The Breakdown Series: #01—Maeda’s arsenal and pace-of-play

Ed note: Welcome John to the front page, as he makes his Twinkie Town debut today. He’ll be providing us with some more analytically-oriented takes going forward.

Welcome to the Breakdown Series, a regular collection of interesting statistics and advanced analysis related to the Twins. With this series I hope to use a data driven analytic approach to uncover insights and provide some deeper explanations of Twins baseball. The offseason has given us plenty to investigate and today we’ll start by digging into Kenta Maeda’s arsenal and looking at the Twins’ pace of play.

★ ★ ★

Maeda’s Arsenal:

Kenta Maeda’s acquisition has been covered and dissected from many angles here at Twinkie town and other places. My colleague Kyle Edelbrock already did a quick dive into Maeda’s approach using MLB Statcast data. Let’s dig into this data a little further to see what else we might learn about Maeda’s pitching style and approach to attacking hitters. At its core, pitching is about making it difficult for the hitter to hit the ball. Throughout baseball history this has most often been done through changing velocity, throwing pitches that move in different directions, and changing the location of pitches in and around the strike zone. How is Maeda equipped to do these things?

First, let’s look at Maeda’s arsenal:

What we see overall is a pretty common arsenal for a right-handed starter: four seam fastball, slider, changeup that make up almost 90% of his pitches thrown, with the occasional curveball and sinker mixed in. The average velocity of these pitches ranges from the high 70s (curveball) to the low-90s (fastball varieties), with both the changeup and slider in the low to mid-80s. The chart below shows the velocity distributions on Maeda’s arsenal from 2019. We can see that he has the entire range from the low 70s to the mid-90s covered with his five pitches.

It starts to get even more interesting when we look at how Maeda deploys these pitches against right-handed and left-handed batters. Look at the gold highlights in the table above – like many pitchers Maeda has a clear preference for using his slider to attack righties and his changeup to attack lefties, while his fastball usage remains similar against both. We also see a clear preference for using his curveball against lefties but not against righties. If you have watched many Major League Baseball games you probably recognize that this kind of approach is a common pitching strategy. Though it may be a common strategy, it’s clear from this data that Maeda has options for varying velocities from pitch to pitch.

Next, let’s dig into his pitch movement:

The chart to the left is generated from statcast pitch tracking data from baseballsavant.mlb.com and shows the movement profiles of the pitches in Maeda’s arsenal. The scatter plot is from the catcher’s point of view (imagine Maeda was throwing the baseball to you) and the vertical line bisecting the chart implies a straight pitch with no horizontal movement. The top of the chart implies a pitch with no vertical movement. All pitches thrown on planet Earth have some degree of vertical movement (thanks gravity) and Maeda’s pitches are no different. What this data shows us is that Maeda’s fastball (red), sinker (orange), and changeup (green) all feature horizontal movement to his arm side - meaning moving in towards a right-handed batter when thrown by a right-handed pitcher. His slider (yellow) and curveball (blue) feature horizontal movement to his glove slide - meaning away from a right-handed batter when thrown by a right-handed pitcher. What’s particularly important in terms of making pitches difficult to hit is these movements relative to one another. For example, on the first pitch of an at bat Maeda could throw a sinker, which averages about 14 inches of horizontal break in towards a right-handed batter. On the next pitch he could throw a slider, which the breaks the opposite direction nearly 5 inches on average. So, from the 1^st pitch to the 2^nd pitch Maeda can vary the movement about 19 inches – a range that is wider even than the full width of home plate itself (17 inches). That’s a lot of area for a batter to try to cover with the bat and we haven’t mentioned these two pitches are also about 9 MPH different in terms of velocity.

To further maximize the pitcher’s advantage, he can also work to ensure that the pitches appear similar to one another. This makes it harder for the batter to discern and identify the pitch out of the pitcher’s hand. Perhaps the simplest way to make pitches look similar is to release them from the same point as consistently as possible - from pitch to pitch and across all the different pitches in the arsenal. How does Maeda do with this?

