Although there are many repertoires among starting pitchers in Major League Baseball, one aspect remains a constant among them, and that is the presence of some form of fastball. The fastball is the pitch that sets up all other pitches by virtue of contrast in velocity and movement. A curveball would become much easier to hit if the hitter knows he doesn't have to deal with a fastball: he can just sit back and let the pitch come to him. The same goes for the changeup. The vertical and horizontal movement of a slider (or cutter) are much less impressive if they're not contrasted with a fastball that has spin carrying it in the opposite direction.
So it seems that any good pitcher in the majors needs to at least have a good fastball, because otherwise, hitters could just try to ignore the other pitches and hit the bad fastball, right? Depending on your definition of a good fastball, it doesn't seem to be that accurate. There are some pitchers out there who succeed, despite throwing a fastball that gets hit hard. And since I find those pitchers fascinating, I decided to write a piece about (some of) them.
Bronson Arroyo
Bronson Arroyo is a fascinating pitcher for various reasons. One: his incredibly high leg kick. Two: his low career BABIP (.282). Three: the variety of pitches he seems to throw and the year-to-year difference in what Bronson seems to throw (I can't make heads or tails of his PitchF/x data), and four: the big difference in quality between his fastballs and his breaking balls. While it's kind of hard to determine what Arroyo throws exactly, he definitely throws a sinker and a four-seamer, possibly a cutter, one or two kinds of (split?)changeup, two variations of a curveball, and possibly a pitch that is somewhere between a splitter and a slider. None of those pitches are whiff-machines, but the curve/slider(s) combination is highly rated by Fangraphs' pitch values.
Arroyo's fastball has average approximately 88 miles per hour over his career, but dropped to 87 in 2011, coinciding with a sharp decline in effectiveness, mostly due to a very large number of home runs given up. He's thrown the fastball (sinker and four-seam combined) about 45% of the time, but only 39% in 2010, which was interestingly also the one season it ranked as a positive by Fangraphs' pitch values. In that year, his four-seam fastball had a contact% four percent lower than in 2011, and a much lower groundball rate (probably positive for a four-seamer) as well. Overall, his fastball has a contact rate that fluctuates between 87 and 92 percent, while throwing his fastball for strikes in the 62-63% range (64.5 is league average).
James Shields
Shields' was the polar opposite of Arroyo in 2011. Coming off a horrible season with an ERA above 5, Shields reduced his fastball percentage from 46.1 down all the way to 36.4%. This drastic change probably helped put up his best season, with a career low 2.82 ERA and a career high 23.1% strikeouts. Pitch values rated his Shields' fastball as above average for the first time in his career, despite it being 0.5 mph slower than it was in 2010. Hitters made contact with it more than 91% of the time, up from below 90% in 2010, but unlike in 2010 the pitch didn't get deposited over the fence much.
Aside from the bad fastball, Shields isn't much like Arroyo at all: his sinker plays a marginal role, his curve gets groundballs (unlike Arroyo's crazy horizontally moving, flyball-inducing offering), his slider is nothing special, but his changeup is one of the best there is. Shields actually uses his fastball in the traditional way: to get strikes and get ahead of the batter so that the batter can be finished off with an offspeed offering. This seems to have changed drastically in 2011, with Shields using much more unpredictable pitch sequences. But while Shields improved in almost every way, he still gave up a relatively high number of home runs per flyball, as he has done over his whole career. Will the league adapt to Shields' changes and will the Rays regret not trading Shields at peak value?
Brett Myers
Myers also has a bad fastball (a combination of two fastballs actually) at 88-89 mph, but he has a great curve to make up for it, while his slider is average and his changeup is below average. He has a penchant for giving up home runs, even more so than Shields. Also like Shields, he recently reduced his fastball% drastically, coinciding with his best season so far (2010, 3.14 ERA), but dropped off the season after, during which he threw the fastball at a normal rate again. Much like the others, his fastball is made contact with in the 89-91% range, and Myers is neither very predictable nor very unpredictable in his pitch selection. The Astros might regret not cashing in on the man's 2010 season, assuming there could've been a decent trade out there somewhere.
A.J. Burnett
Burnett is the first hard throwing pitcher on the list. His fastball used to be good, but has gone severely downhill starting in his age 31 season and coinciding with the start of a trend of peripherals-related underperformance. He did try to stop the trend by cutting his fastball% by 13% in 2011, but while it did improve his xFIP from 4.49 to 3.86, his HR/FB% spiked and led to a 5.15 ERA. A 1.5 mph drop in average fastball velocity over the last two year probably hasn't helped. Contact rates of 88-90% on his still pretty hard (92-93) fastball are odd, and worrisome. For Yankees fans, anyway. His main non-fastball pitch, the curve, is pretty easy to predict: it won't be there when Burnett's behind in the count, but you can count on a curve coming your way in a strikeout situation.
Edwin Jackson
Bad fastball, pretty bad changeup, great slider. Compiling a scouting report on Edwin Jackson is much easier than compiling one on Arroyo, I can tell you that. Even though Jackson's fastball averages over 94 mph, he has cumulated a lot of negative runs on the fastball over his career. Unlike the others in this article he does not have a home run problem, but he has underperformed his peripherals slightly over his career, mostly due to a relatively high BABIP. Jackson has cut down on using his fastball by 10% over the last two seasons, improving his advanced stats slightly but not his ERA (due to high BABIPs). Jackson doesn't exactly pitch backwards, but he's much less predictable in his pitch selection than Burnett. His fastballs have contact rates in the 88-90% range (how unique!).
Conclusion?
I will probably write a "sequel" to this article if it's well-received, and until then any conclusions may be a bit premature. It seems like there might be some correlation between pitch predictability and underperforming peripherals, and between fastball quality and HR/FB% (even if Edwin Jackson didn't fit the description). It also seems like a bad four-seam fastball is made contact with a lot, more than 87%. Sinkers, of course, are even more geared towards inducing contact and less towards missing bats, so they probably should be evaluated differently. Let me know what you think!