In article ,
"Peter Duniho" wrote:

wrote in message
...
[...]
Of course your odds of having an engine failure with two engines is
double of what it would be with one, and quadruple with four.

Only approximately. The only reason doubling (or quadrupling) the number of
engines doubles (or quadruples) the chance of an engine failure
(approximately) is that the failure rate is so low. For example, if the
failure rate were 50%, a doubling of that would cause you to expect an
engine to fail each flight (a 100% chance of failure), when in fact the
chance is actually only 75%.

It is somewhat ironic that you should be the one to point this out in
light of the argument we are having in another branch of this thread
because this is precisely the point I was making. The condition of the
probability of failure on a single trial P being low is precisely the
condition that allows you to approximate the formula for cumulative
failure 1-(1-P)^N as P*N. If you think about it, there is absolutely no
difference in the risk calculation between making one flight with four
engines and four flights with one engine (except insofar as the
probability of failure for one engine over four flights are not quite
independent of each other if it's the same engine each time).

rg