On Tuesday night, I tweeted* that it seemed the Red Sox/Marlins game had been on for "like six hours now." Always a fan of hyperbole, it was more a statement to the fact that Mark Buehrle was pitching in a game that was now over the three-hour mark. While most understood the point I was making, @rRaindog63 (also known as Bill outside of cyberspace, a quick Googlin' tells me) felt the need to challenge by hyperbolized comment with one of his own:
@CeeAngi In part, you can thank the D.H. rule for that. And once they add Instant Replay to every other play, the games will get even slower— Bill Miller (@Raindog63) June 13, 2012
I feel like I know Bill well after reading his 140 character self-description of himself. Bill is a father, a Mets fan, a history teacher turned librarian, and a writer. I'm not sure which library he works at, so I can't confirm he's actually a librarian, nor have I read anything he's ever written—but let's assume all of these things are true, including the Mets fan part. Small sample size, but Bill is wearing a Mets hat in his Twitter avatar, so let's assume that Bill does, in fact, watch the Mets and by proxy more National League baseball (Bill, if you're listening, since you're in Greenville, South Carolina, you should go check out the Drive, the Class A affiliate of the Boston Red Sox and their ballpark that is a replica of Fenway Park).
First, I was willing to overlook the fact that Bill made this comment during the Red Sox vs. Marlins game, an interleague match-up that was played at Marlins Park, meaning there was no designated hitter,but, Bill's comment had me wondering: was there any validity to this idea that American League games are longer and that the designated hitter was the culprit? I knew what Bill was getting at, he even sent me a link to this article, which is from 2009 and basically argues the same thing that Bill did: that more offense is slowing down the games.
I asked around and Tweeters seemed to agree that the more hits, the more walks, the longer the games would be—and that clearly games in the American League would be longer, and according to one friendly tweeter, more boring.
But, I wasn't convinced.
I didn't think it was rational to assume that just because there was one batter different in the lineup, that the designated hitter made a big difference in game length than having a pitcher in that slot would, especially when you consider moves like double switches, pitching substitutions, and pinch hitters, which are used frequently in National League games to compensate for the lack of designated hitters.
I took to the Internet to find more data or articles that had been written in regards to game lengths of the American League versus the National League, but I didn't find much other than articles insisting that the designated hitter is evil, the American League is inferior, and that all of this offensive production was making game so much longer, but I still wasn't convinced. Since I couldn't find a neatly packaged version of point I was trying to argue, I did the only thing I could: I decided to create the argument myself. After a couple of emails to Sean Forman of Baseball-reference.com and Bradley Ankrom of Baseball Prospectus, I had all of the game length data I was seeking.
I looked at several things in the data, which is available if anyone wants to see the full-spreadsheet versions, but I had three main data sets: Average Game Length (AL vs. NL from 1960-2012) which includes extra-inning games, Average Game Length (AL vs. NL from 1970-2012) which includes only nine-inning games in the averages, and Average Time per Out from 1960-2012. After looking through the data and making a few charts, here's a summary of the findings:
Average Game Length 1960-1972 (pre-designated hitter)
Total Seasons: 13
American League Longer: 8 seasons
National League Longer: 5 seasons
Smallest Difference: 1 minute, in 1960, 1963, 1966, 1970
Biggest Difference: 8 minutes ( National League 163 minutes, American League 155 minutes)
Average Difference: 2.6 minutes
While American League games lasted longer in 62% of the seasons from 1960-1973, most season game lengths were close—with the average difference in game length being just 2.6 minutes per game. In two seasons, 1961 and 1962, the National League games were higher than average, lasting 5 and 8 minutes longer, respectively. The time per out data mirrors the game length data in this case, showing that in 8 seasons the American League took longer to record an out, the games were also longer.
What does this tell us? Not too much, other than entering the era of the designated hitter, game lengths and times per out were comparable between the two leagues. Now, enter the designated hitter.
Average Game Length 1973-2012
Total Seasons: 40
American League Longer: 34
National League Longer: 3
Smallest Difference: 0 minutes, 2002, 2004, 2011
Largest Difference: 11 minutes ( American League 1985, 1994, 1995)
Average Difference: 4.95 minutes
Over 40 seasons with the designated hitter, the evidence suggests that games in the American League are, in fact, longer but with one major caveat—they aren't that much longer. With an average difference of just 4.95 minutes per game, it would seem that the perception that American League games are just so much longer! And the designated hitter is to blame! seems to be a misconception. Furthermore, there are so many variables that go into determining the length of a game, that it's difficult to pin that to the designated hitter, when there are many factors that can dictate the length of a game in either league, including but not limited to:
• Commercial breaks
• In-park promotions
• Slow pitchers
• Slow batters
• Coaches who visit the mound frequently
• Bat boys that lack hustle
• Catchers that approach the plate frequently
• Pitching changes
• Bench clearing brawls
• Singing Sweet Caroline*
• Number of pitchers per at-bat (walks, plate discipline)
• Park factors
*there is no evidence to suggest that singing Sweet Caroline makes games longer since it's conducted during a commercial break, but it should be on the list as it's an evil much greater than the designated hitter.
If you look at the chart, the biggest spike in game differences came during the period between 1985-2001, where American League games were clearly taking much longer, but since 2002, that gap has shrunk considerably. Since 2002, the National League games have been longer in three seasons, with the American League longer in five seasons (two were ties). Again, there are many factors that can contribute to the game length, but the data suggests there is an increase in game time for the National League, not a decrease for the American League, which leads me to believe that the factors dictating game length are not closely tied to the use of the designated hitter, but rather a combination of many factors that can dictate the length of the game.
Bottom line: @Raindog63, or if I may call you Bill, you're right—the games in the American League have been historically longer since the switch to the designated hitter—but I'm more inclined to call it a draw on this argument, considering the difference in game times is minimal, with both the American League and National League game lengths in a tie for the 2012 season so far, continuing the trend of the game length gap being eliminated between the two leagues.
At any rate, Bill, thanks for the discussion.
*Tweet is such a stupid word, and I cringe every time I use it. Does anyone else get really uncomfortable actually talking about Twitter, even with people who are on Twitter? All of the words that they choose to use, like Tweet and ReTweet are silly, but when you start getting into things like "I'm following that guy" it sounds just two steps removed from boiling a rabbit in a stranger's kitchen. Not to mention the fact that we get disappointed when someone "isn't following us back," because clearly unrequited stalking is lonely. If someone who doesn't use Twitter asks me what I'm doing on my phone, I'd rather tell them I'm posting on the Miley Cyrus message boards than admitting that I just... tweeted.