Being commissioned for baseball seemed to be very easy at first. Rob Manfred took office in January 2015, announced his intention to accelerate the game and immediately had results. The duration of a game of nine entries was reduced in six minutes from 2014 to 2015, equaling the largest decline in history. At that rate, baseball would have completely disappeared in 30 years and Manfred would have to retire.
But like the two-homer games on Inaugural Day, this pace was not going to hold. Baseball added four minutes to the average per game last year, and this year’s games appear to add more time to a new record.
Despite some minor changes for the 2017 season – requiring managers to decide faster if they were to challenge a move and implement automatic intentional tickets – the average duration of a nine-inning game is now 3 hours and 5 minutes – the highest average in baseball history and a five-minute increase compared to last year. It is also the second highest year-on-year increase since integration. And as the games get slower with those up for the roster in September, the final month might add one or two more minutes to that average.
Manfred, who continues to insist that the league will do something about it, has drawn a distinction between playing time and game pacing. As he told reporters in June:
The playing time is sometimes what happens on the ground competitively. How many races are scored, how many races are based, how many times you change to the pitcher. Those are things I do not want to control. Because that’s what the competition is about. That is in the hands of the teams.
So how much of the nine minutes that have been added to the average since 2015 fall under “what happens on the ground competitively?”
That is a complicated question.
There are more appearances to the plate per game.
The base percentage hit bottom in 2014 and 2015, with the lowest OBP at league level since the early 70’s. Although strikeouts have been increasing since then – in more than one punch per game – everything else Has remained the same or has benefited from the offense: There are two-thirds more per game, 25 percent more homers, slightly more balls and slightly fewer sacrifice strokes, high sacrifice and runners trapped in attempted theft.
The accumulated result is that in each game we have almost an appearance on the extra plate – 0.78, to be more exact. That partial appearance on the plate is responsible for 1 minute and 20 seconds per game.
There are more pitches thrown by appearance on the plate.
This has been happening for decades, as the average shock has increased from 3.57 pitches in 1988 to 3.82 per appearance in 2009. The batters have been more patient while the OBP has been replacing the batting average in prestige. The greater speed of the pitchers and a more deceptive repertoire have led to more strikes fanned (those that usually lengthen the turns). And the punch paradox – in which batters and pitchers have simultaneous strategies that lead to more strikeouts – has led to deeper counts.
After 2009, that figure of 3.82 pitches per appearance remained stable. It seemed we had reached the peak. The pitch-per-turn rate did not vary by more than a hundredth pitch in either direction for six years, until it suddenly jumped to 3.87 by 2016 – the highest recorded rate since it began to carry that statistic in 1988 – and 3.90 this year. A pair of hundredths of pitching per appearance on the plate, applied to 76 appearances per game, add six pitches. Each pitching extends one turn in about 24 seconds, so now we’ve added about two and a half minutes more.
Three minutes, 45 seconds that have now been counted.
There are more appearances on the plate (and thus, more pitches made) with runners on base.
Pitchers universally take longer to throw the ball when there are runners on base – on average, about 7 additional seconds – and the above-mentioned higher percentage of base would tell us that a higher percentage of baseball is now occurring with people on base.
In fact, about 43.2 percent of all appearances on the plate come with people on the trails. In 2015, that figure was 42.9 percent. That does not do well to our watch: That means it is done close to an additional pitch per game with people on base, which on average adds us about 7 seconds.
There are more pitcher changes.
The pitcher changes increase in quantity as the season progresses. But if we take a look at the first halves of the past three years, we can see that the pace of 2017 could set a new record for pitchers per game on the side:
2015: 3.97 (4.11 all season)
2016: 4.02 (4.15)
Not all pitching changes add time to the game. Most of the new pitchers come in to start a new entry, but about one-third of the changes are made in the middle of the episode, meaning that since 2015 has occurred near a mid-entry change for every 12 games. Each pitcher change takes about 2 minutes, 20 seconds, which adds 20 seconds per game.
So we counted little more than four additional minutes. Where does the rest of the time come from?
The defenses take more time between pitches.
Well, this is complicated. FanGraphs uses the PITCHf / x time stamps to record the time that pitchers are taken between their pitches within a turn. They call this statistic “pace” and it has been increasing:
2015: 22.8 seconds
2016: 23.2 seconds
2017: 24.3 seconds
Multiplied that by almost 400,000 pitches, that number only represents five and a half minutes. So, easy: Have pitchers speed up their work by setting up a pitching clock, as they did in Double A and Triple A – with great success – and how Rob Manfred looks set to implement next year with or without Blessing of the association of players.
Except that this is where the thing is complicated a little. Baseball Prospectus also uses PITCHf / x time stamps to keep track of the amount of time pitchers take between shipments within a turn. They also call this statistic “pace” and the same … has gone down:
2015: 21.7 seconds
2016: 21.8 seconds
2017: 21.1 seconds
The difference is at what intervals each method excludes as abnormal. FanGraphs discards any interval of more than 50 seconds, and Baseball Prospectus discards any of more than 60 seconds. This is based on the shared assumption that something else has happened other than that “a pitcher stares for a signal”: a visit from the catcher, a hitter twisting in pain after hitting his leg with a foul, a shortstop Who comes to talk about who will cover in case of a difficult, whatever.
But BP also scraps any interval that includes attempts to catch the runner at the base. And discard any interval after a foul. The foul ball usually takes more time: The pitcher receives a new ball that begins to rub, the defensive has to adjust again after trying to catch the shot, the batter is allowed to leave the batting box and adjust his gloves , etc.
If one of these approaches is “right” it is less important than the fact that they give us two different sets of data with two different paths, and our answer might be in the middle.
In a set of data – the pitcher does the pitching, then gets the ball back, then does the pitching, and then gets the ball back, and so on – the pitchers are not working this year slower than they did before. In fact, they are working slightly faster!
But in the other, the game has slowed down, and continues to slow down. This suggests that slowness is caused by slight interruptions – rubbing a new ball, or a shot to surprise the runner – which causes most of the damage to our calendars. It suggests that pitchers, defenders and batters are taking longer to get back into rhythm after these breaks. Where even the slightest interruption in the rhythm of the game exists, the lost time rushes to fill it.
Manfred will continue to try, and will have successes and failures along the way. The bad news is that “what happens on the ground in a competitive way” has naturally led somehow to longer games – more base runners, more pitches, more pitcher changes. However, the good news is that there are many things going on between pitches that none of us are missing too much. The natural pace of throwing and catching baseball is slow, but that is not what is slowing it down. To fix that, Manfred would have to seal the holes that let the players interrupt the pace of the game.
Why are MLB Games Five Minutes Longer this Season? Manfred continues to insist the league will do something about it. Average duration now is 3 hr 5 min.