For the second part of my ongoing case study investigating how pitcher fatigue could possibly be quantified, I chose to focus on Washington Nationals starter Max Scherzer. Scherzer is fresh off a World Series win, and has been one of the most dominant pitchers in the sport. Scherzer is consistent, and in 2019 he posted a 2.92 ERA, throwing 172.1 innings and collecting 243 strike outs. He remains as one of the game’s hardest throwers even at age 35. I thought he would be an interesting pitcher to look at, as he has been a veteran in the game compared to Syndergaard.
I followed the same steps of analysis of the pitches thrown by Scherzer throughout the 2019 season, but unlike Syndergaard, this included the 2019 postseason. Max threw a total of 3286 pitches in 2019, and while he was sidelined for a month in July, he averaged 100+ pitches thrown per start.
To my surprise, Scherzer’s average velocity on his fastball increased as his pitch count rose. Unlike Syndergaard, the difference in means and medians in each of the pitch groupings was statistically significant according to the Kruskal-Wallis test performed. With a p-value of 0.027 for STATCASTs measured pitch speed against pitch count, this falls under the 0.05 significance level.
Additionally, Scherzer’s spin rate vs. pitch count was not statistically significant with a p-value of 0.08. Again, this is contradictory to the findings from Syndergaard’s data.
Perhaps this is all due to Scherzer being a veteran pitcher, and knowing when to use the velocity he has in the tank when he needs it. Like mentioned previously, he was able to get into 100+ pitches in > 90% of his starts.