I compile our multi-type flying-hours for all of our pilots on 206(R) Sqn at Boscombe Down and I've learned that statistical analysis for groups of pilots will change (enormously) dependant upon whether you include (or not) those pilots who have been posted, or, that you include those who have recently arrived (and that I've updated my database for the previous twelve months flying-hours of those pilot's who have recently arrived). Statistical figures are very helpful; except that there is NO substitute for anything other than the "actual" figures for each pilot. Averages, minimums, and maximums, don't reveal the true nature of any individual's flying activity without going into specific detail for each pilot.
The bottom line is... you can prove anything with statistical analysis; just as long as you understand the baseline of those numbers!
TCF