One of the best advances in the statistical movement deals with the combination of stats and Pitch F/X technology, which measures strike zones, pitch types, movement, velocity, and of course, whether or not the batter swung, laid off, made contact, etc. This is utilized in a few different ways by different people; the people at FanGraphs present sweet stats for players' swing-tendencies and contact-tendencies, on pitches both inside and outside of the zone (awesome, awesome stuff). They can do the same for pitchers and their tendencies for making batters miss. Using pitch f/x data, pitch-types, tendencies, and velocities can be compiled for pitchers, and that's exactly what Josh Kalk does with his player cards. His blog and website absolutely rules, and if you don't frequent it, you should. He collects, corrects, and compiles Pitch F/X information and presents them in convenient player cards.
I recently utilized Baseball Prospectus Runs Prevented and VORP stats to compile a ranking of pitchers from rotation spot 1-5, utilizing only NL pitchers who pitched at least 100 innings in the NL. Today I finally got to a long-time project that I had been eyeing -- combining the quality rankings (VORP, Runs Prevented) and the raw stuff. Now, others do this in much more scientific ways, but I decided to do this in a pure ranking system that divided pitchers by three different types: those with a fastball above 91 MPH, those with a fastball below 91 MPH, and those who throw a sinker.
All of this can be discerned from Josh Kalk's player cards, and I used them heavily for this survey, but I should say that this is my interpretation of the data. This is not his interpretation, and he may well disagree with this usage if he saw it; I'm not sure. But, this is a combination of my pitching survey, and a division of pitchers by their prominent fastball / sinker type.
The number next to each pitcher is their ranking according to runs prevented from the previous survey, and here I ranked them twice: (1) by fastball / sinker type, and (2) by VORP.
I interpreted the horizontal and vertical movement ratings by Kalk's suggestions, so that the data makes sense, but I added my own personal twist. I compared only the pitcher's two most prominent pitches, to make comparison across the board more even (so we don't have to deal with extensive information for a guy like Dave Bush, who throws apprximately 5-6 pitches, versus a guy like Oliver Perez, who basically throws two pitches).
I attempted to incorporate colloquial language to present the pitch f/x rankings. Movement for pitch f/x is measured against spinless balls of the same velocity of the pitch that is tracked, which means that just about every fastball "rises," and even most sinkers "rise." To present this colloquially, I describe vertical rising movement as "riding," which is a term that I think evokes a mental image that properly describes what pitches that actually spin do -- that fastball by Ben Sheets doesn't actually rise 11 inches; it simply doesn't drop as much as a spinless pitch.
So rather than writing, "Ben Sheets' fastball rises 11" and moves in 4" on righties," I write, "Ben Sheets' fastball rides 11" and in 4" on righties." This is something our eyes can see when we watch Sheets' best fastballs; now instead of simply saying "Ben Sheets' fastball rides in on righties," we have the benefit of actually measuring the ball against another spinless ball, and finding coordinates of its motion -- and thus, "Ben Sheets' fastball rides 11" and in 4" on righties."
Now, some pitchers don't throw a fastball or sinker as their primary pitch. So, I simply used the slower pitch -- be it a curve (Wandy Rodriguez), a cutter (Doug Davis), a slider (Chris Sampson, Bronson Arroyo), or a change up (Edinson Volquez), among othets -- as their primary pitch, and then rate the fastball or sinker accordingly. This is sometimes counter-intuitive (I never would have guessed that Volquez throws the change more frequently than both of his fastballs), but it accurately reflects their pitch selection.
Pitch selection is what I wanted to reflect in this survey, not pure stuff. This comes up as an issue with some pitchers, like Carlos Zambrano, who throw several fastballs, but their fastest pitch is not their main fastball selection. Anyone that watches Zambrano knows he can throw the ball faster than 92 MPH, but that's not his main fastball that he selects most frequently.
I don't doubt that most of these major leaguers could easily throw the ball harder than they regularly pitch. Here I wanted to capture what the pitcher is getting at with the selection of his two most frequently selected pitches. In almost every case, these are the two pitches a pitcher throws for more than 50% of his total pitches, so it is important to know what he's throwing more than half the time -- and maybe then we can figure out why.
Now, I incorporated one more interpretation of my own, which is not included in Kalk's original data. I'm not even sure this is correct, but I do it anyway simply because i want to present the data in an intuitive way that reflects how we see the baseball. In judging the difference between the primary and secondary pitches and their differences in vertical data, I don't merely present both of their technical pitch f/x calculations. Instead I present the proper measurement of the primary pitch, and present the secondary pitch in comparison.
For instance, Ben Sheets' fastball "rises" almost 11" vertically, and his second pitch, the curveball, is thrown with so much spin that it actually drops almost 3" more than a spinless ball. Now, instead of simply presenting the 3" drop from the spinless ball, I present the difference between thw two pitches. So, if Sheets' fastball rides 11" through the zone, his curveball (which is actually spinning!), effectively breaks down over 13" compared to the fastball. This is quite a drop, which makes sense given that Sheets most often throws his curve so hard that it breaks into the dirt at a rather high velocity (over 80 MPH on average).
So, to present Ben Sheets, here's what I write:
7. Sheets, Mil (52.4 VORP; 198.3 IP, 74 R): 93.54 MPH, -12.98 MPH
Fastball (93.54) rides 10.76” and in 4.17” on righties
Curveball (80.56) breaks down 13.51” from fastball and 3.15” away from righties
Sheets' basic IP, R, and VORP stats are listed, as is his rating (#7) from the original survey. I also list the top speed velocity (93.54), and the velocity differential (-12.98) between primary and secondary pitches. Now, beware, top-speed velocity of the primary pitch does not always equal fastball velocity.
