The Phillies best pitcher, Cole Hamels, made his first start of the season on Wednesday night. The Phillies normally mighty offense couldn’t muster a runner past first base, and so Hamels was hung with a loss, despite pitching well.
By the numbers:
5 hits (1 HR), 2 BB
From those numbers, we can say that Hamels pitched a really nice game. The 6 strikeouts is slightly below his career strikeout rate, but not much below, and the skimpy walk total is par for the course. As is the home run, for the matter, as Hamels allowed 25 homers in 28 starts last year.
Major League Baseball, last year, began a system of pitch tracking that they call Pitch F/x. This system uses a number of high speed cameras positioned around the ball park to follow the path of the ball from the pitchers hand to the catcher’s mitt (or the hitter’s bat), and calculate the ball’s speed and trajectory. The system has been installed in every ballpark, I believe, and many of the kinks that existed last year have been ironed out. Really, this is one of the biggest advances in baseball analysis in years, as it allows, essentially, for rigorous objective analysis of performance in a way that can only normally be done by the subjective measure of scouts. While in the past, a pitcher might be categorized as having a huge curveball, we can now say exactly how big it is. A hitter might be said to be a mistake hitter–that is, he only makes solid contact when the pitcher accidentally leaves a pitch in the middle of the zone, or if a slider doesn’t, ya know, slide.
What we’re going to do with it is, well, look at some pitchers. Probably mostly just Phillie pitchers, but we’ll see how good we get with the data, and how it all strikes our fancy.
Anyway, and now what you’ve all been waiting for: CHARTS!
It is said (“Who says it?”, “They.”, “They talk a lot, don’t they?”) that a pitcher needs four things to succeed. Namely: velocity, movement, location, and deception. With Pitch F/x, we can talk about the first three things. Velocity, of course, is the most famous of all these, as every baseball fan can name their favorite flamethrower, whether it’s someone like Joel Zumaya recently, or Roger Clemens, Nolan Ryan, or going back a bit farther, someone like Sudden Sam McDowell or Walter Johnson. Flamethrowers get all the love. If you can throw 95, some baseball team is going to want you.
Movement, of course, refers the “break” of the pitch. When Bad News Bears’ ace Amanda Whurlitzer tells you that her curveball breaks 2 feet, she’s talking about the kind of movement she gets on the ball. The most common way to discuss movement is to imagine a straight line from the pitcher’s hand to the cather’s glove. Whatever sort of deviation from this line the ball makes on its actual path is, often, considered the “break”. This isn’t necessarily the best way to think about pitch movement, but we’ll talk about that in a moment.
Location is simply where the ball is when it crosses the front of home plate. The plate is merely 17 inches wide, and your average batter’s strikezone starts about 19 inches above the ground, and extends another 25 inches from there. The upper limit of the strike zone is just over 3 and 1/2 feet up. Throwing a pitch on the corner, low and away, is a strike. Throwing a pitch straight down the middle is a strike as well, but those pitches are bit more likely to end up near the outfield wall. As everyone’s favorite soft tosser, Greg Maddux, once said, “You can do a lot of things when you put the ball exactly where you want it.”
Deception is the one thing we can’t really measure with Pitch F/x, though it is about as important as the other three. Most pitchers, if not all pitchers, involve a certain amount of deception in their delivery of the ball. Whether it’s using their body or their glove to hide the ball before releasing it, or using a slow windup to screw up a hitter’s timing, deception is certainly important. Unfortunately, at this time, we can’t really talk about it.
Let’s look at location first. This chart is easily the most difficult to look at of the three we’ve got, so, don’t worry, if you make it past here, it’s smooth sailing.
What you’re looking at there, is the location of nearly every pitch Hamels threw on Wednesday (a few pitches weren’t documented by the Pitch F/x system), as seen by the catcher. The black square in the middle represents the approximate strike zone.
We can see that Hamels was a bit on the wild side on Wednesday, which is also indicated by the fact that 12 of the 32 batters Hamels faced worked at-bats of at least 4 pitches. Other things we can learn are that Hamels tends to throw his fastball up in the strike zone (despite the wildness, most of the blue dots are in the upper half of the chart), he keeps his changeup low in the zone, and he appears to have little idea where his curveball is going to go. High, low, inside, outside, that curve could be going anywhere. On the upside, Hamels threw a lot of curveballs on Wednesday (19 of them), which is encouraging, since at times last year, he seemed to go throw games where he wouldn’t throw more than a couple of curveballs. Curveballs (and other breaking pitches) are often referred to as “feel” pitches, because the pitcher must grip the ball somewhat delicately (to keep the speed of the pitch down), but apply substantial fingertip pressure on the ball to get the kind of spin neccessary to create break. The traditional wisdom is, basically, practice makes perfect, and so the more Hamels throws the pitch, the more likely that he’ll start to harness it a bit better.
