Yeah, that's a step up in accuracy from "statistical analysis". Still, you can technically do ML (badly) without knowing anything at all about statistics. You just open a can of algorithm and a bag of data, dump them in a pot, simmer and stir, and hope it comes out tasty.