That is true, the QBR is a black box
But with analytics, there are lots of places you can see the math. There is a real-time bot on Twitter that says what their recommendation is. And all the math is different --- slightly.
Also, when I look at it I get the logic ... if you go and miss, percentages tell you that you will gain more points by going for it than kicking FG or punting ... but I have looked at the math and it is way over my head. Not by a little ...
Mina Kimes is one of the biggest proponents ... and I like her fine, but she's so committed she does not like to budge when challenged. With the comment, Nick Saban made about going for it---he said roughly that analytics don't tell you about what happens if you go and fail.
She rightly pointed out that is the one thing that it does. It takes into consideration all possible outcomes if you go for it and fail. It is baked into the cake. So in that specific case, Saban was wrong. But I think well, he's a legend coach, so he pretty much knows what he's doing
The problem for me is these folks don't like it when coaches don't follow the "advice". They will mock the coaches, there is even a bot that rates every punt on a "Coward index". That was the "93rd percentile of cowardly punts since 1999". That is overboard ... not sure them calling a coach's decision cowardly is more of a projection than anything.