Note: This article was originally published at the Statistically Speaking blog at on February 26, 2008.  Since the site is defunct and its articles are no longer available on the web, I am re-publishing the article here.

In Part 1 of this analysis, we examined the league numbers for batting average on balls in play (BABIP) and whether Bannister was able to beat the league BABIP by pitching in favorable counts. We found that he did not gain any particular advantage by inducing more balls in play on two-strike counts, so we turn elsewhere to seek an explanation for his 2007 performance.

Part 1
Part 2
Part 3

What pitches does Brian Bannister throw? The scouting reports tell an interesting tale, especially if you follow them back a couple years. In the minor leagues, the cut fastball was reputed to be his best pitch. His four-seam fastball was thrown in the high 80’s, touching 90, although he was able to locate it well, his curveball was a big breaker that was considered a plus pitch, his changeup was a work in progress, and his slider was regarded as a pitch likely to be scrapped. But in the fall of 2006 in the Mexican League, Bannister worked on a two-seam fastball, and after joining the Royals in trade for Ambiorix Burgos, he scrapped his cutter, experimented with different speeds on his curveball, and started throwing a slider again.

What can we see in the PITCHf/x data regarding his pitch repertoire in 2007?

As far as I can tell, Bannister basically threw four pitches in 2007. I couldn’t find much evidence of a cutter, so I believe he didn’t throw it very often. I couldn’t tell his four-seam and two-seam fastballs apart, so I’ve lumped them together as fastballs. Often, I can see evidence of two clusters when pitchers throw both types of fastballs, even if the exact delineation is hard to make. In Bannister’s case, I didn’t even see two separate clusters. In addition to the fastball, he threw a changeup, slider, and two varieties of curveball. Let’s examine these pitches in more detail.

Bannister’s fastball runs 87-91 mph, and the average spin deflection he gets on the fastball is a 9-inch hop and a 1-inch tail in toward right-handers. Compared to a league-average fastball, that’s 2 mph slower and with about 5 inches less lateral movement. The fastball is Bannister’s main pitch to both lefties (52% of the time) and righties (56%).

His changeup runs 82-86 mph; the five-mph separation between his fastball and changeup is quite a bit less than the league average separation of nine mph between those pitches. The average spin deflection on his changeup is a 2-inch hop and a 6-inch tail in toward right-handers. Bannister uses the changeup fairly often to lefties (19% of the time), but it’s his fourth pitch to righties (only 9%).

Bannister’s slider runs 84-87 mph, and the average spin deflection he gets on the slider is a 3-inch hop and a 3-inch break away from right-handers. That’s a couple inches more lateral deflection than a league-average slider. He uses the slider almost equally to lefties (15% of the time) and righties (20%). It’s somewhat unusual to see a right-handed pitcher throw the slider that often to lefties. Reference this quote from Bannister after his June 29 start against the White Sox:

Bannister rarely faced any trouble the rest of the way. He entered the night with a .299 average versus left-handers, 83 points higher than righties. The right-hander’s goal was to equalize those numbers, and he incorporated a new slider to great effect.

“It’s a new pitch,” Bannister said. “I used to throw a lot of cutters. I have been working on it all year, and tonight, I threw it a lot more. … Mac has really worked with me on things and I think the slider really helped, especially against the left-handers.”

Bannister’s fourth pitch is a curveball, and it also presents something I haven’t seen much with other pitchers. For most of the year, Bannister threw a slow curveball, running 70-78 mph, and there’s some indication he was even purposefully varying speeds within that range. Then, in his August 17 start against Oakland, he changed curveballs, unveiling a harder curve running 77-82 mph, with about five inches less drop than his curveball from earlier in the year. It seems to have been a complete switch from one to the other, since I don’t see any game in which he threw both varieties of curveball. Overall, Bannister threw the curveball 14% of the time to both lefties and righties.

Here’s another graph showing the movement due to the forces of spin deflection and gravity on his pitches in the last quarter-second before they cross the plate, and here are a couple other ways to look at the vertical vs. horizontal deflection over the whole pitch trajectory:

Next, let’s look at how Bannister mixes his pitches in different ball-strike counts, which I’ve split out by batter handedness. Bannister is more consistent than most pitchers about throwing all of his pitches across all counts, but he still shows some tendencies by count.

