Best Evidence In Your Opinion?

[Guest FightingFat]

Putting personal testimony to one side, my favorite 'most convincing' evidence is the thermal Woodpile bigfoot. Those curling fingers... I'm still pretty skeptical about most evidence though.

Is that this one:

?

[Guest]

This is a common misunderstanding. If you ask a number of people to invent a story about a tall hairy monster in the forest, the height of the creatures that they imagine will also approximate a bell curve.

Any probability distribution will graph as a curve, but not as a NORMAL curve. In a falsified dataset, the distribution will be sharply peaked, multiple-peaked, or otherwise skewed. That pattern is consistent throughout naturally occurring phenomina. That's why bell curves are such a reliable predictor of validity.

As I stated, in the case of BF, such a falsified dataset would skew to the right (larger), and would more than likely show multiple peaks. Right-skew because it makes no sense for Joe or Bubba Sixpack to hoax a SMALL track or sighting. Bigfoot is supposed to be BIG, which biases the hoaxing towards the large end. The lack of both technical awareness and communication between alleged hoaxers will naturally lead to a number of divergent peaks as various "hoaxers" pick their sizes, not knowing which sizes the others have picked.

You misunderstood my point. I am not accusing anyone of anything, I am just pointing out that the existance of a bell curve distribution in the reported heights of sasquatch is not proof that sasquatch signtings must be real. It also does not require that the sightings are part of a coordinated conspiracy. The bell curve would also arise naturally from a reasonable number of hoaxers, all independently describing a tall hairy creature, without requring any coordination on their parts to ensure the reported heights fit a normal (gaussian) distribution. Again, I am not accusing anyone of anything here at all. I am just pointing out that the normal distribution is not proof that real creatures were observed.

1) The curve was for MEASURED lengths of cast tracks, not decriptions of creature height.

2) Your statement flies in the very face of the foundation of the validity of the bell curve. There is virtually no incentive or reason for a putatitve hoaxer to make up a story about a SMALL bigfoot. That will naturally throw off the curve as previously described.

Basic logic.

^^ Simple. Human and BF tracks do look similar, but there are differences such that one cannot imitate the other, even if they are the same size. BF has that 'mid tarsal break' and it looks different in the tracks- and not human at all. IOW, the more you know about tracking, the more you know that something is out there...

They also are larger across relative to their length than human tracks (which is adaptatively correct for a larger, heavier creature)

How do we know that the 5" track was from a bigfoot and not a human?

IMO track stats are totally worthless unless some unbiased group throws out all the hoaxed tracks and misidentified tracks. The remaining tracks should then be plotted.

Fahrenbach did do a pass through to weed out those he considered unreliable examples according to the standard he set.

And your inclusion of the word "unbiased" of course is meant to allow you to throw out any group whose position you don't agree with, which is a typical Skeptic ploy.

[Guest BFSleuth]

Is that this one ?

Yes, that is the video.

The lack of both technical awareness and communication between alleged hoaxers will naturally lead to a number of divergent peaks as various "hoaxers" pick their sizes, not knowing which sizes the others have picked.

.... unless of course there is a top secret Organization of Bigfoot Hoaxers... ... it would be a very dedicated group...

[Guest DrBanner]

Mulder, why would you expect right skew instead of just shifting the mean of the bell curve to the right? You're not using the word skew properly in a statistical sense. I disagree with your claim that falsified data sets will not produce a normal distribution. There is no reason for that to be true. Independent hoaxers would pick a range of sizes with some "big" mean. And by the way, bimodal ("multiple peak") distributions can occur from real data sets as well, not just falsified ones. By the way, why wouldn't a hoaxer make a small print once in a while? If Bigfoots exist then presumably there are young bigfoots out there, right? Might as well fake those too. Maybe people like you will be even more willing to believe those small prints are real, because you can't imagine why someone would fake a small print. *shrug*

[Guest BFSleuth]

^ Perhaps as an experiment you could ask the question of potential hoaxers, "If you wanted to hoax BF tracks, what length and what width track would you create?" Then compare the results of your experimental question to the data set from Fahrenbach and see how it compares. My hypothesis of such an experiment would be that potential hoaxers would create a data set that is more skewed to the large end of the scale than Fahrenbach's data set, and may ignore the small end of the scale.

"You're not using the word skew properly in a statistical sense." - from my years of statistical analysis in experimental research I think Mulder is using the term "skew" very appropriately.

I think the fundamental issue regarding this question is whether apparently disorganized hoaxers would be expected to generate a bell curve of track sizes in numerous hoaxes over a wide geographic area, and whether they would in many cases undertake such hoaxed trackways in remote locations with little chance the product of their work would be observed.

[Guest]

Is that this one: <WOODPILE VID> ?

yep

[Guest DrBanner]

BFSleuth: Your description of skew is on the money but I'm not sure Mulder was using it the same way. You could perform an experiment as you propose. It would be good to do such a thing before making unsubtantiated claims such as false data sets not leading to a normal distribution. It is easy to invent a hypothetical scenario leading to right skew. If you'd like, I can invent one leading to left skew for you as well. Your reason for expecting right skew is based on the assumption that you know how hoaxers think, their tendencies, behaviour patterns, etc. I'm not sure that is a fair assumption to make.

In fact, disorganized hoaxers might even be more likely to generate a nice bell curve than an organized conspiratorial group of them would be. Humans are notoriously bad at simulating random events. For example, if you ask people to make a fake data set of Heads and Tails coin flips, it is usually quite easy to distinguish the fake data set from a real one based on the absence of long stretches of the same flip (which are expected in a large, real data set, but avoided by people generating a fake data set).

Edited August 12, 2012 by DrBanner

[Guest BFSleuth]

I'm not sure you would be able to call this a random population, but you might consider starting a Poll question new topic and see what kind of responses you get from the BFF on the question of size. You can PM for more information how to do this if you want.

Regarding the question of anticipating how hoaxers think, I suppose I'm taking my lead from kitakaze to "think like a hoaxer". Putting myself in the shoes (pun intended) of a hoaxer, if I were going to go to all the effort of creating a hoaxed set of tracks, I would likely want to create tracks around 18-20" long and about 8" wide at the ball of the foot... which would be slightly to the larger range of the distribution.

Beyond the actual size of tracks, there is of course the question of how your create a trackway that would show tell tale signs of natural movement, with appropriate long stride length, and trying to choose a location that is remote but with some chance that somebody will discover it. The longer and more remote the trackway and the more difficult the terrain or walking conditions the more unlikely it is that the trackway was hoaxed.

[Guest DrBanner]

The only thing I am arguing here is that a bell curve is not unexpected for fake prints, and therefore a bell curve should not be held up as proof that tracks must be real. It's funny that we were debating right skew...haha. I think you'd have to be pretty knowledgable about hoaxer behaviour to confidently predict a skewed curve instead of simply having a curve with a larger mean. Regardless, it's good to see that someone else understands a little bit of statistics here.

[Guest BFSleuth]

^ +1 for a good post, sir! While my hypothesis is that a hoax data set would be skewed, perhaps we can design a test question at least for the population of the BFF to see what we get. I'll PM you, so we can lay this particular issue to rest in this thread and allow further discussion of the purpose of the OP (original post).