I have a thingo up at Citymetric on how agent-based modelling would have helped the poor sods at Govia Thameslink Railway realise how stuffed they were, at least in time to everyone from getting fired.
There is a bit of dispute* in the extremely nerdy parts of the Internet about whether I’m being reasonable in my decision to equate agent-based modelling with Bayesianism and not-bothering-to-do-agent-based-modelling with frequentism.
I would argue that although they aren’t literally necessary, frequentists who are fervent enough to argue about this point are all the kind of idiots who do their models in Excel and so deserve everything they get.
*the people who disagree with me are right and I am wrong, in absolute terms