ; or the number of episodes of fossil formation with/without a findable bigfoot fossil being formed, or the number of person-days in suitable habitat with/without finding a bigfoot body, etc, etc.

As I tried to imply, a discussion of the theory (s) of evidential (or subjective or Bayesian) probability (and its applications) to the question of bigfoot is not appropriate for this forum. Probability theory is widely misused and misunderstood, and abused. Much in the same way that the word "theory" alone is. In part this is due to the difference in common vs. scientific, mathematical or technical usage of words like "odds", "statistically", etc.

So I have no problem with people saying they are personally 100% certain that they encountered a bigfoot. But that's layman's language...let's not try to imagine that has meaning in the realm of probability theory.

G, If you would like to try to find some rigorous (yet generally understood here) way to apply probablity theory to the issues at hand, by all means, give it a whack. I'll be at the gym.

As a general comment on the numbers of reports:

To me, the number of reports indicate only that there is/are significant reasons why people report bigfoot encounters.

.Some say that one of the reasons why people report bigfoot encounters is that bigfoot exists. At present there is no proof of that. No one that I know of says that there are no other reasons why people report bigfoot encounters. The psychology literature gives many other possible reasons.

The fact that individuals and organizations who collect these reports, cull out large numbers of them as "unreliable", to me further indicates that at least some people (probably most people, and possibly all people) report bigfoot encounters for reasons other than having had an encounter with a flesh and blood bigfoot.

p.

I won't dissect your post, but I should. There's lots to dispute there. In a nutshell you are doing the very thing you are railing against. You claim that the absence of bigfoot roadkill affects the probability that they exist. What exactly is Bayesian about that?

February 15, 2012

As I've said before, it's existance has only been in doubt in modern times since Euro's colonized North America It's recognized in other countries who don't seem to have the problem we do in needing a specimen. In fact given the directive in the 50's from our embassy on a trek to find the Yeti on orders I believe from Nepal not to kill one, or the Army's placing them on a training manuel I'd say your whistling dixie on this one.

unicorns, basilisks, flying horses and (for many people during their youth) flying reindeer were not in doubt before they were demonstrated false. These other countries do not "know" anything about the yeti for instance. They don't know what it eats, how much it sleeps, what diseases it's susceptible to, where they fall on the tree of life etc. How is their position better than a sceptic's? What justification for this idea is better than the null hypothesis?

So the flat earth world view is used to show that root assumptions are the basis of a world view. Science in the 19th century carried a root assumption that black people were less intelligent than whites and followed this with apparent proof by studies of the scull (phrenology ), the nazis continued with this while other countries now changing their world view saw evidence to the contrary. List goes on. There are examples of the ancients and indigenous people and medieval people believing in the hairy man but western science currently doesnt have this fit their world view and discounts it. Likely in future scientists will follow in what they are only begining to understand about the harmony of atoms, quantum physics and the multiverse and declare the great likelihood of BF.

Your examples were all disproved by scientists. Modern science rediscovered atoms because that was true. Science didn't lose this knowledge of the atom, society did. Modern science is not the same thing as the science of Plato or Aristotle. True the ancients did know some things but they got much else wrong. Slavery was widely practiced in most civilizations before modern science came along. Racial views were changed because of scientific study and the propogation of the results not because of an attitude of "oneness with nature" Just so with Bigfoot. If "hairy men" are real then the ancients knew something we have not rediscovered yet. That's all. If "hairy men" aren't real then this would not be much of a surprise either. The ancients were wrong about many other things as well.

No, the point is you are running around making unsupportable statements such as the one below. A lack of proof is not proof of anything.

like invisible pink unicorns? An egregious lack of evidence that should be there (like it is for every other large land mammal) does make it harder to justify the belief in bigfoot. I'm still holding out for Ketchum's report but I'm not placing bets.

February 15, 2012

Google Gregory Panchenko who's getting you your answers. Looks like he and a German Team have been quietly working on exactly what you're asking about. Plus the Russian Gov't has never denied they exist. Plus the body of work by Kaufman in the Caucasus Mt.s.

You're asking for definative proof, my argument was ONLY in North America do we have a bias on Bigfoot, Yeti, Sasquatch existing at all. Like has been done before, despite reports and the indiginous population telling us they exist, we rolled in here and decided it didn't. I would think with all the mistakes we've made in the same vein in the past we'd brighten up.

When a body is finally produced it'll be the whole Mt.Gorilla thing all over again.

February 15, 2012

We have two biases in this country. Some believe and some don't. Making scientific announcements based on either belief is poor science. That's the big thing for most sceptics. I would rally against anyone who tried to base laws on belief alone.

Edited February 15, 2012 by antfoot

February 15, 2012

You're asking for definative proof, my argument was ONLY in North America do we have a bias on Bigfoot, Yeti, Sasquatch existing at all. Like has been done before, despite reports and the indiginous population telling us they exist, we rolled in here and decided it didn't.

Well, how 'bout Nepal? Would you say that belief in yeti is widespread among the indigenous people there? I'd say so, or at least it was.

What did the Nepalese have that indigenous Americans did not? Pieces of yetis. Relics, kept tucked away in their most remote and sacred monasteries. So it would seem that the yeti evidence is much stronger than the bigfoot evidence, right? You've got the ancient cultures absolutely treating these creatures as another kind of organism that they just accept is, and you've got people in those cultures today claiming to have encountered them. You've also got PIECES OF YETIS TO PROVE IT. Here in NA, I've never even seen a claim of a piece of a bigfoot held by native peoples, yet the Nepalese have this evidence and were willing to show it to Western scientists (in this case, Edmund Hillary and Marlin Perkins). They clearly wouldn't have shared it unless they were 100% certain of its authenticity.

There's just one problem: every single piece of evidence that the Nepalese swore up and down was a piece of a yeti turned out to be conclusively non-yeti: The "scalps" were made from tahrs or serows, the skin was from a Himalayan bear, the "finger" was human. So the ancients had nothing, actually, to tie to their yeti except thousands of years of folklore.

February 16, 2012

What is the statistical probability that all sightings are false?

"Slim to none, and Slim just left town."

February 16, 2012

I won't dissect your post, but I should. There's lots to dispute there. In a nutshell you are doing the very thing you are railing against. You claim that the absence of bigfoot roadkill affects the probability that they exist. What exactly is Bayesian about that?

G,

shoulda coulda woulda. nothing substantive.

your "nutshell" nope, not doing that, and didn't claim that. You just made it up.

go back and read it again.

It don't know how I can make it simpler. So if you want to learn about evidential probability, I would advise you to take a course.

p.

Edited February 16, 2012 by parnassus

February 16, 2012

Parn, I served with statisticians, I knew statisticians, statisticians were friends of mine. Parn, you are no statistician.

Edited February 16, 2012 by See-Te-Cah NC
To remove quote of preceeding post

February 16, 2012

"Slim to none, and Slim just left town."

Klaatu Barada Nikto

February 16, 2012

Lively discussion. Nice.

On the old forum, the statistical probability of ALL sightings being false came out to be 0.00034% chance or something like that.

February 16, 2012

COGrizzly, those are stats worth looking into. Out of 1000 sightings, 3 or 4 may be real. Which means there could be 3 or 4 BF's in north America?

I throw that out there for the skeptics to mull over.

February 16, 2012

Hi Kentuckyapeman:

Not to be the math police, but it takes 10,000 sightings at that percentage to yield 3-4.