Imagine you were playing around on Brooks Baseball – as one does – and decided to check in on some Jake Arrieta data after his solid season debut. You think, “You know, it looked like he was commanding his stuff well and batters weren’t all over him, so I wonder if those periodically low radar readings on his fastball were just an in-stadium glitch.” So you pull up the old chart, and you OH GOD WHAT:
That’s what a nightmare looks like in chart form. So, should we assume the fetal position?
Let’s take a step back and sort this out as best we can.
Attendant to yesterday’s discussion about Arrieta’s start – which, again, looked good and featured solid results – you may have seen a number of side conversations in the comments about Arrieta’s velocity, and then Jeff Sullivan at FanGraphs took up the case, asking boldly, “What on Earth Happened With Jake Arrieta?” As you can see from the chart above (and the PITCHf/x charts at FanGraphs register the same drop), Arrieta’s velocity from his start in St. Louis recorded as some 2.5 to 4mph lower than last year, across the board on all of his pitches. That’s the kind of drop that causes panic in the streets, and causes pundits to ask things like “What on Earth Happened With Jake Arrieta”.
The question here is how concerned should we be, given the usual caveats about a single data point. Well, I probably won’t give you a satisfying answer, because I’m going to end up on: I’m not freaking out just yet, but I’m watching.
The primary reason for my relative chill-ness is that we know velocity recording was mucked up in a couple ways on the night of Arrieta’s start. For one thing, the PITCHf/x system did not track all of Arrieta’s (or Adam Wainwright’s) pitches. For another thing, Statcast (which uses a different tracking system) also appears to have misclassified a number of pitches – unless Arrieta really was throwing his changeup frequently much harder than his fastball, which … nah. The Statcast system, like PITCHf/x, also was missing data on a large number of pitches.
Something was wrong with the data collection that night, at least during Arrieta’s start. That does not mean I am saying the data was specifically corrupted in a way to tank his velocity readings, and I’m also not saying the radar readings recorded were incorrect. I am saying only that, when confronted with a shocking set of numbers from a data set that was collected at a time when something was going wrong with the system, I don’t think it’s a sound practice to say that the data definitely proves anything in any direction.
In other words, I see this data, consider the circumstances, and I say: hmm. This is definitely something to track going forward.
And that’s as far as I go.
There’s another wrinkle to this that will take more outings to sort out: the way pitch velocity is being tracked in big league stadiums has changed this season (it was news to me and a whole bunch of other sabermetrically-inclined folks this week). That article from Tom Tango, as well as this one from Dave Cameron, get into the changes, but the upshot is that we may see velocity readings register about a mile per hour higher than we’re used to seeing.
Of course, none of that would explain why Arrieta’s readings were much lower than expected. It does, however, underscore the fact that outsiders may not yet be working with the best available data.
Thinking back on what my eyeballs were telling me during Arrieta’s start – which I can trust barely more than the data, to be honest – I did not see premium, 2014-15 era velocity coming from Arrieta. But I did see batters failing to square up his pitches, and at least a few tardy swings. Arrieta, himself, mentioned that he started over-rotating at times, so perhaps his naturally complex mechanics were off? (That might help explain why his vertical and horizontal release points also registered that night as pretty far off where they usually are.)
We should be fair and note that Arrieta’s velocity actually started to dip last year, particularly in the second half, and it was a point of discussion for a while – so much so that Joe Maddon had to offer his thoughts on whether he was concerned. We should also be fair and note that Arrieta has sometimes started a season with a little less velocity than he hits by the summer months (as most pitchers do), and it is also entirely possible that some pulling back on the reigns was designed over the offseason to improve Arrieta’s command troubles. [Edit – as noted in the comments, one more “it’s fair to point out” thing: although data was missing for Wainwright, too, his velocity readings were not abnormally low.]
As I said, though, it’s possible that the explanation at the end of all of this is going to be simply that Arrieta is dealing with the decline in velocity that so many other pitchers do in their 30s. Even if that’s the case, it’s not the death knell to his effectiveness (as a certain rotation-mate can attest). But, for a pitcher without impeccable command, having plus velocity greatly increases the margin for error.
We’ll see where things stand when Arrieta takes the mound this weekend (his start was moved from Monday to Sunday).