The save stat rewards the wrong pitching decisions and manages bullpens badly

A save gets credited to a relief pitcher who finishes a win while protecting a lead of three runs or fewer, or enters with the tying run somewhere in the mix. That definition, written decades ago, has quietly warped how managers actually deploy their best reliever, almost every team saves its closer specifically for ninth-inning save situations regardless of when the game is actually on the line. The problem is obvious once you watch it happen repeatedly: the highest-leverage moment in a game is frequently the seventh or eighth inning with the bases loaded and a one-run lead, not a clean ninth inning with nobody on base. A manager chasing save totals for a closer's next contract, or just following decades of bullpen-management convention, will burn a lesser reliever in that actual high-leverage spot and save the best arm for a lower-pressure ninth that happens to fit the save definition. The stat was supposed to measure who's closing out games effectively. What it's actually done is talk managers into saving their best pitcher for the least dangerous version of a close game. Sports Illustrated's MLB coverage has covered the handful of teams that have experimented with leverage-based bullpen usage instead.