The Longest Postseason Game Ever, or Not

First, a brief introductory note. Some of you will know me from the writing I’ve done for FanGraphs’ sister site, The Hardball Times. This postseason, I’ve been granted the chance to do some occasional quick-and-dirty playoff analysis here. My thanks to Dave Cameron and Paul Swydan for making this possible. Now I’ll stop wasting time and get to the content.

On Saturday night and (for us Easterners) Sunday morning, the Giants and Nationals played a ridiculously long baseball game. How long was it? For one, Nationals pitchers Tanner Roark aged a year during the two innings he pitched. Yes, that means Sunday was his birthday, and quite possibly the worst one he’s had, since he got stuck with the loss. Happy bleepin’ birthday, Tanner! I’d give you cake, but Brandon Belt already ate it.

Somewhat more seriously answering the question of “how long was it?”, it was 18 innings long. In the past quarter-century of major-league play, there have been just ten regular-season games that went longer. There’s never been a postseason game longer than that.

Or has there? Saying “the longest game” is something like saying “the biggest building.” Do you mean the tallest building? The one with the greatest ground-level footprint? The largest volume? There are several ways to look at the matter.

So I’ll look at the matter several ways. I’ll compute the longest postseason game ever by five criteria, starting with:

Time: There used to be a logjam at the top of this board. Game Four of the 2005 NLDS between the Braves and Astros went five hours, 50 minutes. That is followed closely by Game Five of the 2004 Yankees/Red Sox ALCS at 5:49, Game Five of the 1999 NLCS between the Braves and Mets at 5:46, and Game Three of the 2005 World Series between the White Sox and Astros at 5:41.

There used to be a logjam at the top. The Giants and Nationals went six hours, 23 minutes, starting on Saturday. They beat the old record by more than half an hour. And to think, if Pablo Sandoval had lined out in the top of the ninth, the game would have been over in less than three hours, the only postseason game so far this year to get under that mark.

Innings: Giants/Nats and the 2005 Atlanta/Houston NLDS clincher both went 18 innings, with Game Six between Houston and the Mets in the 1986 NLCS coming in third at 16. I’d be dissatisfied with this tie, except that there’s an easy way to break it.

Outs: The 2005 game ended on Chris Burke’s walk-off home run with one out in the home 18th. Saturday’s game ended on Jayson Werth’s line fly to right, concluding a full 18th inning. That means Giants/Nats went 108 outs, two longer than Braves/Astros. That’s another win for 2014, but I’m not done yet.

Plate Appearances: Looking at innings or outs is actually an imprecise way of tallying a game’s length. It counts the events that have to happen to move a game toward its conclusion, the outs, but doesn’t include all the hits, walks, errors, etc., that produce the score of the game. It’s the bones without the flesh and blood.

Counting how many times somebody comes to bat and does something gets us much closer to the true duration of a game than those other measures. By this gauge, Saturday’s marathon came in at 129 plate appearances, edging out the 1999 “Grand Single” game and Game Three of the 2005 Series by three apiece. But that’s only good for a distant second place. The Braves and Astros in ’05 went a dizzying 144 plate appearances in their 18-inning Game Four.

Pitches: This takes the above concept to its logical conclusion. Pitches are the atoms of baseball games, the smallest irreducible components. More pitches means more actions in the game and, in a strong sense, more of a game. This may be the truest way of measuring a game’s length.

There’s a slight complication to this one, in that our records of pitch counts don’t stretch back that far, even for playoff baseball. The number’s missing for two contenders for the longest postseason game: the 1986 Mets/Astros and the 1999 Braves/Mets. Somebody could watch videotapes of the games and take notes. Nobody has yet done so, or at least no one has done so and then sent Retrosheet the data.

For the games we have fully recorded, once again the Nats and Giants edged out the ’05 ChiSox/Astros contest by three. The Series game had 482 pitches, while Washington and San Francisco combined for 485. And again, this leaves Saturday’s game a distant second, as the 2005 NLDS decider piled up 553 pitches.

So was Saturday’s game the longest in the history of baseball playoffs? It was and it wasn’t. In familiar, even traditional, ways of measuring a game, it was the longest ever. In a couple ways of drilling down into the numbers, it yields pride of place. Ask someone in the Eastern Time Zone who was hoping to turn in at a regular hour that night, though, and you’re likely to find a rock-ribbed traditionalist.

Closing trivia note: Game Three of the 2005 World Series went 14 innings, tied for the longest World Series game ever. The other 14-inning contest in the Series came way back in 1916, Game Two between the Boston Red Sox and the Brooklyn Robins (later the Dodgers). A young southpaw named Babe Ruth went all the way for the winning Sox that day.

The 2005 contest took, as noted above, 5:41 to play 14 frames. Boston and Brooklyn needed 2:32.





A writer for The Hardball Times, Shane has been writing about baseball and science fiction since 1997. His stories have been translated into French, Russian and Japanese, and he was nominated for the 2002 Hugo Award.

17 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Anon21member
9 years ago

All of your links are fucked, man. You did that thing where you start them all with the URL of this piece.

People who write for websites: how does that happen? I’ve seen it many times on completely different sites, and I assume that a bunch of authors aren’t just manually making this mistake, so what gives?

David
9 years ago
Reply to  Anon21

Fangraphs uses a content management system (CMS) to allow contributors to write, editors to edit, etc… without anyone having to know a whole lot about actual web programming. Think wordpress. In fact, it might be wordpress. A CMS usually has some sort of WYSIWYG editor that helps you to enter links to other web pages. When you link to another webpage, that link can be set to be either an absolute or a relative address. Relative means that the URL is assumed to be subordinate to the containing page. Absolute means what you link is what you get. The links in this article are absolute-style links appended relative to this article’s URL. Methinks the wrong checkbox was clicked and all the hyperlinks were set up to be relative when they were supposed to be absolute …. or they were supposed to be relative to the root level fangraphs.com site and not this particular article … or something along those lines

Anon21member
9 years ago
Reply to  David

That’s a great explanation. Thanks, David!

Brandon Reppert
9 years ago
Reply to  David

I can confirm that it is WordPress.

B
9 years ago
Reply to  Anon21

The crazier thing is that you pointed this out 70 minutes ago, and the links are still a mess.

joser
9 years ago
Reply to  B

You think they’re working? There’s a baseball game on!