Notice: Trying to get property 'display_name' of non-object in /var/www/html/wp-content/plugins/wordpress-seo/src/generators/schema/article.php on line 52

Let’s fix the pace-of-play problem. It shouldn’t be that hard. But first, let’s diagnose the patient.

Here’s the average game time for the past 20 years.

We can see the obvious trend line, going mostly upward and adding nearly half an hour to the game. Ouch!

But of course, we’re not just worried about the length of the game. We’re worried about there being too much dead space in the game. And dawdling. Major League Baseball players dawdle like a bunch of 5-year-olds at bedtime. So, here’s a graph of the last 20 years, showing the number of minutes it takes to manufacture one plate appearance. (This is just average game length in minutes divided by the average number of PAs.)

Same pattern again. From the beginning of the graph to the end of the graph, we see that it takes an extra 20 seconds or so just to get one more batter through … whatever it is he’s doing out there.

OK, one more. Maybe the problem isn’t that plate appearances are taking longer. Maybe the problem is that more of them are ending in strikeouts, walks, and the occasional hit batsman. Y’know, the boring stuff about baseball that isn’t fun to watch. Who wants to come to a game to watch a bunch of strikeouts anyway? We have entered an era in which specialist relievers who only have to go one inning are able to just come into the game, air it out, and no one seems to want to swing to make contact.

Anyway, here's the data. This is (K + BB + HBP) / PA.

Again, we’re seeing fairly unabated growth over these 20 years. Games are taking longer, players are obviously taking their sweet time, and the strikeout scourge is ruining everything. We need to do something. Now.

There’s just one little problem. It’s 1997. Don’t believe me? Here’s the number one song on the radio right now.

Did you see the Jeffrey Maier thing? That just happened a few months ago!

Also, I’m having trouble with my graphing program. It keeps shifting the years at the bottom of the graph ahead by 20. For some reason, it’s saying that the graph covers the years 1996-2016. Not sure why. All the graphs above should actually say 1976-1996. Sorry about that, Angelfire readers. I’ll try to fix that by the time the CD-ROM version of this article comes out.




Length of Game

Minutes per PA

Not in Play/PA













(And yes, those are real numbers now.)

It’s worth pointing out that the distance from 1976 to 1996 on all three measures is larger than the difference from 1996 to 2016.

Here’s the (real) 1976-2016 graph for “not-in-play balls per PA." The graphs for all three of these measures across that time period all have the same basic shape.

It’s been a slow, mostly steady climb over the last 40 years, but the majority of the “damage” was actually done over the first two decades in that graph, rather than the last two. Now we know what we need to do. We need to figure out what happened between 1976 and 1996!

When we talk about the pace-of-play issue, there’s a tendency to frame it as something that happened in the past few years. We need a longer view, not just for the “hey kids, get off my lawn” aspect of it, but because it matters if we actually want to address it.

I don’t want this to be taken as a cry that the pace-of-game problem is so far ingrained that there’s no possible way that we can fix it, so why should we bother trying. Whatever your views are on whether the game is aesthetically broken, commissioner Rob Manfred has stated that shorter games and more action in those games are priority goals for him.

The point is that we need to be realistic. This “problem” has been around a lot longer than the entitled batters and the octopus bullpens filled with mutant power arms and PITCHf/x and the “foul everything off” strategy and Moneyball and a host of other things that get blamed. Baseball has been evolving toward a slower, more “boring” game for years. It means the forces that are driving that evolution are probably deep, systemic forces that are much harder to change than a couple of minor rule tweaks might solve.

What if the real reason that players are a bit slower is that media coverage has intensified and that players are now expected to interact with fans more than they were 40 years ago? That might leave them just the tiniest extra bit tired. What if the death of the scheduled doubleheader—which means fewer travel days—makes for more fatigue as the season goes along? What if the reason that we have more strikeouts is that, despite all the concern about pitcher injuries claiming so many young arms, the same advances that make Tommy John surgery possible are actually keeping young arms healthier to the point where the game has actually slowly been over-run with pitching?

It’s possible that with a few of those tweaks, we’ll see some minor progress, but again, I think we need a more sober view of what “progress” looks like. It might be a couple of minutes and basically just holding the line against further spread of the disease. It’s trying to curtail the symptoms, rather than addressing the root cause. Maybe there is no treating the root cause. This might be a fight against evolution. That’s a hard one to win.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
I just came back from Spring Training in Arizona, and while I was there, I immediately noticed that games were usually over in less than two and a half hours. If this can be attributed to something other than lots of beer and sun, it might be worthwhile to imitate during the regular season.
There are no commercial breaks in spring training?
One problem with comparing 1976-1996 with 1996-2016 is that there appears to be a plateau from 1996-2006 followed by a sharp takeoff. That would indicate that something important occurred both in 1996 and 2006, rather than just the continuation of an earlier trend.
Get ready for a 2-phase plan!