To righties, he likes to start out with a fastball or an occasional slider to get ahead of the hitter; both are pitches he can throw for strikes. The changeup features most often at 1-strike counts (17%), compared to only 6% of pitches with 0 or 2 strikes. Like most pitchers, he’ll throw the curveball when ahead (24%) but rarely when he’s behind in the count (3%).

To lefties, he also likes to start out with the fastball, and he’ll throw the curveball at 0-0 (17%) or when ahead in the count (26%) but not when behind (4%). By contrast, the slider rarely gets thrown at 0-0 or 1-0 (5%), but he likes to throw it with 1 or 2 strikes (23%).

He doesn’t seem to have a particular favorite strikeout pitch against either righties or lefties. The fact that he doesn’t have one great pitch to rely on for strikeouts may go a long way toward explaining why he doesn’t get that many of them. Along those lines, let’s examine the results that Bannister gets with each of his pitches. I’ve split the curveball into the slow curveball (Curveball1) and the late-season hard curveball (Curveball2) since he got quite different results from the two.

LHH #Pitches _Ball_ _CS_ _Foul_ _SS_ InPlay _Avg_ _BABIP_ _SLG_ _HR_
Fastball 354 0.37 0.30 0.13 0.02 0.18 0.338 0.295 0.615 0.062
Slider 105 0.32 0.08 0.27 0.10 0.24 0.200 0.200 0.320 0.000
Changeup 132 0.38 0.16 0.15 0.05 0.26 0.235 0.212 0.471 0.029
Curveball1 35 0.71 0.09 0.03 0.00 0.17 0.333 0.333 0.500 0.000
Curveball2 58 0.36 0.17 0.14 0.07 0.26 0.467 0.467 0.600 0.000
RHH #Pitches _Ball_ _CS_ _Foul_ _SS_ InPlay _Avg_ _BABIP_ _SLG_ _HR_
Fastball 350 0.34 0.21 0.19 0.05 0.21 0.192 0.169 0.356 0.027
Slider 127 0.32 0.13 0.18 0.06 0.30 0.368 0.314 0.632 0.079
Changeup 58 0.45 0.03 0.33 0.02 0.17 0.300 0.300 0.400 0.000
Curveball1 42 0.40 0.17 0.14 0.12 0.17 0.000 0.000 0.000 0.000
Curveball2 43 0.40 0.12 0.14 0.09 0.26 0.182 0.182 0.182 0.000
Lg. Avg. ——— _Ball_ _CS_ _Foul_ _SS_ InPlay _Avg_ _BABIP_ _SLG_ _HR_
Fastball —— 0.36 0.19 0.19 0.06 0.19 0.330 0.304 0.521 0.037
Slider —— 0.36 0.14 0.17 0.13 0.20 0.310 0.286 0.481 0.033
Changeup —— 0.40 0.11 0.14 0.13 0.21 0.319 0.295 0.502 0.035
Curveball —— 0.40 0.19 0.13 0.11 0.16 0.310 0.290 0.471 0.029

A good way to give an overall grade to Bannister’s pitches is to use the linear weights run values like we did earlier. In this way it’s easy to see that the fastball is Bannister’s best pitch; with it he can shut down righties and hold his own against lefties. The slider is a strong pitch against lefties but weak against righties, and the curveball is the opposite.

Pitch Total runs Lwts/pitch LHH RHH
Fastball -6.4 -0.009 +0.004 -0.022
Slider +0.6 +0.003 -0.036 +0.034
Changeup +0.8 +0.004 +0.002 +0.009
Curveball1 +0.1 +0.001 +0.044 -0.036
Curveball2 -0.1 -0.001 +0.023 -0.034
Total -5.1 -0.004 +0.001 -0.009

The league average information comes from John Walsh’s article. Now, let’s look at how Bannister locates his pitches and in more detail at his results.

Against lefties, Bannister pounds the outside part of the plate, getting a lot of called strikes and some foul balls on the outside edge. When he misses down out of the zone or too far outside off the plate, lefties rarely swing and miss. When he gets his fastball over the inside half of the plate, lefties make good contact. They hit .338/.615 (avg/slg) on balls in play, compared to league average of .330/.521. In the 65 fastballs to lefties tracked by PITCHf/x that were put in play, 4 of them went for home runs.

Against righties, Bannister lives in the zone and gets surprisingly good results for an 89-mph fastball. His called and swinging strikes and foul ball rates are all about normal. However, when righties put the fastball in play, they hit only .192/.356. We’ll examine this in more detail later since it seems to be a large component of Bannister’s success.

