Steve Fossett and My Baye's Coincidence Theory

[Sephus]

[ QUOTE ]
But what about if they were told that 95% of the land was searched thoroughly?(And that one spot of land is as likely as any other spot to contain the accident) Given that information only, they should multiply the chances of a stunt, whatever they were, by 20.

[/ QUOTE ]

if they thought the chances were 10%, they should expect a stunt 200% of the time.

[David Sklansky]

I have already admitted that I am not intimately aware of the facts of this case. What I AM intimatley aware of is that people who don't understand how to use ratios of probabilities are way off on their assessments when there is significant mathematical information that they haven't taken into account. This includes people who are experts in the specific field.

Perhaps the experts who were asked about Fosset DID include in their calculations the chances of a crash and the chances of a successful search. I doubt it though because they didn't mention it. More likely though they were just similar to the morons who wouldn't change their assesment of guilt when confronted with new information about shoe sizes.

[David Sklansky]

[ QUOTE ]
[ QUOTE ]
But what about if they were told that 95% of the land was searched thoroughly?(And that one spot of land is as likely as any other spot to contain the accident) Given that information only, they should multiply the chances of a stunt, whatever they were, by 20.

[/ QUOTE ]

if they thought the chances were 10%, they should expect a stunt 200% of the time.

[/ QUOTE ]

Obviously the multiplying by 20 method is a good approximation only when the initial probability is quite small. In your example it would move from 1/10 to 20/29. (By the way at least 20 million people in this world can do that calculation in their head in thirty seconds. If you can't, I want to know why.)

[PairTheBoard]

[ QUOTE ]
I have already admitted that I am not intimately aware of the facts of this case.

[/ QUOTE ]

But that doesn't stop you from concluding ...

[ QUOTE ]
More likely though they were just similar to the morons who wouldn't change their assesment of guilt when confronted with new information about shoe sizes.

[/ QUOTE ]

Is that "more likely" than extremely farfetched or just more likely than your chances of getting lucky with Brandy?

PairTheBoard

[Piers]

[ QUOTE ]
Perhaps the experts who were asked about Fosset DID include in their calculations the chances of a crash and the chances of a successful search. I doubt it though because they didn't mention it.

[/ QUOTE ]

How likely would it be for an expert to use the correct Bayesian logic in his calculation but when asked to explain it publicly does not mention the Bayesian logic because he knows it would confuse most of his audience. Instead explaining in simpler (but less accurate) terms that most people will ‘understand’.

Disclamer – Never heard of Fosset before, and never heard any of these experts so probably don’t know what I am talking about.

[Phil153]

Does anyone agree that Sklansky's main conclusion seems valid (that people saying close to zero chance of a fake are morons, if they believe that), but that the rest of his post, including the numbers given, is Sklansky himself being a moron?

[jason1990]

Your original posts are very clear. You said,

[ QUOTE ]
The theory ... has been proclaimed extremely farfetched ... the proclaimers are COMPLETE MORONS.

[/ QUOTE ]
You did not say that those who base their assessment only on psychology are morons. You then said,

[ QUOTE ]
the faked disappearance becomes approximately as likely as the real accident. You read it here first.

[/ QUOTE ]
This is the big claim in your OP. luckyme challenged this part of the OP. He did not challenge the fact that experts did the math wrong. You responded to him by saying,

[ QUOTE ]
You are missing the point. My claim is SELF EVIDENTLY right.

[/ QUOTE ]
I did not misunderstand what you wrote. Perhaps you miswrote what you intended to say. At any rate, now you have clarified your meaning. You are not claiming that your numbers (specifically the lower bound of 9%) have any objective validity. You acknowledge that this is just your personal assessment which could easily be wrong since you are not an expert on the situation. So what is left to discuss? Do we need another whole thread on the fact that "experts" sometimes do not understand Bayes Theorem? I liked it better when it seemed you were making bold claims about specific probabilities. At least then there was something worth debating.