Chris Volstad’s Problems: A Pitch f/x attempt


Before I try to diagnose Chris Volstad, let me diagnose you. You need a healthy dose of Marlin Maniac, which you can get daily right to whatever it is you use to read blogs by subscribing here. Doctor’s orders! Also, follow me on Twitter! I know, I know, I couldn’t help it.

WARNING: This is my first time even attempting to make Pitch f/x visualizations on my own. I might be terrible. They might come off poorly. But I’m trying dab nabit.

We all wish Chris Volstad was pitching a little better. After last season’s awesome start, the team was counting on him being the third wheel in the rotation, backing Josh Johnson and Ricky Nolasco. However, this year Volstad has really struggled, posting a FIP of 4.84. He’s been above replacement, about six runs better to be exact, but that is not what we wanted from our starter, especially one who posted a 2.88 ERA the previous year in 84 innings.

ERA? I know, Cro-Magnon like on my part. Yes, I did that on purpose to illustrate that Volstad’s 15 games (14 starts) were something of a mirage last year. Volstad actually posted a pretty good 3.82 FIP, but that value was not as a result of his 5.55 K/9 and 3.84 BB/9. Rather, his FIP was boosted by the fact that Volstad only gave up three home runs last season, yielding an impossibly low 3.9% HR/FB%. That number right there should scream “regression.”

Well, this season Volstad’s stuff has actually looked improved. He is striking out 6.11 guys per nine innings while walking only 2.8 per nine. His xFIP this season is at a solid 4.04, the expected value of a third or fourth starter on a good team. So what’s up with his FIP?

Like you needed me to tell you. xFIP normalizes for home runs, yielding the league HR/FB% (more or less) at the pitcher’s amount of fly balls for the home run value. Chris has allowed much more than that. On the season, Volstad has allowed an absurd 23 home runs, a HR/FB% of 16.1%! All of his other peripherals are mostly the same. He’s allowing 3% less groundballs on the year while allowing 5% more flyballs, which is contributing somewhat. But he’s still at 50% GB%, and the rate is so high that something must be wrong, right?

For this, I attempted my first try at using Pitch f/x to diagnose of pitchers. Using Dan Brook’s awesome Pitch f/x tool, I gathered up all of Volstad’s pitches from all of his starts and tinkered around with Excel. First off, let’s look at his pitches in general (not sure how useful this is, but it does show off that I can do this). Here’s a horizontal vs. vertical movement chart. (NOTE: I wish I knew how to watermark pictures into the graph area, because then you’d be seeing little pictures of Chris Volstad after home runs. Blah.)

From that graph (and from the labels on it) you can tell that Chris throws mostly a four-seam fastball with a changeup and a big curve. The big blob of blue is his fastball because it’s the pitch with the most “rise” against normal gravitational effects. The changeup has similar movement with a bit less rise and more “tail” towards righties, and the curveball is the blob of fuschia in the lower right hand corner of the graph, as it has more topspin and drops more than gravity would generally have it do.

For this piece, however, the horizontal movement may or may not be important. My hypothesis is that the pitches Volstad is giving up for home runs are “rising” a bit more than his normal pitches, so our focus should be on vertical movement. In addition, we should also take a look at velocity, as mistake pitches sometimes have poor velocity behind them as well. For that, we go to this graph.

This shows us his pitch distribution more clearly, as Volstad has good separation in terms of velocity on his pitches. You see his fastballs are clumped in the low 90’s, with his changeup in the low to mid 80’s, and a high 70’s curveball to top things off. FanGraphs’ Pitch f/x data confirms this; Volstad’s average fastball is 91.6 MPH, while his average changeup is 83.5 MPH, a good 8 MPH difference.

All right, let’s get to the meat of the problem, the home runs. Here’s the same previous chart for just his 23 home runs.

Obviously, even 23 home runs isn’t a large enough sample size to make a great determination. You can see that the concentrations of the pitches is more around the 9-10 inch break mark for his fastball, a few over 10 and a few under. The changeups are also on the high end, but there’s not enough of those to even come up with a concentration of points. Let’s focus on the fastball then. As a comparison point, here’s a graph of his four-seamers on groundouts and his four-seamers on the homers.

This graph was zoomed in to the fastball area so we could discern any differences. The spread doesn’t seem any different here than in anywhere else. The pitches for groundouts still are focused in the 9-10 region, though there are a lot more pitches underneath that region as well. The average vertical break on the groundouts was 8.8 ± 3.2 in 96 pitches, while the average on the home runs was 9.3  ± 3.6 in 18 pitches. In other words, there was a lot of variation, and these means aren’t particularly significant; they aren’t even one standard deviation from each other. FanGraphs has the average vertical break on Volstad’s four-seam fastball as 9.3 in 1325 fastballs thrown, identical to the fastballs for his home runs.

Obviously, this is still a small sample size, so there can’t be too much determined through this. I have a feeling, however, that the vertical break isn’t any higher for Volstad’s homers than they are for any other pitch. Thus, it might be a location issue, or it might be that hitters are getting good swings at pitches. I’d love to see a graph similar to the one Dave Allen did for Josh Johnson showing the distribution of Volstad’s fastballs vertically in the zone compared to the league norm, but I don’t think I’ll be able to do something like that (I’m just not that good yet).

In this little bit of work, we can guess that Volstad’s pitches are working just fine. It would seem that the issue is location, plus a good deal of bad luck. In other words, regression time, this time the other way. I hope.