I forgot to include the fact that there were 20 pulses leading up to
the muon(actually you need to use 21 pulses since one was missed), and
just calculating the probability of having two pulses separated by more
then 320usec is not enough to calculate the probability of a rollover
being missed in the entire buffer up to that point. In that case as
Chris points out:
prob = ( 1 - (1-.03)^21) = (1 - .9997^3) = .6%
So then using our previous 393 event number we would expect 2.4
events(we saw three). Now it should be pointed out that this number
is going to be too high since for one thing we assumed that 25% of the
events passed through one of these boxes. On the other hand the
measured number doesn't include events which happen to have had
radioactivity in the right place by accident.
It is roughly consistent however so I conclude that the rate of data
in this buffer is consistent with having missed one roll-over word.
I also think that Chris is right that by unplugging one input to a
vertical box it probably means that you have something like a 1%
chance of missing a rollover in that WFD channel.
-Chris