Anatomy of a player: Brian Bannisterby Josh Kalk
April 15, 2008
The Royals' Brian Bannister is off to a tremendous start for the Kansas City Royals. I mentioned him in my last article, saying that not a lot of people thought he could repeat his excellent season last year. The main reason was that he wasn't striking out enough batters compared to the number he walked.
Last year, his ERA checked in at a cool 3.87 but his FIP was a more mediocre 4.52. This is mostly from his K/G being a mere 4.4 and his BB/G of 2.5. Now 2.5 walks per game isn't bad at all, but only 4.4 strikeouts generally is going to lead you into danger. So how has he turned it around this year so far?
Well, first he has kept his walk rate at 2.5 BB/G but has upped his strikeouts to 6.5 per game. That is a huge increase and while he probably shouldn't have an ERA under one, his FIP is an excellent 2.59 so far. His microscopic ERA of 0.86 has certainly been helped by a Defensive Efficiency Ratio of .825 behind him and a LOB percentage of 81.2 While both of these numbers will cool off, he still could be one of the better pitchers in the majors if he can keep the strikeouts coming.
So how has he increased those strikeouts? Well, let's take a look at his PITCHf/x data and see. Here are the raw data straight from Sportvision and MLBAM.
Wow, that is all over the map. MLBAM has him throwing four different fastballs, a slider and a curve. Clearly, a few of these are grouping errors. This isn't too surprising; analyst Dan Fox has mentioned they are still working out the kinks to the pitch classification system.
First, let's look at the splitters. While Bannister says he would like to throw a splitter, I haven't read or seen anything that makes me believe these are really split-fingered fastballs. They are more likely change-ups, as they are in the same location as the change-ups he was throwing last year.
Next, what about those cut fastballs? Bannister's four-seamer lives right on the edge of being a cut fastball, breaking only 1.84 inches towards a right-handed batter compared to a ball thrown without spin last year. A good cut fastball has about the same break, but in the opposite direction, so while I believe these are truly four-seam fastballs, a few had a little more (less?) break on them and were classified as cut fastballs. This is one of the problems with classifying pitches individually instead of grouping them together. Pitchers who throw a certain pitch that is on the border between two pitches will most likely be shown as having thrown some of each.
Next, let's look at the curveballs that have negative horizontal movement compared to a pitch thrown without spin. These are somewhat disconcerting pitches because they are moving like a lively screwball would move. Bannister didn't throw anything like that last year, and what is worse all came in his start in Detroit. While the PITCHf/x cameras are supposed to be calibrated better this year, it looks like maybe Detroit's cameras weren't having a great day. For now, I'd keep them considered as curveballs and assume that they really aren't breaking in to a right-handed batter.
Last, what about the group of fastballs way over to the left that are being classified as about half sinkers and half fastballs? This is a very interesting set of pitches and deserves closer attention.
First, it is important to note that while Bannister also didn't throw any pitches that looked like this last year, these appear to be real—he threw several of these pitches in each game he played. This makes it much less likely that they are due to camera error. So, assuming they are a new pitch, are they really two seamers (sinkers)? Last year in spring training, Bannister mentioned that he was working on a two-seamer, but my classifcation algorithm didn't find any two-seamers last year. It is possible he started using this pitch before the PITCHf/x cameras were installed and then scrapped it by the time they were turned on. I can't find any mention of him working on that pitch this spring ,though.
I don't think these are two-seamers. First, their vertical movement is around 10 inches, compared to a ball thrown without spin. While some pitchers do have large vertical movement with their two-seamers, it is rare. Second, he is throwing this pitch at 89 mph, nearly two mph faster than his regular fastball. Almost always, when a pitcher throws both a four-seamer and a two-seamer, the two-seamer is a couple of mph slower than the four-seamer. So if this pitch isn't a two-seamer, what is it?
I believe it is just his regular four-seam fastball thrown at a different arm angle. Here is Bannister's release point by pitch type. I have changed all the cutters to fastballs and splitters to change-ups, and I am listing all these new pitches as sinkers on this plot to separate them from the fastballs.
You have to look closely because they are being hidden by some change-ups, but all these new pitches are being thrown at an angle more side-armed than his regular release point. A more side-armed release point would impart more horizontal movement and less vertical movement on the pitch while keeping the speed about the same. This is exactly what we see here. So I think these pitches are four-seamers thrown at a different arm angle.
When we break the pitches up this way, we see that Bannister throws at this new arm angle early in the count, then uses the regular fastball much more later. In fact, on two-strike pitches he is throwing his regular fastball nearly 60 percent and many of them appear to be up in the zone. This is in contrast to last year, when he used his slider often to try to put away hitters. If this is his new strikeout pitch, it will be interesting to see if the league adjusts. If so, will he then readjust and go back to the slider?
Speaking of the slider: Both this year and last, Bannister has thrown a very hard slider—again, a pitch very close to being called a cut fastball. It is strange to see a pitcher who throws most of his fastballs in close relation to his sliders. Because both of these pitches are close to a cutter, the difference between Bannister's fastball and slider is much less than it is for most pitchers.
It is possible that this is why Bannister wasn't having much success putting away hitters with his slider. Even if they were fooled and guessed fastball, it wasn't too much of an adjustment to reach the slider. With the introduction of this new arm angle, the difference between this fastball and his slider is huge and something else for the hitters to be thinking about.