TED: Peter donnelly shows how stats fool juries

Especially starting from 11th minute... reminds me about the need to review the measures of the model: specificity, sensitivity... more here: https://en.wikipedia.org/wiki/Sensitivity_and_specificity