R.A. Dickey is quite the esoteric baseball player. He throws a knuckleball. He has hiked Kilimanjaro. Even more odd, he did so for charity. He is a Star Wars nerd. He names his bats after objects from Lord of the Rings, Beowulf, and other epics. He is leading the league in strikeouts.
Wait, what? I am sure most of us knew the odd (or rather, AWESOME) stories about his adventuring and use of his English Lit degree, but I was shocked to find out he is co-leading the MLB in strikeouts with Justin Verlander. Several bloggers have covered the reasons beyond his success so far, but I want to delve into the question of sustainability. Put another way, should he be seriously considered as one of the best in baseball, or is he due for a major correction? Is there an impending negative regression?
The timing of this post was fortuitous, as it comes right after Dickey threw his second consecutive one-hitter. It would be very tempting to use his recent dominance as a sign of continued excellence, but that is not sabermetrically rigorous analysis. So, as usual, we turn to advanced stats to evaluate his performance thus far this season and then try to make some predictions for the remainder of the season.
The first stat that jumps out is his increase in K/9. His highest value before 2012 was in his first year as a consistent starter in 2003 with the Rangers, where he posted a 7.25 K/9. Certainly nothing of note, but higher than his career average of 5.95. He has only been above MLB average in that category twice: 2003 and 2012. Therefore, one must be skeptical of his 9.36 K/9 this season. One would expect to see this regress towards his career number. That being said, most people agree that K/9 can become consistent around 150 batters faced. Dickey has seen 374 batters thus far, so perhaps his K/9 will not regress to the point that it would hinder his peripherals. Just for kicks, I regressed his K/9 and got a 2012 rate of about 7.22. Certainly not a bad result.
Going further, what makes a K/9 high? Certainly getting swings and misses help. After watching some film (always a useful practice) and consulting Pitchf/x data (also a useful practice), there is something to be said for Dickey getting more missed bats. To wit, his Z-Contact % is currently 77.6%, much lower than his career average of 85.7%. What this tells us is that batters are not able to get contact on pitches in the strike zone as often as they have been. I understand the debate about how much control a pitcher has on getting swings and misses, especially in the zone, is a contentious one, but Dickey has clearly demonstrated he is somehow missing more bats than usual. I, for one, credit his increased adeptness of his knuckleball. Before he was able to master it better, the ball was getting more rotation and therefore batters could hit it more easily. (Keep in mind the physics of a knuckleball; spin and rotation are the nemeses of this pitch). His biography talks in-depth about this process and how switching from the traditional three finger to two finger knuckleball helped him gain a measure of control over the pitch.
To the Pitchf/x charts and graphs! One quick fact jumps out; this kid is throwing more knuckleballs. He throws his ‘fastball’- if we can even call it that- about 14.0% of the time, which is a career low for him as a starter. His knuckleball is being thrown for about 85.0% of his pitches, the second highest percent of his career. Along with his memoirs, this gives credence to the above idea; he is mastering a documented tough pitch to hit, and it is getting him results.
To the heat maps! Courtesy of Fangraphs, we find this:
Conclusion: against left handed hitters, which would be a platoon-split disadvantage for the pitcher, we see Dickey keeping the ball on the edge of the plate versus in 2011 where the knuckleballs were decidedly more down the middle, ripe for trouble. Again, this is far from explaining his dramatic increase in effectiveness, but it begins to explain the increase in strikeouts and weaker contact.
So, is all this sustainable? Overall, I would expect to see some negative regression. His LOB% is 83.8%, which is certainly lucky. This strand rate is simply not going to stay that high, runners are going to start scoring and his peripherals will decline. That being said, his FIP and xFIP are nearly identical (2.72 to 2.70). Conclusion: he is performing about as expected. Assuming his BABIP regresses to his career average of 0.291 from his current 0.246, the LOB% will decline, but I doubt it will dramatically reduce his effectiveness.
R.A. Dickey is currently going bananas. This streak of dominance will most assuredly come to an end, but it appears to be safe to say that his growing mastery of the knuckleball will continue to make him an odd yet effective pitcher. It is nice to see one of the games ‘good guys’ do well, and I, for one, am glad it is R.A. Dickey.