1. Force the batter to stay in the box between pitches. Not really a rule change, just an enforcement change... umpires should be discouraged from granting timeouts to batters.

2. Levy a dawdling tax. Figure out how long it realistically takes to get the ball back, get a sign, and pitch, e.g., 20 seconds, and then estimate roughly how much a wasted second costs the league. Every time a pitcher takes more than 20 seconds to throw a pitch, charge the pitcher's team a set amount per second.

This will speed up pace of play in the most natural way possible. Teams will take a flexible approach to decide how to reduce the tax (and by how much). For example, once the Yankees find out that a mound visit will cost the team $30,000, maybe ownership asks Jorge Posada to keep his mask on. And if he refuses, then great, the Yankees will have to pay that much more into the revenue sharing pool.

Soon enough, teams will take this into consideration when negotiating with free agents, and players will begin to understand that dawdling will affect their earning potential.

And teams will start inculcating minor-leaguers with fast habits.

I have no idea what a wasted second might be worth, but I'm sure there are ways to estimate it based on whatever the correlation between fan interest and game length happens to be. Advertising is already charged on a per-second basis, maybe the correct value of the dawdling tax is similar.
I actually like the dawdling tax.

The most obvious thing to eliminate is relief pitcher warmups. In no other sport does a substitute come in and warm up for 5 minutes before playing. Let them throw 3 pitches because maybe the depth perception is a little different but that's it.

Also limit each team to 3 timeouts which includes all mound visits, again...another thing that is in every other sport.
Question: how is the game length here quantified? Does it go from the first pitch to the last out, counting everything in between? Or does it count just times where actual game action is happening?

If it's the former, then it seems that the answer is pretty intuitive: commercial advertisements having more of a place as more and more baseball is put on TV. Makes sense that there'd be a steady climb in TV exposure from 1976-1996, and that there would be a big jump in the mid-2000s when became a truly viable means of watching the game.

To get to rofldude's point, I'd be curious to see a study that compares the length of televised spring training games vs. non-televised ones. I bet that there'll be a not-insignificant difference.

Sure, pitch clocks, batter's box rules, and "encouragement" to not take forever will put a bandaid on it, and maybe stem the increase in game length. The way to meaningfully shorten games is to cut ads out.

This, obviously, will never happen, but it seems like every article discussing this problem constantly dances around a pretty clear solution because there's no way to solve it.

It sucks to basically just throw our hands up and say, "well, there's nothing we can do," but until press outlets or players or something start to call out the league for causing a problem that it's blaming on other factors, it's really the only thing that we can do.
So a cursory glance at game lengths for the Tigers spring training games this year indicates that the opposite of what I posited is actually true.

Televised games (12 total): 166 minutes on average
Non-televised (20 total): 180 minutes on average

This is rife with sample size issues and doesn't control for things like pitching changes or runs scored, so it's mostly useless, but still interesting.
Nice article.
The biggest thing is keeping the batters in the box, and stop catchers from habitually going out to the mound. Correct those and you'll short games AND have more actual play.

By the way, NFL and college football games are getting longer too.
I'd like to see the length of game data mapped against the television commercial break time over the years. In the long-ago past, virtually no games were televised on a daily basis; now, courtesy of advances in cable and satellite television, almost all games are televised, which means that time between half-innings is dictated by commercial breaks. Why are games in Spring Training shorter? Because there are relatively few televised games, so they can be played at a more natural course. jonvanderlugt has it precisely right, but I bet that there's some way to provide correlation. Stopping blaming what's going on during the actual game and starting laying the cause where it belongs: Greed.
Commercial breaks and money are definitely part of the problem. But so are the batters who have to step out of the box after every pitch no matter what. And replay may be a factor as well. As for the multiple pitching changes, it would be interesting to see those numbers over time.....probably another upward trend.

Games have gotten too long, with too much downtime during the "action." I hope some systemic changes can be made....
Television breaks are longer but it depends if the problem being addressed in pace of play or length of game. I don't mind the breaks between innings as it's the time to go the bathroom or get a snack.

The issue, in the minds of some including myself, is the lack of action between the breaks. Too many strikeouts, walks, mound visits, etc. and not enough balls in play.
Unscientific estimates from barring starting points, but I'm on my phone and going for direction and magnitude (some evidence via below links, some wild ass guesses).

Extra :30 commercial per break: +8.5 min
3 extra pitchers per game (mound visits) :+3 min
3 extra pitchers per game (warm-up): +6 minutes
22 extra pitches per game (since '88): +7 min
1.7 extra relief pitcher seconds per pitch (24 pitches): +0.7 min
2 extra seconds between pitches: +9.8 min

That all adds up to 35 extra minutes. It's not really one thing, it's a bunch of things, some of which can't be changed (e.g. We can't force a batter to swing on pitch 3.6). Slowness begets slowness it seems.

...from *varying* starting points.