# Historical Look at Pitcher Usage

#### Introduction

In a recent post I compared two seasons, 1968 and 2018 with respect to a number of different variables such as the game length, runs scored, and margin of victory. I also briefly looked at the number of pitchers used in a game. I thought it would be interesting to explore pitcher usage more carefully, looking at the dramatic changes in this variable and also exploring how current pitcher usage varies between teams.

To illustrate how things have dramatically changed, I have copied the pitching component of the boxscore for the first game in the Phillies 1968 season. For this game, the Phillies used a single pitcher, Chris Short, who threw a complete game four-hit shutout.

Let’s contrast this with the pitching score for the first game in the Phillies 2018 season. Aaron Nola pitched pretty well in this game, but he was taken out in the 6th inning and he was followed by five relievers who got the remaining 10 outs of the game.

#### How Many Pitchers are Used in a Game?

Using Retrosheet play-by-play datafiles, I collected the number of pitchers used by each team for each of the seasons 1960 through 2018. Below I graph the mean number of pitchers used per team per game against season. From 1960 through 1980, teams tended to use 2-3 pitchers per game (average close to 2.5). But starting with the 1980 season, things changed — the average number of pitchers per game per team increased up to the current average around 4.5.

To get a better understanding of this phenomenon, I graphed the number of pitchers used for all games for all teams in the 1968 season. I have shaded the bars corresponding to 3 or fewer pitchers used. For all teams, it appears it was uncommon to use more than 3 pitchers for the 1968 teams.

Compare this with the number of pitchers used per game for teams in the 2018 season. The pitcher use is dramatically different. Now it is rare to use three or fewer pitchers. We know there are few complete games in modern baseball so the bar heights for one pitcher are very low. Also looking at the table, we also see considerable variation between the teams. For example, Detroit appears to have a relatively large number of games where they used 3 or fewer pitchers. The Phillies rarely used one or two pitchers in a game.

#### How Many Batters Does a Pitcher Face?

Another way to look at pitcher usage is to examine the number of batters faced (BFP). For example, if a pitcher faces 27 batters, he is going through the batting order three times. I think a team is conscious of this number — they might be reluctant to have a pitcher face more than 18 batters since then he will be facing a batter for the third time. (It is commonly believed that a batter’s success will improve against a specific pitcher for more plate appearances.)

For all pitcher appearances in the 1968 season, I collected the number of batters faced. Note that a relatively high proportion of these appearances corresponding to starters who faced between 20 to 40 batters. Over half of the appearances corresponding to relievers who faced between 1 to 15 batters.

Let’s compare this distribution of BFP with the distribution for the 2018 season. This is very different. Relatively few of the pitcher appearances correspond to starters. Pitching currently is dominated by relievers and over half of these appearances correspond to relievers who face 3, 4, or 5 batters.

#### Differences Between Teams

Okay, pitching usage has changed in modern baseball. But the manner of using pitchers varies across teams. For example, one strategy is to bring in a relief pitcher to face a single batter — a left-handed specialist is a type of pitcher who is supposedly effective in retiring a left-handed hitter. Below I graph the number of pitchers facing exactly one batter for all teams in the 2018 season. Teams like Cleveland and both Chicago teams commonly used the pitcher against a single batter strategy. In contrast, this was very rare for the Pirates or Yankees to do this.

Also we are starting to see different strategies in the use of starting pitchers. Below I have displayed histograms of the number of batters faced by the starters for all teams. If you look at these 30 histograms, one outlier should be obvious. The Rays had an unusual strategy of starting the game with a relief pitcher and they had a high number of starters who faced 9 or fewer hitters (which means that they would face each batter in the lineup a single time.) Maybe other teams will adopt the Rays’ strategy in the 2019 season.