The Captivating Improvement of Tyler Chatwood’s Curveball Mechanics
Since joining the Cubs on a 3-year, $38M salary, Tyler Chatwood has faded into the background entering his final season on that contract. He started the contract with a painfully low percentile outcome in 2018 and followed with 162 less TBF and only 5 GS. It’s understandable that the hopes are low entering a shortened 2020 season.
Even with the shortened playing time in 2019, Chatwood arguably exceeded expectations. He finished last season with a 4.28 pCRA, 4.41 SIERA, 22.8 K%, and 11.4 K-BB%. His Barrel% was also exceptional, ranking just behind Marcus Walden in the entire league at 2.9 with a 200 BBE minimum. It wasn’t a great season by any means, but it was quite average. Noticeable improvement and a refreshing breath of fresh air is both a great statement to describe this and some rocky foreshadowing into the topic ahead.
To start, it’s clear that Chatwood’s curveball has been one of the better weapons in his arsenal. In the Statcast era, the pitch boasts a .243 wOBA and .219 xwOBA. Both of these rank noticeably above the league average to varying degrees in the same timespan.
One of the best features of Chatwood’s curveball is its spin. During the Statcast era, Chatwood’s curveball spin ranks 5th at 2952 average RPM with a 500 curveball minimum. However, his 2019 season was his first season reaching an average RPM above 3000, at 3092. This average RPM also ranked at 5th in baseball with a 100 curveball minimum. It was also one of seven curveballs to have an average RPM of 3000 with the same minimum.
Compared to the year prior, this was an increase in average RPM of 266, effectively bringing his curveball spin from Baseball Savant’s 91st percentile to the 98th. For a visual reference of this change in RPM, here’s a 2826 RPM curveball from Chatwood in 2018 (his average then), and here’s 3092 RPM curveball from him in 2019.
Another interesting tidbit of Chatwood’s curveball spin is its standard deviation. Here’s the normal distribution chart spanning the range of the spin of Chatwood’s curveball in 2018.
And here’s the normal distribution chart spanning the range of his curveball spin in 2019.
The tighter distribution is a bit noticeable and, indeed, the standard deviation backs that claim up. It improved by 10.2, from 134.3 to 124.1. This should allow his spin to be more repeatable. Given that his standard distribution became lower, it clustered more around the mean. This makes the data more reliable than if it were more spread out, like in 2018.
Another thing to look at involving spin rate is active spin. Essentially, this is the percentage of spin that contributes to active movement. In 2019, Chatwood’s curveball was 73.7%. With a minimum of 1000 pitches, this ranked 72nd and in the 65th percentile. In 2018, however, his active spin was 79.8%, which ranked 31st and in the 75th percentile with a 1500 pitch minimum.
The thing about active spin is it’s best with spin rate. This is because, ideally, a pitcher would want a high number of both rather than one or the other. When combining active spin with the spin rate on the pitch, Chatwood’s 2019 prevails in spin contributing to the movement. In 2019, the number is 2279 when rounding to the ones place compared to 2255 in 2018. It’s a small difference, but still one nonetheless.
Spin contributing to the movement isn’t the only possible method to measure it, however, and the pitch also improved in that area from 2018. When initially looking at inches of break, the curveball’s vertical break decreased by five inches while its horizontal break increased by 0.3 inches. However, when looking at his inches vs average, his vertical break increased by 0.1 inches (11 to 12% break vs average) and his horizontal break increased by 1.3 inches (-9 to 5% break vs average). That’s a noticeable improvement.
Chatwood’s movement on his curveball not only improved from 2018, but it was also quite competent. Out of qualified pitchers in 2019, his vertical break ranked 23rd in the league and in the 92nd percentile. His horizontal break, while being noticeably worse, still ranked 124th and in the 55th percentile.
The average velocity of his curveball also noticeably improved. From 2018, it saw a healthy increase of 3.2 mph, from 78.2 to 81.4 mph. With a 150 curveball minimum in 2018 and a 100 minimum in 2019, those rank in the 50th and 78th percentile, respectively. 2019 also marked the first season in Chatwood’s career with an average velocity on his curveball above 80 mph. His previous high was marked in both his 2011 and 2017 seasons at 79.3 mph.
Not only that, but the standard deviation of Chatwood’s curveball velocity also saw noticeable improvement. Below is the normal distribution chart of the range of Chatwood’s 2018 curveball velocity. It has a standard deviation of 2.0 mph.
Below is the normal distribution chart of the range of Chatwood’s 2019 curveball velocity. It has a standard deviation of 1.6 mph.
The strikeouts and the potential of it improved on the pitch as well. To start, Put Away Rate is a good stat to describe strikeouts. Basically, the stat measures the strikes compared to the number of pitches on two-strike counts. Below is the relationship between the stat and strikeouts for relief pitchers with a 300 two-strike count and 50 strikeout minimum in 2019.
The coefficient of determination (r²) is quite strong on this data set at 0.684, and that makes sense. The two stats are descriptive and, theoretically, the higher the PAR the more strikeouts. With a more condensed volume of pitchers, it makes sense that the data is more well-correlated.
With all of that said, Chatwood’s PAR on his curveball has been quite adequate each of the last two seasons. In 2019 it was 25.0%, and in 2018 it was 25.4%. While this is a decrease in the percentage, his 2019 ranked at a slightly higher percentile than his 2018 with a 50 curveball thrown at a two-strike count minimum. Both round to the 72nd percentile, however.
Not only that, but Chatwood’s whiff% on the pitch greatly increased. This stat measures the number of swings and misses compared to the number of total swings. A linear comparison of the stat and strikeouts with a minimum of 500 swings and 100 whiffs for relief pitchers in 2019 is charted below.
As the graph shows, the stat has a stronger correlation to strikeouts, albeit with a lower sample. In 2018, the pitch’s whiff% in 2018 was 32.8, which ranked 70th and in the 53rd percentile with a 50 swing on a curveball minimum. In 2019, however, it was 47.4, which ranked 6th and in the 96th percentile. That’s an extreme difference, however, it’s a very good one.
So, with all of that describing the strikeouts talked about, it’s time to talk about the actual result. His K% itself on the pitch decreased by 0.3% from 2018-2019, however, similarly to PAR, it ranked better in comparison than at face value. With a minimum of 25 TBF in each season, Chatwood ranked 15th and in the 91st percentile in 2019 and 18th and in the 89th percentile in 2018.
The wOBA and xwOBA of the pitch also had really intriguing differences. In 2018, the wOBA and xwOBA mean was a very capable .242, which ranked 23 points below the league average of .265. In 2019, however, his xwOBA decreased from .231 to .228 while his wOBA saw an immense increase from .252 to .325. His wOBA and xwOBA mean was just two points above the league average of .275, so it was still right around average.
Overall, Chatwood’s still not a good pitcher. He saw some strong, noticeable improvements, but it’s unlikely that he’ll repeat his 2019 season as of right now. However, that doesn’t make the improvements that he saw any less impressive. As I said before, it’s a nice breath of fresh air for Cubs fans who are still suffering from his disastrous 2018 campaign. And, even if he regresses back to his 2018 season, the Cubs should find a home for his 2 DRS in less than 100 innings pitched last year in the field. You never know.
