Playoff Games Tend to Take a Very Long Time
Less than a week ago, in Game 5 of the NLDS, the Cubs and the Nationals played one of the more strange nine-inning games that I can remember. I don’t need to go back over the details; Nationals fans don’t want to revisit them, and Cubs fans currently have more pressing matters on the mind. But in the end, Wade Davis struck out Bryce Harper to wrap up the bottom of the ninth. The game was never dull, as the Cubs escaped by only one run. Yet it became the longest nine-inning game in baseball’s postseason history. The first pitch was thrown at 8:08pm local time. The last one was thrown 277 minutes later.
Clearly, no conclusion can be reached based just on one extreme. It’s not like nine innings of playoff baseball always take four and a half hours. But, you might have a gut feeling that playoff games have been taking a while. They have been! Because they almost always do. For a variety of reasons, when it comes to planning your day around a playoff baseball game, you should carve out a bigger chunk of free time. By roughly half an hour or so.
This is really quite simple, thanks to the powerful search tools available at Baseball Reference. In the plot below, I’ve covered the wild-card era, stretching back to 1995. I calculated average game duration — in minutes — per nine innings, and you see lines for the regular season and the playoffs. Easy enough to understand. For the sticklers among you: I realize it’s sometimes considered a sin to not begin the y-axis at zero, but, come on. Let’s not worry about that today.
In every year, the average playoff game has taken longer than the average regular-season game. Again, per nine innings. The smallest gap between the two averages was 21 minutes, in 1998 and 2001. The largest gap between the two averages was 46 minutes, in 2009. Over the whole window, regular-season games have averaged 173 minutes, while playoff games have averaged 203 minutes. Over the past five years, those averages are 180 and 210 minutes, respectively. This year alone, we’re presently looking at 185 and 224 minutes. The latter is the largest average in the sample. Regular-season games are taking longer than ever. So it’s not a complete shock that this has carried over to the postseason.
Let it be said that Jerry Crasnick wrote about this years ago. This is obviously not a new phenomenon, and if you want an explanation, there’s an easy place to start: the TV breaks are longer. It’s not just about place of play on the field. MLB and the networks are out to make as much money as they can, by running as many advertisements as they can without pissing too many people off. Excerpt:
Between-innings commercial breaks account for a significant chunk of the longer running times in October. The standard break for an MLB regular-season game is 2 minutes, 5 seconds, and extends to 2:25 for a nationally televised game. During the postseason, the between-innings break increases to 2:55.
Even if you didn’t know that, you could probably feel it. The TV breaks drag, but, remember that live sports are one of the only ways for networks to make money anymore. You can forgive them for insisting on a greater opportunity. So, longer breaks. That’s part of it. What else? You can’t really blame runaway offense.
There tend to be fewer runs in the playoffs, because the pitching is extremely good, and, historically, run-scoring has been depressed by about 10 – 11%. In one sense, more baserunners and more run-scoring would lead to longer games, because games don’t end until there have been sufficient outs. But because playoff runs are at a premium, games tend to be closer, which makes every pitch more important. That’s likely to slow down the average pace. A game like yesterday’s will pick up, when one team is beating the other by eight.
Also related to pace, here’s a proxy. A major reason why playoff run-scoring goes down is because playoff contact goes down. There are more strikeouts in October, because, once more, the pitchers are selected for their effectiveness, which is strongly related to whiffs.
So far in these playoffs, batters have struck out more than a quarter of the time. That would stand as a postseason high. Strikeouts mean deeper counts. Indeed, the average playoff plate appearance lasts more pitches than the average regular-season plate appearance. One of the things I’ve come to learn about pitcher pace over my years of being interested in it is that pitchers tend to slow down as a count gets deeper and deeper. Some of it might be accumulating fatigue, and some of it might be having to think a little longer, since a batter has seen more pitches. Deep counts slow the game down. The playoffs see more deep counts.
And, as you could’ve guessed, the playoffs see more pitchers used.
This is another regular-season trend that’s carried over into becoming a postseason trend. This regular season, the average team used 4.3 pitchers per nine innings, which is the highest mark in the 23-year window. And so in these playoffs, the average team has used 4.8 pitchers per nine innings, which is also the highest mark in the 23-year window. In theory, a pitching change only slows things down when it happens mid-inning. But, well, more pitchers also mean more mid-inning pitching changes, and rare is the pitching change that isn’t preceded by a meeting or two at the mound. Mound meetings are delays, even if they’re thought to be strategic, and I would assume that there are more meetings in October.
You’ve got longer breaks between action. You’ve got more pitchers being used. Although you’ve got a reduced amount of overall run-scoring, you’ve got deeper counts and higher stakes. You also just have more relievers and more hard throwers, and these pitchers take longer between pitches, maybe because each pitch is simply more taxing. Remember that, for everyone out there, it’s exercise. Everybody needs a little bit of physical rest. And in the playoffs, when everything is so important, it also stands to reason that each decision might be thought through a little more thoroughly. The end result is that nine innings of baseball in the playoffs take longer than nine innings of baseball in the regular season, by 30 or, in this year’s case, 39 minutes. It’s less of a viewer concern, because the playoffs are also exciting and stressful. The games can get away with taking the time that they do. But, we’ve written so much about how the playoff version of baseball is different from what we see for six months. You can go ahead and add duration to the list.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.




Prior to this year’s 4:37 NLDS Game 5, the longest 9-inning playoff game in MLB history was…last year’s NLDS Game 5 between the Nationals and Dodgers, taking 4:32 for a total of 54 outs and seven runs.
Maybe Dusty Baker is just getting really slow at walking out for all those trips to the mound.
Which of course is a wonderful reason to start the Cubs game at 8 pm. On a weeknight.