This chart, using Fangraphs data shows that Maeda’s release points on all the pitches in his arsenal are tightly clustered together. It is especially important to note is that we can hardly see the fastball and sinker plots here because they are covered by the slider and changeup plots. This indicates that Maeda’s three most prominent offerings are released from very similar points which helps make them harder to distinguish from one another, and harder to hit.

The last tool in a pitcher’s toolbox is pitch location - Varying where pitches are thrown in and around the strike zone. What might Twins fans expect from Maeda in terms of pitch location?

Again, we can turn to Statcast, and find plots of the locations of Maeda’s pitches. This is 2019 data from baseballsavant.mlb.com and is shown from the catcher’s perspective.

From this, we can discern some clear patterns. Maeda uses his fastball pretty much all over the strike zone – up, down, inside, outside, arm-side, glove-side. There is a clear clustering pattern on his slider – low and away from right handed batters seems to be the preference. For the changeup, the opposite seems true – low and away from left handed batters. Given what we learned above about these pitch’s movement profiles earlier these location patterns make logical sense.

Throughout baseball history, a pitchers most effective tools have been changing velocity, throwing pitches that move in different directions, and changing the location of pitches in and around the strike zone to make their pitches hard to hit. When we consider Kenta Maeda’s arsenal and approach we find he releases his pitches from a very similar release point, with significant velocity differentials, very different movement profiles, and can locate his pitches in all different parts of the strike zone. Hitting is hard, guys. And Kenta Maeda is well equipped to make it harder.

★ ★ ★

Picking Up the Pace:

The 2019 Twins were the 7^th slowest MLB team in average time per game at 3 hours, 14 minutes, per Baseball Reference. The overall Major League average game time in 2019 was 3 hours, 10 minutes—which is a new all-time high as the pace of play continued its recent trend of slowing year over year. Curiously, the three fastest teams in MLB last season were the Twins’ AL Central division foes: Detroit, Kansas City, and Cleveland.

Digging into why the 2019 Twins played slowly, we quickly uncover that the pitching staff worked very slowly, tied with the Red Sox for slowest in baseball with an average of 26.2 seconds between pitches. Fangraphs tracks a statistic called “Pace” that essentially measures the average number seconds elapsed between pitches thrown by a pitcher. The MLB league average in 2019 was 24.9 seconds. The Twins starting rotation (26.0 seconds, slowest in baseball) and relief unit (26.4 seconds, 8^th slowest) each contributed to slowing down the Twins game times. In general, games tend to slow down when teams get into their bullpens – MLB wide, starting pitchers worked with an average pace of 24.4 seconds, while relievers averaged 25.7 seconds. Let’s take a look at how the Twins’ off season additions to and subtractions from the pitching staff might help pick up the pace in 2020:

The tables above show the Twins off season pitching additions tend to work faster than the pitchers they are replacing by almost 2 seconds per pitch. For simplicity, let’s just assume these additions only replace the 8,243 pitches thrown by the departing players (a conservative estimate given the larger, starting roles the Twins are hoping Bailey, Hill, and Maeda play). If these pitches are replaced by the faster workers added to the roster, we can estimate that we might expect Twins games to move along about 250 total minutes more quickly over the course of the season. Over 162 games, that alone would reduce the average game time by about 1 minute and 32 seconds.

Unfortunately, there is little correlation between pitcher pace and other performance statistics like ERA, FIP, or strikeouts. Working faster does not necessarily mean a pitcher will perform better and working slower does not necessarily mean a pitcher will perform worse. However, according to research done at fiverthirtyeight.com pace is highly correlated across seasons, meaning a pitcher who works fast or slow in one season is very likely to continue to do so in future seasons.

So, while this data might not suggest we’ll see improved pitching performance from the Twins pitchers, at least we might get a better fan experience and a little bit of time back in 2020. Quicker working pitchers make it easier to stay focused on the game, whether you’re on the field, in the stands, or on the couch.