Monte-Carlo Modelling

Monte-Carlo Modelling is a technique invented for predicting the behaviour of particles inside atomic bombs. It is named after the casinos of Monte-Carlo, because it uses probability theory in random samples: if you play long enough, you can discover the probability of winning without having to calculate it predictively . It works like this:

The PeopleSize program generates large sets of numbers, each unit of which is random, but which collectively are 'normally' distributed, in the same way that human dimensions are distributed. A Normal Distribution is a set of numbers (such as human dimensions) in which most values are near the average, and large and small values are progressively rarer.

In a normally-distributed set of random numbers each number is random, but the set of numbers as a whole represents human size distribution in some important ways.

The program generates the distribution, then examines it to determine the proportion of data points which meet all the dimension percentile criteria. In this way it answers the question hypothetically, but on a very large sample.

The extent to which each pair of dimensions vary independently of each other is central to this process, and is expressed as a 'Correlation Coefficient'. Correlation coefficients are calculated statistically in anthropometry datasets, as matrices of coefficients. They link a majority of the dimensions available in PeopleSize, but not all.