When I wrote last week about the slowdown of MLB games in recent years, I noted that the New York Yankees and Boston Red Sox were the two slowest teams. Both are in the American League, and maybe that’s part of the explanation. After all, AL games are played with the designated-hitter rule, which means pitchers don’t have to bat. That allows managers to change pitchers without affecting the lineup, which could prompt them to yank more pitchers mid-inning. Since pitching changes lengthen games — by about two minutes each, according to my analysis — games played in AL parks could be longer simply because of the DH rule.
Reader Gabriel Haro wondered as much:
I checked the data provided to me by ESPN Stats & Info, focusing on 1997 to 2013 (interleague play started in 1997 and 2013 is the last full year in the data set).
My first check was the simplest. I divided all games into four categories: games between AL teams, those between NL teams, interleague games hosted by AL teams (which have the DH) and interleague games hosted by NL teams (which don’t). Games between AL teams were, on average, two minutes and 15 seconds longer than games between NL teams.
But that gap might be because certain AL teams — notably Boston and New York — are slower than NL teams and not because of different rules. Luckily, interleague and World Series games provide a useful test, because teams typically have played each opponent roughly the same number of times at home and away — albeit not necessarily in the same season. That should control for any effect from particular teams or matchups.
So, when the same two teams played each other in an NL park or an AL park, which game was longer?
On average, surprisingly, the longer game has been in the NL park — by 15 seconds. That calculation is based on more than 2,000 games each in NL and AL parks — defining interleague to include regular-season and World Series games between an AL team and an NL team.
That’s solid evidence that it’s the style of individual teams, rather than the DH rule, causing the discrepancy in length of AL and NL games. And, in fact, if I isolate the 1997-2013 data set to just games without Boston or New York, then AL-only games are faster, on average, by a minute and five seconds relative to NL-only games. (Interleague games without the Red Sox and Yankees continue to be roughly the same length with or without the DH rule — one second longer without the DH.)
But maybe that’s unfair — of course AL games will look faster once I’ve removed the league’s two slowest teams. So I tried two more tests.
First, I removed the Red Sox and Yankees, but also the two AL teams with the fastest games, on average: the Blue Jays and White Sox. AL-only games without the league’s four biggest outliers averaged two hours, 55 minutes and 5 seconds. All NL-only games averaged two hours, 55 minutes and 2 seconds — just three seconds faster. Meanwhile, interleague games without those four AL teams were 18 seconds faster, on average, in NL parks. So, again, there’s no evidence that the DH lengthens games.
Finally, I instead removed the NL’s two teams that played the longest games, to complement my removal of the AL’s two slowest teams. From 1997 to 2013, those two NL teams were the Mets and the Dodgers. Without them, and without the Yankees and Red Sox, the average NL-only game was nine seconds slower than the average AL-only game. And the average interleague game in an NL park took 13 more seconds than the average interleague game hosted by an AL team.
So while the DH theory made sense, I’m confident the DH doesn’t lengthen baseball games.