The Cause of Lengthening MLB Games

Over at ESPN, Jayson Stark talked to Rob Manfred about the fact that, a year after chopping six minutes off the length of the average Major League game, those gains have been almost entirely lost in the first six weeks of 2016. Included in that piece was this chart, which shows the trend over the last 11 years.

The four minute and 12 second gain from last year to this year is actually larger than any of the per-season gains made during the 2011-2014 stretch when MLB games lengthened quickly; that kind of rise in game length is clearly frustrating to Manfred, especially after the gains they made last year. As the commissioner notes to Stark in the piece, MLB believes there are a variety of factors contributing to the longer games, with players not taking the pace-of-play initiatives as seriously this year, cold weather, and simply the structural change in results all contributing. Stark points out that walks and strikeouts are both up again, so overall pitches are up, and more pitches equals more time. But let’s try to go beyond that and look and see if we can quantify the differences in game length this year.

At the risk of oversimplifying things, there are essentially three buckets of variables that affect game length: the number of batters per game, the length of each at-bat, and then the dead-time elements; pitching changes, managerial arguments, replays, second baseman punching the lights out of bat-flippers, etc… The dead-time bucket is the hardest one to control; you can’t really legislate number of pitching changes easily, and while the league can work on speeding up replay, as Stark notes, there just aren’t enough of them to really be driving a big increase in game length.

So let’s focus mostly on the first two factors; the quantity of batters and the length of each at-bat.

At-Bat Length

This is the bucket that got the most attention in Stark’s piece, noting that walks and strikeouts are both up significantly, and those are the at-bats that require the most pitches, thus lengthening each at-bat. The move away from contact baseball has been a factor in the lengthening of games, and it isn’t a coincidence that games are getting longer at the same time that ball-in-play rates are at historic lows.

But this part can also be easily overstated. We’re averaging 3.88 pitches per plate appearance in MLB this year; it was 3.82 last year, and in that same area for the past decade. At 38 plate appearances per team per game, that change results in an addition of 2.14 pitches for each side, or 4.3 extra pitches for each game. The league takes approximately 23 seconds between pitches, so we’re talking about 95 seconds in added time from the increase in pitches per plate appearance.

When dealing with a four minute increase, this is part of the problem, but it certainly isn’t the whole issue. The issue of whether more walks and strikeouts are good or bad for baseball is an issue worth discussing, but we should be careful not to tie the lengthening of games from last year to this year entirely to that somewhat separate issue. There have been more pitches to batters this year, and time of game is up, but we can’t draw a straight line between those two things and call it a day.

In fact, a bigger issue in the at-bat length is the issue that MLB was trying to change so significantly last year; the time between pitches. The biggest driver of the reduction in game length last year is that pitchers cut almost a full second off time between pitches last year, going from 23.0 seconds in 2014 to 22.1 seconds in 2015. This year, we’re back up to 22.6 seconds between pitches, and with the league throwing 289 pitches per game, that half second increase adds 144 additional seconds to each game.

The total increase in game length over last year is 252 seconds, so the increase in time between pitches results in 58 percent of the the total increase. And this is just straight dead time. While additional pitches may not be ideal, they are at least still game action; time between pitches is just players standing around not doing anything. This is where the league and the player’s association will have to work together to find some differences in how players were acting last year versus how they’re acting this year, and get them back to their 2015 rituals.

Batters Per Game

As you probably guessed from the above paragraph, the fact that we’re at a 252 second increase over last year, with 144 seconds being eaten up by time between pitches and 95 seconds going to extra pitches per batter, there isn’t a lot of time left for much else; those two variables account for 95% of the total increase. So everything else is kind of fighting over the scraps, or are offset by changes in other areas.

But those scraps can still be worth talking about. As noted above, an average MLB plate appearance averages just under four pitches, so each at-bat lasts about 90 seconds, not including the roughly 20 to 25 seconds or so that it takes for a batter to walk to the plate and begin his at-bat; including that, each additional batter adds almost two minutes to a game. So, any kind of increase in number of batters can quickly extend the time of game, which is one of the reasons why the desire to add offense back to the game is seen as anathema to the idea of shortening game length.

But even with offense spiking up a little bit this year — league wOBA is .318, which would be the highest season average since 2010, and we haven’t even played the warm summer months yet — we’re not actually seeing a huge increase in batters per game. Per Baseball-Reference’s league index page, we’re at 37.97 batters per team per game this year, up from 37.8 last year. That means the league is seeing one additional at-bat roughly every three games, which works out to about an extra 30 to 40 seconds added per game. It’s not nothing, but it’s definitely a smaller factor than the at-bat length increases discussed above.

Of course, if league offense continues to trend back upwards, this will be an issue the league will have to account for in the future. Teams averaged over 38 plate appearances per game from 1990 until 2012, so the drop under 38 PA/G is a recent effect, and probably one that won’t last. If the game is going to see more offense, it’s also going to have more batters per game, and that could push the game length higher, which will have to be offset by changes elsewhere if the league wants to get back under the three hour mark.

But for this year, the issue isn’t really extra hitters, or the small upwards trend in offense. The issue is primarily a pace problem (again), with pitchers standing around holding the ball for too long between deliveries. The increases in pitches per plate appearance are also a real contributing factor, but should probably be a lower priority fix than the pace issue, since those additional pitches are at least still something worth watching.

Game length and pace of play are inextricably linked. For MLB, the story might currently be about the average game time going over the three hour mark again, but if the league can get the players back on the pace-of-play initiative that started last year, they can not only speed up the games, but shorten them as well.