Let’s take a look at pitch speed now:
Unfortunately, I’m not sure how to get Excel to create a scattergraph with a line of best fit, so you’ll have to make do with my squiggly lines for the time being. The vertical axis here indicates pitch speed, while the horizontal axis refers to time. That’s what the (blue) line for the fastball is slightly longer on both ends than either of the other lines: Hamels’ first and last pitches were fastballs. Things to note: Hamels fastball was greatly exaggerated when he first came up. First, Hamels was reported to throw in the low to mid 90s, then just the low 90s. The truth is the Hamels tends to throw in the mid to upper 80s. Or, at the very least, that’s how he was throwing on Wednesday night. Personally, I don’t have a problem with this, though I’d be curious to know how many other pitchers are as successful as the Phils’ ace with a fastball as slow as his.
Also worth looking at is the separation between Hamels’ fastball and changeup. Now, the idea behind the changeup is that, coming out of the pitcher’s hand, the pitch looks like a fastball. The hope is that the pitcher can do everything just as he does with a fastball–same delivery, arm speed, arm angle, and release point–but pitch the ball at least 7 mph slower than the heat. Hamels is outstanding in this regard, as his changeup often sits 10-12 mph slower than the fastball.
Hamels curveball is interesting. It’s just a smidgen slower than his changeup, but it breaks significantly more, so I wonder how it appears to a hitter. When Hamels can throw it for strikes, especially, it appears to be quite the formidable third pitch.
A few facts:
Fastest FB: 88.5
Slowest FB: 83.1
Average FB: 85.6
Fastest CH: 77.9
Slowest CH: 74.4
Average CH: 76.3
Fastest CB: 74.8
Slowest CB: 69.8
Average CB: 71.8
Now, getting on to movement. This is an easy graph to look at, but it’s somewhat difficult to understand. A quick primer: as we previously mentioned, people intuitively talk about the break of a pitch thrown relative to a straight line, drawn from the pitcher’s hand to the catcher’s mitt. This line, however, isn’t realistic. A ball cannot be thrown straight like that, as the effects of wind resistence and gravity cause the ball to sink, and, to a lesser degree, move left or right. The amount of spin put on a ball (backspin on fastballs, front and sidespin on various breaking pitches) cause the ball to drop and rise, cut and tail.
The graph below shows the break of pitches relative to a pitch thrown with no spin whatsoever. That mean, essentially, a ball that is thrown and the only action on it is caused be gravity. As you can see, pitches thrown with backspin resist gravity a bit–thus, a fastball with a substantial amount of backspin appears to “rise”. The ball isn’t really rising, of course, it’s just sinking less than expected. As for the horizontal deviation from the center of the graph, this results in tailing or curving action. On this graph, the pitches that are shown to the left of center are pitches that bore in on a righthanded batter (and away from a lefty), whereas the pitches to the right of center move away from righties and in toward lefties.
It’s perhaps worth noting that how crazy some of this is. Toronto Blue Jays blog recently took a look at RHP AJ Burnett’s first start of the year, and provided a graph much like this one. You can find it here. Burnett throws the same three pitches as Hamels. Both Hamels and Burnett throws a fastball that rises about 10 inches, and Burnett’s cuts substantially, moving back toward a righthanded batter between 5 and 10 inches. Hamels doesn’t get that kind of lateral movement, but the movement he does get is more interesting, in a way. Because, you see, the ball can go anywhere. Only 20% of his fastballs behave the same way Burnett’s do–that is, by tailing back toward a lefthanded batter (again, Burnett’s pitch is the same because, he’s a RHP with a FB that tailed toward a righty batter). The other 80% actually move away from lefties. Now, I’m not sure how he accomplishes this. If he’s throwing a different fastball, it only shows up in the movement. Average speed of fastballs that head left: 85.6 mph. Average speed of fastballs that head right: 85.6. I haven’t looked at Hamels’ pitch data beyond this start, but it’s very interesting (to me, at least), that he can make a baseball go both ways. If I get the chance, I’ll try to figure out if he’s doing this on purpose–he may have a tendency to throw the one kind of FB to lefties, and the other to righties.
Comparing Burnett and Hamels, again, we can see that Burnett throws an occasional changeup that moves back toward righthanded batters quite a bit, but doesn’t rise or drop much at all. Hamels throws his change frequently (about 30% of his pitches on Wednesday were changes), and while his doesn’t move laterally quite as much as Burnett’s, it rises quite a bit more.
And now the curveball. Burnett’s drops between 6 and 11 inches, it would seem, and curls away from a righthanded batter about 1-4 inches. Hamels? Well, he throws some kind of gem. The pitch sinks about 4 to 8 inches, but can move laterally, away from a lefty, anywhere from 1 to 7 inches. In fact, nearly half of his curveballs broke more than 5 inches away from a lefty bat. That’s a nasty pitch, it would seem.
There’s a ton of information available here, and I’m just learning to manipulate it. We’ll try to keep using it throughout the season, though I doubt we’ll regularly post things as lengthy as this. Generally, I suspect, we’ll use Pitch F/x to illuminate only one aspect of a player’s performance: just the movement or location or speed of a pitcher on a given night, or to see what, exactly, Shane Victorino is able to hit, if anything.