I don’t get this. Who thought that was a great idea? The game will be lucky to end by 12:30 Eastern Time. Don’t you want fans to watch the sport’s most important and exciting games? Don’t you want kids to be interested in this so you can bring in new fans? These games are long and start too late.
Bingo.
I’m sure someone can tell me the reason why not to, but knowing MLB is not going to take reasonable steps to shorten the duration of games due to commercials, pitching changes, and general tomfoolery, I really wish they would have start times at 6 PM local for all games (regular season as well). We have these things like the radio, streaming services, and wi-fi today that they didn’t have during Cy Young’s time, and it may allow kids in the Eastern and Central Time Zones the opportunity to watch a game to completion.
Only had to read the title to know that this article was correct
One thing I seem to notice more this season are the visits from the catcher to the pitcher to “get on the same page.” Maybe they have always been so numerous.
It’s still enjoyable because I am a rabid baseball fan. However, no way would I sit through such a long contest (even at home) if it was the NBA, or without the daily fantasy contests that supplement the NFL’s pauses.
They’ve picked up in recent years because players are paranoid about other teams using TV cameras, apples, and other widgets to steal signs.
Just tacking on…
While pitcher success strongly correlates to K rate, there’s almost no reverse correlation for batters, which may surprise some people. So the PS tends to pit high-K pitchers against batters with normal K rates.
Pitches per batter this PS is 4.06, up from 3.90 in the regular season. So even though total batters is down, total pitches per game by both teams is up by 6.
These are probably minor factors in the rising game length, but they’re real and persistent PS effects.
Thanks, John — I was surprised that the author didn’t show total pitch counts to begin with. To a first approximation, that accounts for all of the “live time” (as opposed to dead time) that goes into overall game length.
Tangential to your article, but the admonishment to start y-axes at zero applies to bar charts or other area-based ways of communicating information. Non-zero start values ruins the proportionality of areas and obscures the message of the visualization.
Could you make a point about fielding percentages with a chart that starts the y-axis at zero? Every team was between .980 and .988 this year. The differences would be almost imperceptible, yet they can be meaningful. So moving the start of the y-axis in that case is the only way to enable meaningful visualization.
The essential point of a line graph is to make the trend clear, whether rising, falling, or static. If you have to jigger the axes to do that, so be it.
Please stop this absolutism. These charts are aimed at people who know how to read a chart in context. The lines aren’t the ONLY thing that conveys meaning; that’s why the axes are labeled. And you’re not going to win.
The point is that the advice to start y-axes at zero doesn’t apply to the figures in this article – none of them are bar charts. Your comment makes multiple references to line charts, which are, of course, different from bar charts.
It’s not about absolutism, it’s generally about constructing visuals in a way that doesn’t mislead the reader. The visuals in this article don’t mislead, in part because Mr. Sullivan chose lines rather than truncated bars.
And a bar chart probably *would* be a poor way to visualize fielding percentages, but that’s because the variation between teams is small relative to the absolute value of the variable. One way to overcome this is to transform the variable – instead show the distance of each team from the league-average fielding percentage. This technique is pretty common on fangraphs.
You’re right, I missed that you said bar charts, which are not featured in the article. But I’d take the same view there.
Any number of visuals are “zoomed in” to enhance small differences. It’s not necessarily misleading. I read a lot of charts, and have never felt misled by a non-zero axis beginning. It seems like you want it to be misleading, rather than it actually is misleading.
I think you’re also more astute than most people who read charts, John.
The rise in number of pitchers used probably understates the time impact. It’s the mid-inning changes that add time, and those have been much more frequent this PS than the regular season. I don’t have those exact numbers, but it can be inferred somewhat from the number of outs each RP gets:
— Reg. season: 51% 3 outs exactly … 27% 0-2 outs … 22% 4+ outs.
— Postseason: 31% 3 outs exactly … 36% 0-2 outs … 33% 4+ outs.
In the regular season, most RP stints are exactly 3 outs, and most of those involve one discrete inning; that’s how managers like to run things for the long haul. Most of those changes don’t impact game length.
But in the PS, a big chunk of the normal pattern is redistributed to where they’re probably adding at least one extra mid-inning change, if not two.
John A, you need to be writing longer-form pieces again.
At what point does “very long” become “too long?”
I’m not there yet, but when you combine a seemingly never-ending increase in game length with an increase in three true outcomes, the result is a longer game (duh) with less action. That’s not a good trend.
Bring on the pitch clocks! And while we’re at it, robot umps would speed things up, as would a better replay system, one that doesn’t rely on manager challenges.
As someone who has no team in the fight, these playoffs have been tough to watch. From the 46 minute inning to open the wildcard games to the longest 9 inning game on record, it just isn’t fun.
So if every commercial break is 50 seconds longer, the games are nearly 15 minutes longer off the bat for just the between half inning breaks. Add to that the mid-inning pitching changes; let’s call that 3 per game and you’re at about 17 minutes.
You also have significantly more time after the commercial breaks before a pitch is thrown than in the regular season. The announcers have to get their face time in. We don’t have a metric on that, but if it’s as much as 10 seconds per break, then we’re already to 20 minutes extra without talking about on the field slowness. That’s more than half of the difference.
I’m not saying the game pace can’t also improve, but there are many out there saying this is a pace of play issue which is very misleading.
I was hoping this article would also take a very long time.
I don’t have numbers to back this up, but it sure seems like the post season articles are significantly longer than the regular season articles…
Yeah, but have these 4+ hour games been dull? Heck no. They’ve been GREAT – proof that faster isn’t always better, in baseball, as in other pastimes.
I’m guessing I’m the only one here who feels this way, but I find playoff baseball almost unwatchable. Get in the box, throw the ball and let’s get on with it. I can accept the extra 10 minutes from commercial breaks. It’s the pace within the innings that kills it for me.