[testing312]

Because it is ridiculously easy to overfit your models, and the combined human behaviour producing 'history' is so complex and ever-changing that it cannot be simulated.

Also, if such a law were proposed, a widespread (dis)belief in that law could undo it.

[totemandtoken]

Oh, I don't believe in a law of history or anything like that. I'm just saying the coupling the practice of history with data analysis is some cases could be fruitful - kind of like anthropological history, but with a bit wider focus