Analytics should support experience and knowledge in the field, not replace it.

/ begin rant
Please allow me a tangential vent. Colin Cowherd mocks analytics (often mere hours after offering them in support of his position, but save that for later) and nerds for being stat obsessed. He even has a nerd voice for the routine. But what completes the circle jerk is that when he is discussing Hall of Fame cases, he starts by decrying stats nerds, claims he knows a Hall of Famer when he sees it, and then begins to rattle off total career numbers (All-Pro, Pro Bowl, HRs, doubles, rebounds, etc.).

One day you will here about a sports radio listener who has wrapped his car around a tree because instead of driving, he was yelling at Cowherd in the car. That will be me, sometime between 9AM and 1PM or so. I actively try to avoid him, but n some places in the State, you can't get any other sports station.
/end rant

Back to the topic at hand, analytics offer support and context, occasionally clarity where there was fog before. M3 didn't have clue one why his team was always more injured until he hired a sports science and nutrition guy who changed the way the players were fed, hydrated and altered the schedule to something unheard of in M3's experience. If you want to solve a problem that has resisted your solutions for a long period of time, you have to look for information and expertise elsewhere. This is something analytics and data can help with.

They also help you correct normal human biases and help place context around events. Thompson doesn't need help identifying who can play, but data and analytics might help him project better where the player should be drafted (possibly even where he likely will be drafted) and might help identify the blindspots.

I don't want RoboTed to be GM, but I want RoboTed sitting next to mostly human Ted.