Jake Arrieta’s Been Off for Several Weeks
Ah, this part. I always enjoy this part. Not because I enjoy when superstars struggle, but because I enjoy seeing the standards to which we hold those same superstars. The standards are revealed when people start to worry. Over his last…I don’t know, eight starts, Jake Arrieta has a 3.77 ERA. He’s allowed a slugging percentage of .331. This is Jake Arrieta in a rut. I mean, this really is Jake Arrieta in a rut — people aren’t making up the rut — but you can see how silly reality is. That all-time amazing version of Arrieta became the version we allowed ourselves to take for granted.
But, heck, where’d he go? Arrieta himself would admit he hasn’t been firing on all cylinders. He’d like to pitch better than he has. The Cubs would like him to pitch better than he has. While Arrieta hasn’t been bad, lately he’s fallen short of expectations. And if you examine the record now, you see signs that Arrieta has been off to some degree for several weeks. This isn’t all that recent of a thing.
To repeat, it’s not as if Arrieta has been broken. For the most part, he’s continued to get outs. His strikeouts aren’t yet suffering, and he’s not getting hit all around the ballpark. By Statcast’s exit-velocity statistic, early Arrieta was solid. He’s stayed pretty solid. Jake Arrieta isn’t getting knocked around. The issue is that his command has suffered. Arrieta hasn’t been locating like he likes, and though opponents haven’t taken advantage by racking up extra-base hits, walks are walks, and baserunners are baserunners.
Some handy plots can give us a sense of the timeline. The walks? Here are Arrieta’s walks, in rolling seven-game samples:
This year doesn’t look much like last year. It doesn’t look much like the year before. Arrieta has spent a lot of this year north of that average baseline, whereas last year he was almost always under it. All right, let’s follow this. Arrieta says he simply needs to stop falling behind. He’s not wrong. Here’s another rolling-average plot, this time showing first-pitch strike rates:
A few things are of note here. One, obviously, of late Arrieta has been throwing too many first-pitch balls. But it might be some consolation to see that he did this for a stretch later last season, and also in the season before. This isn’t something he wants to be having happen, but it doesn’t necessarily have to be toxic. Arrieta, in the past, has been able to rally. This next plot is, I think, the most meaningful one. It’s also simple: A rolling-average plot of overall strike rate.
This is the one I don’t like, for Arrieta. The walk-rate plot wasn’t great, but that shows walks over all plate appearances. This shows strikes over all pitches, and that comes with a way bigger denominator. This is where I really see the recent problems. Over his last seven games, Arrieta has thrown just 60% strikes. He spent most of last year close to 65%. Arrieta hasn’t been this low over a seven-game stretch since settling in with the Cubs. By no means do I intend to suggest that Arrieta has somehow regressed to what he was as an Oriole, but his pinpoint location has been failing him. He hasn’t thrown enough strikes early, and he hasn’t thrown enough strikes to battle back. Left-handed hitters, in particular, have been content to wait Arrieta out. Last year, overall, Arrieta threw strikes with 65% of his pitches. He hasn’t hit that mark in even a single game since May 25. That’s eight starts ago.
As isolating the issue goes, it’s obviously very complicated. You always wonder about some injury, but for whatever it’s worth, Joe Maddon isn’t concerned about that.
“If he was injured, I’d be more concerned,” Maddon said. “He’s not injured. That’s the reason I feel he’ll be fine.”
We probably just have to assume Maddon is correct. He might not be, and sometimes injuries are hidden and relatively pain-free, but we don’t have enough to state conclusively that something is badly wrong. Now, a lack of command is worrying. It’s also worrying that Arrieta is coming off his lowest average velocity in a while. But, some kind of mechanical kink is more probable. Let’s just say it’s mechanics. Borrowing from Brooks Baseball, you can see that hitters have been less willing to go after Arrieta’s slider and curveball:
Maybe you think of it as a cutter, and not as a slider. I don’t care! It doesn’t matter. This plot shows swing rates against those two pitches, combined, and the swing rate has dropped. Don’t get too thrown off by that point for July 2016, since that’s just one game, but it doesn’t mean nothing. Arrieta has been less able to throw strikes with his breaking balls. Hitters have been less willing to swing at them. This has all put more pressure on Arrieta’s two-seamer, and no pitcher wants his repertoire to be narrowed.
And now here we see something interesting. A few years ago, Arrieta threw 48% fastballs. Last year, he threw 51% fastballs. Earlier this year, he was at 61%. More recently this year, he’s been nearer to 70%. So even to start with this season, Arrieta was throwing more fastballs and fewer sliders and curves. This pattern has only been amplified during recent struggles. This is a big deal. Arrieta’s curve has been important for him. And the previous two years, Arrieta’s slider was one of the very most valuable pitches in the game. It was arguably the whole difference-maker, and you don’t see pitchers go away from their best pitches like this for no reason. The best version of Arrieta has a dominant slider. Of late he’s been less willing to throw it.
