I've been reading The Bill James Guide to Baseball Managers this week—it's $6 on Kindle after 15 years or so out of print—and because it's less about evaluating managers or players than observing them the refrain he returns to over and over, across 150 years of baseball history, is that great managers don't just have a type of player in mind—they have an understanding of those types that is nuanced enough to get the best performance possible out of them.
It's stronger before the front-office era, when managers were responsible for the scouting, composition, and discipline of their entire roster, but the book picks up a throughline that connects Connie Mack and John McGraw's comprehensive player development philosophies and contemporary managers' preference for certain kinds of players and bench structures.
Daniel Descalso has played most of his defensive innings over the last two years at shortstop. This year Daniel Descalso has taken 19 of his 50 plate appearances in high-leverage situations—17 with the game late and close, 14 as a pinch hitter. Yesterday he batted in the bottom of the 12th inning, as the final piece in a comprehensive double-switching regimen that made the Cardinals worse on offense and defense and allowed no single pitcher to go more than one inning.
I think sometimes it's too easy to look at the omnibus stats, WAR etc., and determine that Mike Matheny is foolish for keeping somebody like Daniel Descalso on the roster, but in this case it's too easy because defaulting to overall value is not hard enough on what Mike Matheny is doing.
Mike Matheny, like Tony La Russa, like Connie Mack, like Harry Wright, will have his guys. Every time he'll take a utility infielder who looks like he's working hard over a talented malingerer or an unproven rookie. That isn't the problem. The problem is that over and over this year he has set his guys up to fail.