I’ll ask again. If you don’t believe stats should be used to help set lineups, what do you propose, choosing names out of a hat?
I’ve never had a problem accepting that almost all coaches know who their key hitters are, but I will never accept that without some kind of data, they can accurately assess much more than that.
IOW, there’s no trick to identifying Joe, Tom, and Bill as the key players. But the trick is, what do you do beyond that? Its very unlikely that all 3 will have the same skills at the same level, so its unlikely they’ll be completely interchangeable. All the data does is allow an accurate comparison in order to give the best chance of putting each player in the spot that best suits him.
Even if a coach doesn’t keep track of data with a piece of paper and pencil, he’s doing it somehow. Memory is the most usual, and that’s what shapes perceptions. All I’m saying is, there’s so many individual data points to try to remember for even 1 game let alone 5, 10, or more, memories can and do easily get “muddled” . What’s the dear of having evidence of one’s perceptions?