April 12, 2019

Manfred's Three-Batter Minimum For Pitchers, If In Effect This Year, Would Have Reduced Game Times By A Maximum Of 40 Seconds

Cliff Corcoran, The Athletic:
To get a sense of the impact the three-batter minimum might have [in 2020], I compiled every pitching appearance lasting fewer than three batters this season. Through Tuesday night, there had been 167 games played in the young season. In those games, there were 153 pitching appearances lasting fewer than three batters, an average of just under one per game. ...

Had it been in place this year, the three-batter minimum would have extended just 53 of those 153 appearances. That's roughly one every three games. With pitching changes limited to two minutes (120 seconds) under this spring's rule revisions, the data suggest that the three-batter minimum would result in a maximum reduction of 40 seconds in the average time of game.

I say maximum because ... many of those mid-inning pitching changes wouldn't have been eliminated, but simply delayed. ... Only six of those 53 appearances saw the pitcher in question enter the game with no one on base and retire every man he faced only to be replaced mid-inning.

Only seven others saw the pitcher enter with men on base and retire everyone he faced without one of those baserunners coming around to score. That's 13 total instances of what I would consider extraneous pitching changes in 167 games, or roughly one every 13 games. ...

Think about that for a moment. These are pitching appearances lasting no more than two batters, and in 40 of the 53 either a batter reached base, a run scored, or both. ...

Forcing those pitchers to stay in the game for another batter or two isn't going to result in anything beneficial to the game or its fans. It's certainly not going to make games shorter, and it's not going to improve the quality of the game they're watching, either.

What it is far more likely to do is alter the outcome of the game. ... Of the 53 pitching appearances in our sample ... 41 were high-leverage situations. The average leverage index of the entire sample was 1.92, which ... is roughly equivalent to having the tying run at the plate with one out in the eighth inning.

Ask yourself this: Would you favor baseball instituting a rule that prevented managers from removing a pitcher who had allowed a 1.352 OPS on the season when the tying run strides to the plate with one out in the eighth? ...

[T]he most significant impact of a three-batter minimum ... will be forcing managers to stick with struggling pitchers in high-leverage situations. That will alter game results, and could impact the standings, in turn. That's a much larger effect, and a negative one. Even worse, consider what will happen when a pitcher has a [poor] outing in a crucial postseason or World Series game under a strict three-batter minimum. It could turn an entire series and possibly alter the ultimate outcome of the season. That's far too high a price to pay for a benefit that would be negligible, at best, and very likely non-existent.

2 comments:

FX said...

This is all nonsense. The reason games are longer is there are many more pitches per game than there used to be. There are many more pitches because the whole philosophy of hitting has changed, seeking one of the three true outcomes. However, with the increasingly trite exception of the home run, the TTO not only make games longer but are boring to fans. What is needed is something to incentivize batters to put balls in play. But stop with the pitching rules, and the number of seconds per pitch, and the time between innings, and the length of commercials. That's all BS. The problem is, there are many, many more pitches being thrown in games. And so games are longer.

allan said...

The reason games are longer is there are many more pitches per game than there used to be. ... The problem is, there are many, many more pitches being thrown in games. And so games are longer.

The facts do not bear this out.

This BRef 2010 blog post compared data from 1988 (the earliest year for which complete pitch data exists) and 2009. I looked at data for 2018 myself.

1988: Teams averaged 136.2 pitches per game
2009: Teams averaged 147.4 pitches per game
2018: Teams averaged 148.4 pitches per game

While there has been an increase of 12 additional pitches thrown per team per game over the last 31 years, there has been an increase of only 1 additional pitch per team per game over the last 10 years.

1988: Pitches per PA: 3.59
2009: Pitches per PA: 3.83
2018: Pitches per PA: 3.89

Back in 2004, The Hardball Times used the "Basic Pitch Count Estimator" equation (3.3*PA + 1.5*SO + 2.2*BB, where PA = 3*IP + H + BB) to estimate the number of pitches per team per game.
1950: 146
1955: 144
1960: 144
1965: 142
1970: 145
1975: 144
1980: 143
1985: 144
1988: 136 (BRef)
1990: 144
1995: 148
2000: 149
2003: 146
2009: 147 (BRef)
2014: 145 (BRef)
2018: 148 (BRef)


Finally, Grant Brisbee wrote an award-winning article in 2017 comparing two similar games played 30 years apart:

April 13, 1984. Mets at Cubs. Home team won 11-2. There was a total of 27 baseruners, 74 batters, and 1 mid-inning pitching change. Total pitches thrown: 270. Time of Game: 2:31.

April 17, 2014: Brewers at Pirates. Home team won 11-2. There was a total of 27 baserunners, 75 batters, and 1 mid-inning pitching change. Total pitches thrown: 268. Time of Game: 3:06.

Note that the 2014 game took 35 more minutes to play, despite having 2 fewer pitches thrown.