But if one looks at a much more robust set of data based on a much wider variety of countries, one gets both a model with much better fit, and with a very different message.
Dr Farley believes that if fed data on the number of members of a network who have been apprehended, an ordered-set model can give the authorities an estimate of the probability that the network has been disrupted.
Once a model is created, data scientists ideally will begin working with others to help test the models and set them in motion toward commercial application.