Anatomy of a player: Randy Johnson
by Josh KalkApril 29, 2008
Randy Johnson was at one point in his career the most dominant pitcher in baseball. He struck out everyone in sight and could throw 130-plus pitches in a game with seemingly no ill effects. Now Johnson is 44 and coming off an injury-marred season. Many thought he would retire instead of rehabbing to pitch this year and when he began the year on the DL many questioned if the Diamondbacks would get anything out of him at all again this year.
But, on April 14, Johnson came off the DL and started against the Giants in San Francisco. While he got a no-decision in that game and took the loss against San Diego six days later, he picked up his first win last Friday in Petco. Despite just a 1-1 record, he has an excellent 2.70 ERA and his FIP isn't much higher at 3.36. He has struck out 20 batters in just 16.2 innings, reminding many of the Big Unit of old. The big question is whether he can he keep this up all year or whether wear and tear will take its toll. A look at his PITCHf/x data might shed some light.
When looking at any data this early in the season, it is always good to start by going back to last year and seeing what has changed. Sadly, PITCHf/x didn't track a single pitch of Johnson's last year. So the 2008 data are all we have to look at. The problem with this is the 2008 data isn't great.
As was the case last year, researchers are either going to have to look at just a pitcher's home starts and deal with the imperfections or go through messy corrections to try to fix the data. Because we aren't even a month into the season, correcting the 2008 data doesn't seem very likely right now. Actually, I should say correcting the data, as I did last year, isn't feasible at this point but it is possible that another method could provide good corrections.
Getting back to Johnson, here are the data from his one home start and two road starts.
As you can see here, it would be hard to do any real analysis with this so I am going to implement a hack to try to get something useful. I am going to move the road data so the center of the road data matches the center of the home data. This involves moving the horizontal movement left 5.42 inches, the vertical movement down .49 inches and the speed down 1.68 MPHs. What this should provide is at least consistency in the data to tell us something about how Johnson is throwing. All of the numbers and plots that follow should be constant, but ultimately might be off by quite a bit (after all, the horizontal movement here is different by almost half a foot).
After cleaning up some obvious mistakes in the pitch identification, I end up with this plot for the movement of Johnson's pitches compared to a ball thrown without spin.
Now that is more like it. Johnson's fastball is now clocking in at almost exactly 90 mph, but still has excellent movement, just shy of 10 inches both horizontally and vertically. While we don't have any previous data from Johnson, the movement at least seems in the ballpark as his delivery has always been more three quarters than over the top. It is very possible that the velocity numbers here are low, but I would be surprised if Johnson was able to throw much harder than 92 mph for his average fastball. That said, it does appear he can dial this up and down as he likes, having thrown about 20 of his 155 fastballs at 94 mph or above. That ability to dig deep in key situations could be very important for him going forward.
Let's face it, though: When you think Randy Johnson you don't think of his fastball, no matter how good a pitch that is. You think of his slider, which basically has been unhittable at some points in his career. His slider has always been touted as a hard slider with excellent movement and faster than a league average slider. Now, Johnson's slider seems to be a shell of what it once was. It has slowed to nearly league average from 2007, and its movement has become league average or below. He still is throwing it a lot to left-handed batters, and it still can probably be an effective pitch for him in lefty-lefty confrontations because of his wide release point, but it probably won't fool nearly as many right-handed batters as it once did.
To compensate, Johnson has begun throwing a split-fingered fastball in the last few years. While he calls it that, it behaves almost exactly like a change-up. In fact, the pitch identification algorithm MLBAM has created to use the PITCHf/x data calls it a change. Whatever you want to call it, Johnson throws this pitch only to right-handed batters to keep them off his fastball. This isn't surprising—many left-handed pitchers don't throw their change-ups to left-handed batters.
As for Johnson's long-term prospects: If the data are at all reasonable here, I have to worry. Pitchers can get away with throwing only 90 mph, especially if they are left-handed, but most of them have strong breaking pitches to back that up. It appears that with the decline of Johnson's slider, that could be an issue for him. Also, most left handers who throw that fast have excellent control; Johnson have never been known for that. This season he already has walked nine batters, so if his strikeouts should fall, it could spell trouble.
Lastly, Johnson has faced the Giants and the Padres twice, and once each on the road in great pitchers parks. How will he fare against a lineup like the Phillies or the Cubs, especially in his home stadium? That said, there is a good chance that this is the last year Johnson will be plying his craft on the mound, so if you get a chance to catch one of his games I would highly recommend it. Maybe he can continue to spin his magic for a few more months and keep bringing back memories of how truly dominant he was in his prime.
Josh Kalk is a physics and math geek who can also be found blogging at http://www.baseball.bornbybits.com/blog/blog.html. He enjoys good conversations about baseball and can be reached at .(JavaScript must be enabled to view this email address).