[extropy-chat] calling all bayesians

spike spike66 at comcast.net
Thu May 12 05:20:25 UTC 2005


> ...confirms what I did conclude: that there is not enough
> information in the problem as stated to calculate an
> answer.  I turned it around: by assuming a droob reliability,
> what is the probability that the destructive testing of
> the 19 spare units would result in all-good?  That
> method demonstrates the circular reasoning inherent
> in the problem.
> 
> spike
> 

This is not to say there is *no* value in destructively testing
the 19 spares.  If two or three bads show up in the 19, then
I know there is some serious problem with my initial reliability
assumption.  If it show no bads, (as I expect it will) then that 
test tells me very little.

spike








More information about the extropy-chat mailing list