If I were to rank the months of the year in terms of difficulty of writing for FanGraphs, February would probably be #1; there’s just nothing going on in the second month of the year. The off-season is over, only the most meaningless parts of spring training are going on, teams aren’t aggressively signing long term contracts yet, and there’s just generally no real news on which to comment. February is the best month for me to take vacation, basically.
But behind February, April might be the second hardest month to write for FanGraphs, because while we have games to watch and data to look at, the overriding reminder of history that is drawing conclusions about anything this early in the season is probably foolish. Last April, Justin Upton was Babe Ruth. Last April, the Texas Rangers looked like the best team in baseball. April games matter in the standings and April performances do have some predictive value, but the samples are so small that we should rarely be willing to believe that a player has made a dramatic transformation from what they were before hand. Realistically, the conclusion of almost every data point we currently have is “That’s interesting; who knows what it actually means?”
But we still have to write about baseball in April, and we have to try and make it as interesting as we can. You don’t want to read five or six pieces a day that tell you to all the numbers right now are useless any more than we want to write them. But most of the numbers right now are useless, so we hunt for stories that are interesting and numbers that might be less useless than the rest. The overarching conclusion is still Beware Small Sample Size, but there are things that are least worth monitoring going forward. They might not continue, but if they do, it’s news. This is one of those things.
For the last few years, Kenley Jansen has been one of the best relievers in baseball. He has dominated with a repertoire of essentially one pitch; the cut fastball. He throws it 90 to 95 percent of time, and like Mariano Rivera before him, he throws it so well that the predictable nature of his pitch selection seems to not matter at all. He just pounds cutter after cutter, and hitters swing through them at ridiculous rates. Jansen is a one trick pony whose one trick is pretty amazing.
And now, for the first few innings of 2014 at least, Kenley Jansen has apparently found a new gear. Traditionally, his cutter has sat around 93 mph with minimal variance; 40% of the 1,245 pitches that he threw in 2013 were beween 92.0 and 94.0 mph. That’s his normal velocity range, and it has been for several years now.
Well, so far this year, PITCHF/x has recorded 57 pitches from Jansen — there were no readings for the Australia games — and only 23% of them have been between 92.0-94.0 mph. For his first few outings here in the U.S., his normal range has been 95.0 to 97.0. To show the difference, here’s a chart with the percentage of pitches thrown in each velocity bucket above 90 mph, where the buckets are simply the whole number, so 95 is 95.0 to 95.9.
Yes, that top line is correct: PITCHF/x has already classified more 97+ mph pitches from Kenley Jansen this year than it did all of last year. Maybe more staggering is the cumulative percentage of the top three lines; of his pitches greater than 90 mph this year, 51% of them have been recorded at 95.0 mph or faster, compared to just 7.2% last year. We’re dealing with a sample of just over 50 pitches over three games in two stadiums, so this is about as small a sample as I’ll ever quote here, but the magnitude of the difference is pretty staggering. If you prefer graphs to charts, then here’s a picture of what his velocity jump has looked like.
At the risk of beating a dead horse, this might be nothing. Last week, I wrote about the danger of reading too much into Stephen Strasburg‘s low first start velocity, and not too surprisingly, it picked back up to near-normal (for him, anyway) in his second start. And Strasburg threw twice as many pitches in his first start as Jansen has thrown over the three relief appearances we are measuring. Jansen could easily go right back to throwing 92 to 94 for the rest of the season and no one would blink an eye.
But for his first few appearances here in the U.S., 92 to 94 has not been the norm. Jansen has been excellent throwing 92 to 94, but for a few outings, he’s been sitting 95 to 97 instead. Is a couple ticks on his fastball going to make him a better pitcher? That seems difficult to believe, given that his cutter is already basically an unhittable weapon; it’s hard to be more unhittable than he already was. In fact, there’s some anecdotal evidence that velocity spikes might actually be a precursor to a pitcher reporting an arm injury, and if that theory was correct, this could actually put Jansen in a higher risk category for future DL trips.
Or, again, it could be nothing. Given that the rest of the Dodgers’ pitchers have mostly stable velocity readings from last year to this year, it probably isn’t PITCHF/x measurement error, but that doesn’t mean this is Jansen’s new established velocity. If you were building a future velocity prediction model, you’d regress his 2014 numbers significantly, and probably end up with a future expected velocity of closer to the 93-94 that he’s historically sat at, with a little bit of bump to account for the most recent higher numbers. But, in terms of numbers after six or seven games, this is one of the more interesting data points to watch, because Jansen has been ridiculous in the low-90s, and I don’t even know how anyone could regularly hit him if his cutter decides to stick around in the mid-to-high 90s instead.