Using a Pitcher to Pinch Hit

Starting pitcher Jason Marquis pinch hit for the Washington Nationals in the fourth inning of last night's game and the fifth inning Monday, grounding out both times.

Marquis has a .201 lifetime batting average, but that comes with a paltry .516 OPS. In other words, he's just as weak a hitter as any other pitcher. Yet manager Jim Riggleman often uses starting pitchers to pinch hit if that day's starter is knocked out early in the game, in hopes of saving his bench players to pinch hit later.

If the Nationals' pitcher is being removed so early in the game, it's likely that he's been struggling and that the team is facing a large deficit. Therefore, Riggleman is sending up a pitcher to hit because it is a low-leverage situation (i.e., the at-bat has a small chance of affecting the outcome of the game).

Or so the thinking goes. Yet I have to wonder if the strategy is flawed.

Suppose the Nationals have five bench players, not including the backup catcher (who is rarely sent into the game, because if the other catcher gets injured and the backup is no longer available, the team is in deep trouble). More often than not, Riggleman's strategy would give plate appearances to Marquis and four or fewer of these bench players, when he could no doubt expect better results from plate appearances from five or fewer bench players instead.

The latter combination would give the Nationals a better shot at being in a close game in the late innings, which would justify the risk of using up all the bench players. A run scored in the fourth inning counts just as much as a run scored in the ninth. A surplus of bench players in the ninth inning does the team no good if it's trailing by four runs. Even if Riggleman does blow through his whole bench, it's unclear that having Marquis possibly pinch hit in the ninth is worse than having him for sure pinch hit in the early innings.