hal@rain.org (Hal Finney) writes:
>Peter C. McCluskey, <pcm@rahul.net>, writes:
>> Here's a scenario which I would guess has a 0.1 to 10% chance of happening:
>
>I wonder if you could describe your thinking in choosing to express this
>as a range of probabilities. Is it, as Robin and Nick have suggested, that
>you could theoretically pin it down more closely if you thought about it
>longer? But based on how much time you have spent thinking about it, all
>you can say now is that eventually you would have settled on a probability
>in this range?
Approximately. I'm expressing a good deal of uncertainty about what the
correct number is, although it isn't obvious whether more thought would
improve on that, or whether more data is needed.
>Can you express in probabilitistic terms how likely it is that the
>probability is in the range from .1% to 10%? More, could you estimate
>how likely it is that the probability is in the range from say 2% to 6%?
>And even more generally, could you give the shape of your belief function
>in terms of a probability distribution over the possible probabilities?
Yes, these are possible, but would take a fair amount of thought.
-- ------------------------------------------------------------------------ Peter McCluskey | Critmail (http://crit.org/critmail.html): http://www.rahul.net/pcm | Accept nothing less to archive your mailing listReceived on Tue Apr 28 21:10:51 1998
This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:30 PST