Steve Fossett and My Baye's Coincidence Theory

[David Sklansky]

I had no public audience for my coincidence theory (explained in Poker Gaming and Life and endorsed, he tells me, by Persi Diaconis) when I suspected freezing temperatures caused the shuttle disaster, despite expert's views to the contrary.

This time I have a chance to go out on a limb and publically say that I think it is quite likely, if not actually probable, that Steve Fossett faked his disapperance. The theory has been looked into and has been proclaimed extremely farfetched. But again the proclaimers are COMPLETE MORONS. They could easily turn out to be right as far as what the ultimate outcome is. But they are definitely wrong as far as what the price is (unless I am mistaken about the facts of the case.) Because they don't compare one theory's probability to the others.

Steve Fossett may have had no personal or financial problems to give him a reason to fake his death. But he did have an extreme obsession to be in the limelight. That alone brings the price down. When you combine it with the fact that he was an expert aviator, the fact that its a million to one shot that any random plane goes down, and the fact that a massive search hasn't found him, you have a situation where the faked disappearance becomes approximately as likely as the real accident.

You read it here first.

[luckyme]

Need to watch that this isn't one of those -

He honestly crashed - "sure, but it was a good probability it was faked, even though it's honest this time."
He faked it . - "See, I told you."

What does it take in how this is presented to elevate it above, "My leg hurts, there's a good chance of a storm".

Seriously, how do we test/demonstrate this in a one case scenario?
( I agree with the increased likelyhood of a Fosset fake over a luckyme fake).

luckyme

[David Sklansky]

[ QUOTE ]
Need to watch that this isn't one of those -

He honestly crashed - "sure, but it was a good probability it was faked, even though it's honest this time."
He faked it . - "See, I told you."

What does it take in how this is presented to elevate it above, "My leg hurts, there's a good chance of a storm".

Seriously, how do we test/demonstrate this in a one case scenario?
( I agree with the increased likelyhood of a Fosset fake over a luckyme fake).

luckyme

[/ QUOTE ]

But the experts are saying his faked death is highly unlikely. The easiest way for me to prove that I am not just trying to get a freeroll is to say that I will take a 10K to 1K bet. Push if nothing is discovered either way in a year.

[luckyme]

[ QUOTE ]
[ QUOTE ]
Need to watch that this isn't one of those -

He honestly crashed - "sure, but it was a good probability it was faked, even though it's honest this time."
He faked it . - "See, I told you."

What does it take in how this is presented to elevate it above, "My leg hurts, there's a good chance of a storm".

Seriously, how do we test/demonstrate this in a one case scenario?
( I agree with the increased likelyhood of a Fosset fake over a luckyme fake).

luckyme

[/ QUOTE ]

But the experts are saying his faked death is highly unlikely. The easiest way for me to prove that I am not just trying to get a freeroll is to say that I will take a 10K to 1K bet. Push if nothing is discovered either way in a year.

[/ QUOTE ]

That's the easy way to show that you are confident in your claim. How can we show that your claim is deserving of that confidence?
My granna just offered 5-1 on the basis of her ouija board . Does that add validity to her confidence?

Again, how do we get to that validation stage?

luckyme

[David Sklansky]

You are missing the point. My claim is SELF EVIDENTLY right. Those who say otherwise are doing so because they don't understand how to use probability ratios. You agreed that it was more likely that Steve Fossett would fake his death than others, but that was only part of the reason for my contention. The more important reason is the equally farfetchedness of the other alteranative. Given the safety of modern airplanes and the lack of success in finding him despite a small area and a massive search.

The bottom line is that mathematically illiterate people when assessing the probability of a theory, often do it in a vacuum without assessing the probability of the alternative theories. This occurs most especially when the method to assess one probability uses different techniques (eg the psychological makeup of someone) than what would be used for the second one (eg card combinations).

In other words if you are almost positive that someone won't move all in without the nuts, you still must change your price when going from ace to five to Badugi.

[Phil153]

[ QUOTE ]
The more important reason is the equally farfetchedness of the other alteranative. Given the safety of modern airplanes and the lack of success in finding him despite a small area and a massive search.

[/ QUOTE ]
So for any small plane that goes down (and isn't found after a week), the one in a million odds of the plane going down, plus the fact that they're not found, make faked disappearance more likely? Or does Fossett's enjoyment of the spotlight tip it over the line?

What are the relevant values that you're plugging into Baye's Theorem? For example, how much does the one in a million chance of plane going down impact on the probability? Should we automatically think someone is faking when their plane disappears? Or something like 30%?

Given that you're only taking 10 to 1 odds, does this imply that you think the odds of you being correct are less than 20%?

[PairTheBoard]

Canadian Press Story

Starting Price Unknown.

Quotes from link:
Raising the Price -
" "I have no idea about Mr. Fossett, but I know that it has happened in the past where we have had guys just disappear and stage things," Washoe County District Attorney Richard Gammick said. "

Raising the Price -
"No trace has been found of his single-engine plane despite a small air force that has scoured the canyons and hillsides along the Sierra Nevada's eastern front for 11 days, raising the prospect that he's just not there."

Lowering the Price -
"Another possibility is that Fossett strayed much farther afield than the search area, which already covers 44,000 square kilometres. The plane he was flying could have taken him deep into neighbouring California, Oregon or Arizona, all states with vast areas of wilderness."

Lowering the Price -
" "With his notoriety, we believe he couldn't walk away from this type of event," he said. "People would recognize him." "

Lowering the Price -
" High winds kept most search planes grounded Thursday. "

Were there high winds the day he was flying? If so, would he have risked flying anyway - he was a daredevil type guy after all. So Sklansky's remark below is not reliable:

[ QUOTE ]
DS -
the fact that its a million to one shot that any random plane goes down

[/ QUOTE ]

This was not a random plane in a random area. It was a plane piloted by a daredevil in an area that recently has had winds so high as to keep search planes grounded.

A wild guess -
[ QUOTE ]
DS -
the faked disappearance becomes approximately as likely as the real accident.


[/ QUOTE ]

A no-lose guess by DS. If the wreck is found it will not prove DS's probability estimate wrong and like predictions made by psychics his guess will be quickly forgotten. However, if DS gets lucky and the guy did stage it, DS will crow.

Another case of Opinion all dressed up in fancy verbage about logic and probability. A commonplace at SMP in general and for DS in particular.

PairTheBoard

[Phil153]

Your hard on for Sklansky really amuses me. It means he's doing his job.

[PLOlover]

does the same logic apply with jfk jr death regarding possible foul play?

[Siegmund]

I am happy to agree that Steve Fossett is more likely to pull a complicated stunt than the average pilot is.

I am not so happy to agree with DS's opinions of how rare small-plane crashes are, especially of experimental aircraft, even if they have experienced pilots, or of how easy he thinks it is to find plane wrecks. (The number of previously unknown wrecks they found during this search is pretty good evidence of THAT!)

I could be persuaded that the chance of Fossett being alive is anywhere between say .01% and 2%. I think 9% is much too high, and would be happy to give 10:1 that he won't turn up alive within the next year. I don't have 10k sitting around, but I'd happily bet $100 to get $10.

If, that is, I either owe $100 or receive $10 in a year's time - I wouldn't go putting the $100 in a non-interest-bearing account (like a sportsbook's coffers) for a whole year while I wait.