Monday, April 29, 2013

What has improved with Clay Buchholz?

Clay Buchholz has been the Red Sox best pitcher, and even edges out Dustin Pedroia as the Red Sox best player overall. He currently has a spectacular 1.19 ERA, and even more surprisingly, is striking out more than a batter per inning. While the ERA obviously won’t stay below 1.50, I thought it was worth a closer look to try and figure out where his strikeouts are coming from.

What Hasn’t Changed?
Before we get into the details too much, it is important to round up the usual suspects. Buchholz is not throwing a new pitch; he is still using the fastball, cutter, change up, and curveball that he has since introducing the cutter in 2011. Nor has he changed his approach. Buchholz still throws his four seam fastball about half the time and his cutter about 20% of the time, followed by his curveball and changeup with about equal frequency. Since Buchholz was hampered by injuries to start 2012, you might expect an increase in velocity, now that he is healthy, but we also don’t see a surge in velocity for Buchholz. In fact, Buchholz’s fastball is about one mile per hour slower than 2012. Buchholz’s control isn’t better either – he is still walking about the same number of batters (3.11/9 IP) and he is throwing the same number of first pitch strikes, too. Finally, and most perplexingly, we don’t see a change in his swinging strike rate.

Swinging Strikes and Strikeouts 
The big surge in strikeouts without an increase in swinging strikes is very odd. Swinging strikes and strikeout rates are tightly linked (R2 = 0.65 in 2012 for qualified pitchers), and swinging strike rate usually stabilizes before strikeout rate. Without an increase in swinging strikes, we should probably expect to see a big drop in Buchholz’s strikeout rate over the rest of the season. Needless to say, that would mean that he wouldn’t be the ace we’ve seen in the first month. Based on a linear regression from last season’s stats (it is still a bit early to be using 2013 stats), we should expect Buchholz to have a K/9 of 6.67, instead of 9.32. That strikeout rate would mean his expected ERA (xFIP) would be about 4.00 – hardly an ace, and maybe not even a #2.

A Silver Lining
There is one metric that Buchholz has improved significantly in – batters are swinging much less often at pitches in the zone this year (58%, compared to 65% in 2012). Currently, a 58% swing rate at pitches in the zone is the 10th best in the majors, and would have led the league last year. If batters aren’t swinging at pitches in the zone, that could mean two different things: 1. Batters are getting fooled by Buchholz’s pitchers, or 2. Buchholz is working the corners effectively, so although the pitches are in the zone, they are not good ones to hit. Looking at Buchholz’s heat maps showing where he is throwing pitches, nothing immediately jumps out at me to indicate that Buchholz is working corners better than previous years, however it is rather hard to compare the 2013 season to previous years because of the big difference in number of pitches thrown.

Trickier Pitches? 
If Buchholz is fooling batters, we might be able to see some improvement in the movement of his pitches. Thanks to PitchFX data, we have an amazing treasure trove of data. There is not a significant change in the amount of movement for Buchholz’s changeup or curveball, but he has added an inch of movement on his fastball and cutter (this combines both horizontal and vertical movement). Most of the movement for the cutter has been horizontal movement, which is not surprising as cutters often have more run when thrown slightly slower, and Buchholz is throwing his cutter two and a half miles per hour slower than last year. However, of the two pitches, only the four seam fastball has shown improvement. A big part of that improvement has been in the swing rate within the zone, as hitters have offered at fastballs in the zone 7% less often in 2013. The biggest improvement in that category is with the curveball; hitters have gone from swinging at curveballs in the zone 42% of the time, to an incredibly low 22% of the time. Although the overall movement of the curveball hasn’t changed, how it is moving has. Over his career, Buchholz has gone from having a fairly traditional 12-6 curveball (one that drops pretty much straight down, from 12 o’clock to 6 o’clock, if you were looking at a clock face) to one that has a lot more horizontal movement; at this point, he’s almost throwing a slurve (combination of a slider and a curveball), with over 10” of horizontal movement. Traditionally, most scouts prefer a 12-6 curveball, and a slurve often has a bad reputation, however, this pitch is working for Buchholz.

Looking Ahead 
Whether or not the slurve remains effective, particularly in getting called strikes, will be key for Buchholz’s continued success. I think a lot of that will depend on whether or not hitters adjust to the new, more horizontal movement of the pitch. If he can continue to fool hitters, he may be able to maintain an elevated strikeout rate, without having a typically high swinging strike rate. A regression to model strikeout rate using swinging strikeout rate and zone swing percentage indicated that a lower swing rate at pitches in the zone leads to a higher strikeout rate overall, and greatly improves the fit of the model (R2 = 0.83, compared to 0.65 with just swinging strike rate). According to that model, Buchholz’s K/9 IP would be about 7.50, leading to a much more palatable predicted ERA of about 3.50. Now that looks more like a #2 starter.


  1. Great article. Are there any pitchers who have posted high strikeout rates without high swinging strike rates throughout their career? It would be interesting to see what that group looks like.

  2. It depends on how high you define "high". I think the upside comparable is Doug Fister, who consistently is in the mid 50% range for in the zone swing rate and put up a 7.63 K/9 last year. CJ Wilson also has consistently outperformed his swinging strike rate, which is middling, while averaging 7.83 K/9 over the last three full seasons. I could run some more analysis to see who consistently falls into this group, but right now the model is only working with 2012 data, so I'm not sure how consistent this skill is (beyond the two pitchers mentioned above).

  3. Great analysis. I wonder if the Slurve pitch is better disguised than his previous 12-6. Not sure if you can compare release points from year to year using PitchFX, but that might be an explanation as to why so many more hitters are being fooled by the slurve. Not convinced that would explain all of the difference, but something to consider.

  4. This comment has been removed by the author.

  5. This comment has been removed by the author.

  6. Doug Thorburn at Baseball Prospectus has a nice piece on the mechanical changes for Buchholz. I'm no scout, but the general gist of it is that Buchholz has greatly improved his mechanics under John Farrell. Also, his mechanics are much more similar to 2010, his last good year.

    (Subscription required)