Bannister throws the slider for strikes to both lefties and righties, throwing 68% strikes compared to league average of 64% for the slider. He pitches it down and in to lefties, and when he gets it right on the corner, he has a good shot at get a swing and a miss. Otherwise, it’s likely to be fouled off, 27% of the time by lefties, compared to 17% league average. When lefties do put the slider in play, mostly groundballs, they hit a weak .200/.320, compared to .310/.481 off a league average slider.

Righties fare better against the slider. He doesn’t get many swings and misses nor as many foul balls. In fact, righties put the slider in play 30% of the time. Bannister works down and away a lot, but he doesn’t look terribly effective there. He either misses the strike zone for a ball or gets a little bit of the plate and lets the batter make contact. Righties hit .368/.632 against the slider when they put it in play, including 3 home runs out of 38 balls in play in the PITCHf/x data set.

Bannister has a harder time throwing the changeup for strikes than his other pitches, getting a strike 60% of the time, which is in line with league average for the change. To lefties, he works away and down. He doesn’t get many swings and misses. He gets some called strikes on the outside part of the plate. When lefties get the ball in play, they hit .235/.471, compared to league average against the changeup of .319/.502. The change is only an average pitch for him to lefties because he can’t generate more whiffs with it.

To righties, Bannister works the changeup inside, generating a lot of foul balls, and down, mostly below the knees for a ball. Those two events cover 78% of his changeups. He doesn’t get many strikes, either called or swinging, only a mere 5%, compared to a league average of 24%. That’s what makes this pitch somewhat below average for him because when hitters put it in play, they hit .300/.400, which is slightly better than average.

During the first two thirds of the year, Bannister used a slower curveball, and he almost never threw it for a strike to lefties, staying well down or well away from the zone, for 71% called balls. He also threw it down off the outside corner for a ball to righties fairly often, but he was more willing to get in the zone for a strike to them, and he had fairly good results when he did, including holding opponents hitless on 7 balls in play. The slow curveball grades out as a plus pitch to righties and minus pitch to lefties, but it’s in a small sample and mostly has to do with hitting the strike zone or not.

During the final third of the year, Bannister switched to a harder curveball and was more willing to challenge the hitters with it, particularly lefties, who went from seeing 29% strikes with the slow curve to 64% strikes with the harder curve. He still worked down but now was willing to come across the outer half of the plate with the curve. It garnered him a few swings and misses, whereas no lefty whiffed on the slow curve. However, it didn’t end up being any improvement for Bannister since lefties pounded the new curveball for 7 hits in 15 balls in play. Again, it’s hard to say with the small sample size whether the ball in play results for the curve have any significance.

To righties, the pitching pattern with the hard curve was similar to that with the slow curve. Again, they didn’t have a whole lot of luck on balls in play, netting 2 singles in 11 attempts, mostly off pitches on the outside edge. Once again, small sample size prevents us from drawing strong conclusions.

Combining the ball in play performance for the two types of curveballs yields slightly bigger samples–righties hit 2 singles in 18 tries, lefties knock 6 singles and 3 doubles in 21 attempts. I’m still not sure if the curve is really more effective against righties; in any case, Bannister doesn’t throw it often enough to make a huge impact on the overall ball-in-play numbers we are investigating.

Next, let’s look at Bannister’s pitch sequencing. Here’s a table showing what pitch a hitter is most likely to see from him based on what the previous pitch was.

Previous Pitch Fastball Slider Changeup Curveball
Fastball 48% 13% 22% 17%
Slider 59% 18% 18% 6%
Changeup 67% 13% 13% 8%
Curveball 46% 16% 16% 22%
Previous Pitch Fastball Slider Changeup Curveball
Fastball 56% 19% 12% 13%
Slider 60% 22% 4% 14%
Changeup 56% 13% 15% 15%
Curveball 60% 18% 6% 15%

To righties, Bannister doesn’t have many discernible patterns in how he mixes his pitches, although he doesn’t throw the changeup to them much following a breaking ball. To lefties, there are a couple more noticeable patterns. He likes to follow a changeup with a fastball, and he doesn’t throw his curve much after a slider or changeup.

Now that we know a bit more about Bannister’s repertoire and how he employs it, we’re ready to move on to Part 3 of the series, which will examine the reasons Bannister’s 2007 performance on BABIP was better than the league and whether we can realistically expect those things to be repeated in the future.

Part 1
Part 2
Part 3