For decades, virtually all baseball teams have been slavishly following the same pitching strategy: The pitching staff is composed of five starters, half a dozen or more relievers—some long, some short—and a closer. The starters are left in the game to pitch as many innings as humanly possible. They are allowed to pitch to the point of failure.
Any seasoned baseball fan can recognize what this failure looks like. The starter does not have "command" of his pitches. His fastball isn't fast. His "stuff" isn't working. He's tired. He's wild. His pitch count is high. The results, too, are obvious. Hitters are hitting, walking, getting on base in other ways and, sooner or later, scoring.
In the dugout, as this tableau unfolds, managers and coaches become restive. They squirm, they confer, they consult the lineup to see who's available in the bullpen, in case a pitching change is unavoidable. Visits to the mound are made. The problem is solemnly diagnosed and discussed. The pitcher says he’s fine. He stays in. He either overcomes his problems or he doesn’t.
But the truth is that no one—not the pitcher, the pitching coach, the manager—knows whether the problem will be overcome before it leads to a tragic end. When, and if, the pitcher is in extremis, he is at last removed, but only after his collapse has caused damage: runs scored, runners on base bequeathed to his successor, and, all too often, injury to the pitcher.
Baseball is big business, as all owners of major league teams know. Payrolls of large-market teams have swollen to exceed more than $200 million a year. Players—even those expected initially to play in the minor leagues—are routinely paid millions of dollars a year. Multi-year contracts with superstars dole out tens of millions annually. Profit in major league baseball is dependent on winning, and winning is dependent on pitching. Remarkably, though, baseball’s pitching system has long been broken. There’s a way to fix it—and, not incidentally, improve the bottom line.
But first, why is this pitch-to-failure system uniformly employed? At bottom, the system is based on faith. Those in the dugout are pros. They have been in baseball their whole lives, often as former players. They can diagnose and cure the problem, supposedly; it is a “trust-me” system. The box scores, however, don’t inspire trust. With rare exceptions, they usually report the collapse of at least one of the two starters. Rarely do both starters battle for an entire game, never fail, and either win or lose by a small number of runs. In the great majority of games, at least one starter fails.
Once again, the truth: No one can dependably recognize failure in advance of the event. Which is why I propose a reliable way of knowing when to remove a starter before he fails: Every starter must be relieved no later than the end of the fourth inning. The rule would be inflexible: no discretion, no exceptions; the starter is out of the game after pitching four innings. Before the number four is reached, the manager has the same discretion he now exercises. He can remove the starter earlier for all the usual reasons.
The number four is both arbitrary and useful. It is arbitrary because you can’t empirically show that collapse occurs in the fifth inning. It is useful because most starters do not collapse in the first four innings. The game is usually not irretrievably lost by the end of the fourth inning. (And in any case, while four innings may seem like a random choice, the current system employing five starters—each of whom must rest for four days—is itself somewhat random. Where is the evidence that a system employing three starters, each resting two days, would not be better?)
The objections, of course, are obvious. You’re taking out successful pitchers too soon, preventing them from pitching a great game, even a complete game. How can a starter notch wins if he won’t be allowed to pitch the requisite five innings? Records and memorable feats—including the shutout, the no-hitter, and the perfect game—would become meaningless or impossible to achieve. But the new system has one goal: loss prevention. It isn’t aimed at boosting pitchers’ records or managers’ records, or any result except preventing failure on the mound resulting in a loss. The mantra is: save arms, win games.