Can you actually shorten the game?

Just before struck out in the bottom of the 7th in Monday night’s game, Michael Kay said that it would be important this season for to have success as a 7th-inning reliever because it would turn a game into only a six-inning affair. If you’re anything like me then this is the kind of absent-minded statement about baseball that makes your hair stand up on end. Never mind that teams will play the last three innings of a game, no matter who is on the mound, this statement also suggests that stocking a bullpen with relievers is a viable alternative for a team that doesn’t have a strong rotation. Given how unpredictable relief pitching is, any team that follows this strategy is asking for trouble.

Unfortunately, given the potential strength of the Yankee bullpen this season, fans can expect to be treated to statements of this kind all season long (or at least until the relief unit has another game like Tuesday’s). The fundamental reason it is unwise for a team to bank on being able to shorten a game with its bullpen is because all pitching is volatile and relief pitching is the most volatile of the bunch. Setting this logic as a baseball strategy aside for a moment, however, is it actually feasible mathematically? If the Yankees wind up getting the kind of performances from , , , , and Mariano Rivera that they are banking on, will they get enough innings to actually shorten the game?

According to Baseball-Reference the Yankees tossed 1,442 innings in 2010 and 1,450 innings in 2009. (That actually comes to less than nine innings a game because the away team typically doesn’t toss a full nine frames in road losses). Splitting the difference, one can assume that a team needs about 1,446 innings of pitching in a given season. On the Yankees, is good for roughly 238 of those, or seven innings for each of his 34 starts. That leaves 128 games pitchers other than CC will need to start. If we assume that never reaches for any of the relievers listed above before the seventh inning (a generous assumption, but hear me out) then the two through four starters and long-relief will need to give the team an additional 768 innings. That brings the total innings pitched for the starters and the non-elite relievers to 1,006. Can the cream of the bullpen give the Yankees the remaining 440?

The table below lists the innings pitched in 2010 for each of the relievers on the Yankees.

Those numbers come to a total of 358 innings pitched for the relievers. That would bring the entire Yankee total to 1,364, which is 82 innings shy of what the team would probably need from its staff. That difference is material. None of the pitchers in the table above pitched that many innings last season. To employ this strategy the Yankees would need to get an additional 13.2 innings from each of the relievers the Michael Kays of this world assume will be available to shorten the game. Even if pitches an extra twenty innings this season (which, by the way, might not be a good thing) many of the other names on the list would not be available to make up the difference.

There are two reasons it is folly to suggest that the Yankees have a bullpen so strong that the team can effectively shorten the game. The first is that this is not actually possible. Relief pitchers who usually pitch one inning or less, no matter how good they are, do not traditionally give a team enough length over the course of a season to allow them to be brought in after only six innings on a regular basis. As we’ve seen above, the team will find itself lacking for innings. The second reason it is unwise to employ this strategy? This past Tuesday night.