Your comment selcome on the folliwng draft "letter to the editor"
JOHN BIBBY
=========
"Intelligence" is never 100% true and reliable, and military intelligence
may be less true than most.
What then should we do about it?
The answers may be elementary - each statement should have an associated
'confidence level' ranging say from 10% to 90%. These would indicate betting
probabilities and would order the intelligence statements from less to
greater certainty. The percentages could also themselves be 'tested', as in
the long run 90% of the statements with 90% certainty would eventually be
supported by the facts, while only 10% of the statements with 10% certainty
would turn out to be true.
JOHN BIBBY (York, England)
[log in to unmask]
|