Upvote:2
I would use the assumption that the age ranges were approximately normally distributed, and then work on determining:
The mean of the distribution; and
The standard deviation of the distribution.
and how both might vary by maritime speciality - fishing and particularly whaling, for instance, perhaps attracting a younger demographic due both to the need for greater strength and being more dangerous
Why a normal distribution you ask?
Because the Central Limit Theorem states that if we can make the assumption that the arithmetic mean of the ages on individual ships are statistically independent, then the distribution of those means across the industry (and its specialities) is normally distributed even if the age distributions on each ship are not.
The Central Limit Theorem further states that, if the distribution is actually Binomial, then due to large N the Normal Distribution is a suitable and accurate substitution.
There is no reason to believe or expect that a specialized distribution, such as the Poisson or Pareto to take two common possibilities, would be applicable.
A reasonable sized sampling of crew lists from ships of the period would then allow one to estimate the mean and standard deviation of the whole population from that of the sample.