Experiment

Paul had an intreasting idear. Take the matrix of distances we have of people against each other.

We have 50 odd people so we get 50 * 50 matrix then * 12 for 12 experiments.

This would give a set of distances most likely normally distrbuted.

So IF we take he distances of one person ( 1 * 50 * 12 ) then perhaps we can do a Z test or something like it and then get a measure of how random that person is. So if everyone puts A and B together then we should find A,B or all of A's distances are not a random sample of the distances between people.