Not just to evert team, but by in game situations. The games are played on a field not on a spreadsheet. The analytics should be a tool, not the final decision. I want my coach to know what the analytics say, but also to know their limitations. If the chart says go for 2, but it's a close call, and your go to 2 point conversion play leans on your right guard winning a battle, but your right guard has been getting beat all night, may want to consider kicking. I think I heard Fleck say once that a decision to go for it or punt came down to how amped up and confident the players were in the huddle when he asked if they should go for it.
People say things like "if it works the coach is a genius, if he doesn't, he's any idiot" as though that is an unfair way to evaluate coaching decisions. On the contrary, that is why coaches get the big bucks. The decision needs to factor in countless considerations, most of which fans don't have insight into. Nuances of data, what the data does and doesn't factor in, personnel, game conditions, how each individual player is doing. At the end of the day, the coach needs to make a call on whether, in light of that, they think they can get it. Seems obvious, but going for 2 never pays off it you don't get it, and it's always worth it if you do.