Check out Edinson Volquez:
12. Volquez, Cin (44.3 VORP; 196 IP, 82 R): 83.07 MPH, +10.97 MPH
Change (83.07) rides 1.83” and in 6.98” on righties
Fastball (94.04) rises 7.59” from change and in 4.47” on righties
Now, we certainly know that Volquez throws the ball harder than 83 MPH. But his primary pitch selection happens to be 83.07 MPH. I think this is incredibly telling about the way he pitches. His most frequently thrown fastball, then, is 94 MPH, almost 11 MPH faster than his change up, and it "rises" over 7" against the change up, which is thrown with less spin than his fastball.
I did not adjust the horizontal movement ratings at all because that contains crucial information about the actual trajectory of the pitch. For instance, while it makes sense to talk about how two pitches compare vertically, each pitch's horizontal trajectory has a more specific trajectory than simply rising or falling; for instance, while Volquez's change up breaks over 5" lower than his fastball, it makes a lot less sense to say that his change up breaks in 2" more than his fastball -- the horizontal trajectory is more important to capture in its original statement, rather than in comparison. This will make the most sense when you come across a pitcher, like Jeff Suppan, who throws a pitch that is fairly neutral in terms of horizontal movement, which means that its vertical movement is more pronounced:
59. Suppan, Mil (3.6 VORP; 177 IP, 110 R): 87.3 MPH, -3.98 MPH
Fastball (87.3) rides 9.56” and in 6.19” on righties
“Cutter” (83.32) breaks down 5.24” from fastball and in 2.45” on righties
Now, Suppan's main fastball is not a sinker (surprisingly), but his regular, "riding," cross-seam fastball. Thus, comparing a pitch that Kalk labels a "cutter" -- likely one of Suppan's various little pitches he throws in the low-to-mid-80s, we can see that the cutter does not ride as high as the fastball, meaning that its spin drops it into the zone lower than the fastball; but compare the horizontal movement. Both pitches move "in" on righties, but the cutter moves much less so. Instead of saying that the cutter drops 5.24" from the fastball and almost 4" away from righties, however, the proper trajectory is still maintained with the "in 2.45" on righties," because even though it breaks in less, it is not proper to say that the pitch breaks towards lefties. It might break less towards righties, but a pitch that breaks less towards righties is not the same as a pitch that breaks away from righties.
Thus, horizontal movement is a bit more complex to write about in a colloquial manner, because there is really only one way that a spinning pitch can vertically "break;" but the spin of a pitch can actually horizontally break a pitch in two different directions.
A fastball cannot break "up;" a rising fastball doesn't actually rise. But a fastball can break "in" or "out," depending upon its type. And that's why I presented the horizontal information in its original form, which Kalk corrected and presented on his player cards.
So, once again, this is my interpretation of his data, and it may or may not be correct, and he may or may not agree with it; even I admit that there are lots of issues that remain to write about, which is a good thing, because analyzing pitching is a ton of fun.
***
Pitch-ratings by rotation slot: Here are the basic results of the survey via rotation spot. Oddly enough, the 1st slot of the rotation, the best of the best, did indeed throw harder and produce a higher top-speed velocity. It's rather unfair to all the others, isn't it? I mean, Sabathia, Santana, Sheets, Lincecum, Billingsley, and Oswalt each finished in the top 16. And remember, this top speed average includes Volquez's change up, the pitch he throws frequently. So, even his 94 MPH is left by the wayside, and that pitch would raise the average even more, if we were not incorporating pitch selection into the survey.
But, I think this survey raises some unique questions: below the 1st pitchers, primary-pitch velocity and velocity differential between first and second pitches does very little to actually differentiate performance and results. For instance, compare 1st and 3rd spot pitchers, and 2nd and 3rd spot pitchers. Velocity differential and primary pitch velocity might suggest that the 3rd spot pitchers should be better -- their primary selections were thrown faster, and their velocity differential was greater. But now we have more reason to look at horizontal movement patterns, and vertical movement patterns: I'd suggest that, rather intuitively, it matters much more how a pitcher is spinning the ball, than the velocity or velocity differential. And those main exhibits produced great seasons in the 2nd rotation spot: cf. Jamie Moyer, Ted Lilly, Jorge Campillo, Wandy Rodriguez, Carlos Zambrano, Aaron Cook, and Hiroki Kuroda....
First Starters: 498 G, 490 GS 3217.7 IP, 1214 R
First starters: 3.40 runs average, 6.46 IP/G (37.0 to 73.4 VORP range)
Avg: 91.59 MPH top speed; -8.09 MPH differential
Second Starters: 488 G, 474 GS 2895.7 IP, 1318 R
Second starters: 4.10 runs average, 5.93 IP/G (17.6 to 43.2 VORP)
Avg: 89.63 MPH top speed; -4.89 MPH differential
Third Starters: 463 G, 433 GS 2607 IP, 1366 R
Third starters: 4.72 runs average, 5.63 IP/G (7.7 to 26.5 VORP range)
Avg: 90.10 MPH top speed; -8.89 MPH differential
Fourth Starters: 465 G, 440 GS 2370.7 IP, 1448 R
Fifth starters/swingmen: 337 G, 208 GS 1304.3 IP, 827 R
4th / 5th Starters: 802 G, 648 GS 3675 IP, 2275 R
5.57 runs average, 4.58 IP/G (-12.4 to 19.9 VORP range)
Avg: 89.25 MPH top speed; -5.88 MPH differential
More on the way...