Sorry, another reply.
I'm familiar w this statistical analysis bc, overseas, we had many--smart people all--working to analyze the highest kinetic threats we should go after.
Generally speaking, they used Complex Adaptive Systems as a base model, w statistical overlays, including page rank, though admittedly it was not fully comprehensible to me, and my particular group, bc we were just were dumb-ass mouth-breathers.
I fully realize that the NFL, with recognizable rules, is not the same as modeling bad guys w (often) stochastic behaviors.
But the results, w geniuses and every NTM available, and $billions, were... not quite as we had hoped.
So, I guess, given the much stricter set of variables in the present NFL case, what are the identifiable weaknesses in these calculations?
I only ask to understand how to assess this modelling w greater precision.