Start spreading the news... I'm leaving today...
Just as I imagine many of you guys are departing from 2013 Mariners fandom, checking out and rooting for prospects, I'm headed halfway to Mexico for a three-week research cruise. I'll be wading through oodles of data, getting seasick, and spending all my free time with mariners - so it'll be just like a normal month, except with more nausea. And the ever-present risk of falling overboard and being left behind, of course, but who worries about things like that? Not me!
Many of you may recall that I promised you a going-away present. Well, here it is!
It's an enormous Pitch F/X spreadsheet, containing every full-season pitch thrown by every major league pitcher with more than 30 IP during the Pitch F/X era. What that means is that it has the average velocity of Arthur Rhodes' fastball in 2011, and the average horizontal movement of Andy Pettitte's curve in 2010, and (most crucially) the average horizontal and vertical release points of A.J. Burnett's knuckle-curve in 2012. See, while the movement and velocity data is easy to get, I've never found a data set that has the release points all listed together. So, being an idiot, I set out to compile my own.
Why is this useful? Well, there's a crapload of stuff you could do with this data set - look for a correlation between horizontal release point and platoon split, look for links between release point and movement, and so on - but my initial goal was to build a pitch comp creator. Basically, when this thing's done, you'll be able to type in any pitch's velocity, release points, and movement, and the list'll spit out the twenty most similar pitches thrown from 2007-2012. Then you can look at the pitch values and the pitcher names and it'll give you some idea of how good the pitch is. If it spits out names like Armando Galarraga and Jeremy Bonderman, uh-oh. If it spits out names like Felix Hernandez and Clayton Kershaw, hooray!
So, even though the methodology there would be somewhat flawed due to the potential for midseason release point changes and deception and pitch mixing, I think it's a pretty cool project. That's the good news. The bad news? It's not done.
Turns out, in addition to the twin problems of me being incredibly irresponsible and me spending the last fortnight holding down one and a half other jobs, I ran into the difficulty that Pitch F/X data is a bitch to work with. Specifically, the Fangraphs classification algorithm and the Brooks Baseball classification algorithm don't agree. They confuse sliders and curves, they can't tell what's a changeup and what's a split, and they always get the fastballs wrong. Apparently Brooks Baseball doesn't believe in two-seam fastballs. Anyways, that meant I had to reclassify a lot of the data, which meant spending a lot of time. Correction: it meant spending a super lot of time, because I was already manually adding in the data from the Brooks Baseball pages. It took goddamn forever.
As a result, I got through, uh, the A's. The players whose first names start with "A". That's 5% of the spreadsheet. I'm sorry it's not more, but when I get back from the trip I should have a lot more free time to get this finished, and I promise you that it will get done. In the meantime, enjoy the current list of about 600 pitches - perhaps you can do some rough work with it. I don't know.
That was a long introduction, for an OffTop! I'll leave you, then, with the last questions of mine you'll see for three weeks:
1) Have you checked out yet? If not, what are you watching for? If so, what would it take to get you back?
2) With the release of Monsters University... what's your favorite Pixar film and why?
3) If you
had to wanted to abandon the Mariners and start bandwagoning for another team in 2013, which team would you choose and why?
4) Choose your own prompt!
Have fun, guys. See you on the other side.