Note: This post is all prologue. If this makes no sense to you, you should A) watch The Shield and B) read it anyway.
Alright, step forward if you were one of the league's top ten pitchers in June. Welcome aboard, Trevor Bauer. Come on in, Mr. Fernandez. Not so fast, Zack Greinke. Sorry, Mr. Syndergaard, you may not pass. But wait, what's that? Mr. Norris? Why, go right ahead.
Yes, it's true. The man who was banished from a rotation featuring Williams Perez had an absolutely resurgent June. And no, this wasn't some kind of .150 BABIP, 0% HR/FB, 90% strand rate gift of the luck dragons. We're talking a 2.15 ERA, 2.00 FIP, 2.85 xFIP, and 3.19 SIERA. Those are all top-15 out of the 101 pitchers making four or more starts in June, and top 10 by FIP and xFIP. So hey, he's real, and he's spectacular. Apparently.
Of course, the big question on everyone's mind is... how? Or perhaps, why? Or even, "Should I pledge allegiance to the Mole People and/or Prepare My Meteor Bunker?" Well, I can't answer the last one (however, all signs point to "yes"), but there's at least some indication of the other two. But it turns out that Bud Norris is a mysterious guy, so the answers may beget more questions.
Let's flash back to Monday. Bud Norris just completed an outing where he shut out the Mets for seven innings, striking out eight, and most importantly, walking no one. Again. (He's only had two starts this season where he's walked one or zero batters, and both have come in June.) So I'm thinking, "Hey, I wonder why Bud Norris is suddenly good now? To the Fangraphs!" And, I'll show you pretty much what I saw:
If you read from the bottom up, the red box is his slider usage rate during his first, highly unsuccessful stint in the rotation. The blue one is his slider usage rate after he returned to the rotation. You'll see that it also crept up as he spent more time in the bullpen. You'll also see that as relievers can get away with a lower pitch mix, it seems like Bud Norris maybe adapted some of his lessons learned from the bullpen into his strategy as a starter. A two-pitch starter isn't a great bet, but when your other pitches weren't working, can you blame him? Furthermore, do you really want to argue with results? Fewer changeups and two-seamers, more four-seamers and sliders, and much more success.
So, hey, the mystery of how Norris got his groove back was solved, right? Right? Wrong. Luckily for me, I avoided embarrassing myself by not actually having the opportunity to post this when I first noticed this. Because life got in the way, and it bought me just enough time to see that Jeff Sullivan at Fangraphs had an even better idea.
(If you haven't read it yet, now would be the time to quickly look at this. I know I'm always telling you to click and read stuff, but if you're reading an article about Bud Norris, you should go ahead and do this too.)
Spot the difference? There were no cutters in the image above, right? So... what's the deal? Another Bud Norris mystery, yet not really about Bud Norris at all. Some internet sleuthing reveals this:
What makes Brooks an even better site than it appears is how they "correct" the PITCHf/x data to augment classifications, discuss grips, and make sure that park effects (some cameras are positioned differently) are properly considered. You’ll sometimes notice that our PITCHf/x data doesn’t line up with theirs; that’s because they’ve devised a strategy for fixing "errors" made by the automated PITCHf/x algorithm. Their data, also, is updated immediately following the game, so you can pull up pitch charts in near real time.
So, to summarize: Bud Norris is actually apparently not leaning more heavily on sliders, but is instead incorporating cutters into his arsenal. The cutters are working. Yay, cutters. But, there's something about them that makes Pitch F/X think they're sliders. Had I been more astute, this may have been immediately visible: if you go here and compare his April "slider" vertical movement to his June "slider" vertical movement you'll see it has less downward break. Which makes sense, since the data are really mixing sliders and cutters. This investigation also led me to discover this helpful chart, in case pitch names just sound like weird professions from a crappy sci-fi novel to you:
So, I suppose that one's mystery solved. Or, really, two mysteries. Bud Norris is throwing cutters, and they're working for him. But, they're being (mis)classified as sliders by the raw Pitch F/X output, which makes it look like he's throwing more sliders. What he's definitely doing less of, though, is throwing his bad pitches. To wit:
I'm not a fan of pitch values for many reasons. Part of it is that they're not really predictive. But, Norris's pitch value numbers are illustrative in a pretty interesting way. Take a look:
There's one happy number there, and a lot of sad numbers. That two-seamer is especially brutal. Even the whole slider/cutter classification issue aside (and noting that his new cutter is apparently different enough from anything classified as the old cutter to not really be the same pitch), it's pretty clear that Norris has made his living so far off of his slider, and everything else has been pretty poor.
What's especially notable here is that for a guy who has stuck around in the league for as long as he has, Norris has never actually gotten particularly good results with his fastball. Given that he, like most pitchers, throws his fastball more than his other pitches, that's kind of a problem. So, it's not surprising that by minimizing use of pitches which have not worked for him, and emphasizing something that works more like a pitch that has worked, he's having more success. I also looked for pitchers with similar fringy success to Norris and their own pitch values:
Not surprisingly, these guys don't tend to have good results on their fastballs. (Also, Bronson Arroyo's fastball, yeesh.) That might be why they've been average starters with some career ups and downs over time. But, Norris has had a worse fastball than his peers in this bucket, though he hasn't really slid value-wise because he's always had that slider to rely on. Adding an effective cutter to the mix and minimizing use of a fastball that hasn't done much for him to date could further augment his success.
Just a few more things about Bud Norris. The big knock on him (and why I thought he'd be awful in 2016) is that he cannot bring himself to limit walks. Even in 2016, after accounting for his really good run in June, his walk rate is in the bottom 22 percent of all starters with 70+ innings in 2016, and when you only look at starters that don't get demoted to the bullpen or sent back to the minors, it would be something like the second-worst. His walk rates, both on a percentage and per-nine basis consistently rank as "really bad" or "awful" among his peers. Another thing that's plagued him (until now, apparently, per Jeff Sullivan's article) is an inability to get lefties out, which is kind of important for a starting pitcher. To that end, I found this pretty interesting:
Essentially, Norris' most common fastball misses the zone. If you delve deeper into it here you can see that this problem is mostly limited to lefties, as his zone chart against righties looks more reasonable. But not being able to consistently throw strikes (or not being willing to?) to lefties with your primary, most-used pitch? That's a recipe for disaster, and explains the walk rate pretty well.
His 2016 usage is all over the place right now in terms of location, with lots of pitches thrown low and outside to righties (low and in to lefties). But what's most critical is that he's not going to the fastball well and throwing it out of the zone as he has been consistently in the past. There's not a clear enough pattern to say whether he can consistently throw the cutter and slider in that corner zone and reap dividends from it, but anything that pushes him away from the ineffective fastball to lefties is probably a win for him.
How Norris will fare going forward, and whether his time as a Brave is limited are additional mysteries that we're all going to solve together. But at least it looks like his contract this offseason will pay off better for the team than it seemed it might, and that's a good thing.