We have a great pitcher with great breaking balls, throwing fewer breaking balls. Hitters are also swinging at those breaking balls less. There are a few ways one could try to interpret this. Could be Arrieta has lost some confidence in his secondary pitches. Maybe he just doesn’t have the right feel, and that’s also why he’s more often missing the targets. Or, could be this is out of an interest in self-preservation. Arrieta knows he just threw a ton of innings. Maybe he figures this way he’s less likely to get hurt. A season like Arrieta’s 2015 takes a toll, and it’s not like the various Mets starters have all been healthy versions of themselves. Pitching is dangerous. You might’ve heard.
Whatever’s going on, Arrieta isn’t broken, probably, but he isn’t quite Jake Arrieta. Strikes just haven’t been there in more than a month, and where a year ago Arrieta’s slider was worth +24 runs, this year he’s sitting just below +2. The Cubs say Arrieta isn’t hurt, and Arrieta himself says the fix is to more frequently get ahead. Of course the fix is to more frequently get ahead. The question concerns what he’ll have to do to do that. It’s been a while since Arrieta was on top. It’s been a longer while since he wasn’t.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
So have I.
I date it back to the unsubstantiated but public claim (by one national blogger, subsequently castigated by everyone else as slanderous) that he had to be doing testosterone to go from fodder to elite. Got into his head? Or just coincidence?
Solid stuff as always. As a Cubs fan, I’ve definitely noticed two things–the low pitches out of the zone, combined with what appears to be a new ability for hitters to not swing at them. All the graphs support this, and the real question will be how he responds. Every rational Cubs fan knew his second half of 2015 was historic (without being hyperbolic), but when he started this year and continued to dominate, we started thinking maybe, just maybe, we were seeing something special.
It’s not over yet, and he can certainly rebound, but it’s clear hitter’s approach to him has changed. His pitches per at-bat have been rising by month (3.53 per PA in April, 3.74 in May, 3.99 in June), and that limits him to six innings or less as the pitch count rises. His ground ball rate is dropping, and he’ll have to adjust his approach. It’ll be interesting to see what he does, as well as extremely important to the Cubs postseason success.
I was going to say the same thing. I saw the weird game vs the diamondbacks and it was clear that it wasn’t so much a matter of him losing his control as the hitters realizing that his two-seamer was going to drop out of the strike zone so they just laid off, and from the parts of games i’ve seen him pitch since then it looks like hitters on other teams are doing the same. He just isn’t fooling them any more nearly as much.
It’s very similar to what happened with tanaka in the second half of 2014. He looked unhittable the first half of the year because batters hadn’t learned to lay off his sinker. Of course the number of balls went up when they learned to lay off and people who weren’t looking at the games said he was suddenly having control problems, but his control was the same, he just wasn’t fooling hitters nearly as much.
Maybe he’s throwing more fastballs as a result of batters taking his breaking pitches for balls. Batters watch video and see that his slider-cutter usually breaks out of the zone.
In the game i saw it wasn’t just the slider they were laying off of, it was the two-seamer also.
Great post, and a scary one for Cubs fans. Couple of additional observations:
o Arrieta’s peripherals this year are in line with his career rates … except for walks. As Jeff suggests, Arrieta isn’t walking batters at Baltimore levels, but he’s significantly exceeding his recent walk rate.
o Even ignoring July (just one start as I write this), Arrieta’s velo is down across the board. He’s lost just a little on every pitch.
http://www.brooksbaseball.net/velo.php?player=453562&b_hand=-1&gFilt=&pFilt=FA|SI|FC|CU|SL|CS|KN|CH|FS|SB&time=month&startDate=03/30/2007&endDate=07/05/2016&s_type=2
o Arrieta has abandoned the 4-seamer. This isn’t a sudden change – he’s been gradually de-emphasizing it since 2014. Maybe this is a pitch the Orioles were pushing him to use against his will. But ditching it completely doesn’t seem to have helped.
http://www.brooksbaseball.net/velo.php?player=453562&b_hand=-1&gFilt=&pFilt=FA|SI|FC|CU|SL|CS|KN|CH|FS|SB&time=month&startDate=03/30/2007&endDate=07/05/2016&s_type=2
A lot of this is probably regression to the mean – Arrieta’s not nearly as good as he was last year, and not nearly as bad as he was in Charm City. But his recent decline isn’t random – there are specific and identifiable failure paths here.
Regarding velocity, it may be an attempt to reduce strain on his arm. After he burned out in the playoffs last year, and especially because it was the only time in his career he’d thrown a qualifying number of innings, a decision was made that he should be used this year so as to conserve strength in preparation for the postseason. With all the balls he’s been throwing this year, especially in the last month, this is putting even more strain on his arm. He’s been averaging well under 6 innings a start on pitch count alone.
When all the sudden everyone stops swinging at breaking balls, I have to worry there is some type of tipping of pitches. Any evidence of that in video, that anyone has seen?
No one has suggested this and I don’t think this is the explanation. The different pitches look clearly different from each other, the problem before wasn’t with distinguishing the pitches, it was figuring out how much they were going to break.
there was a lot of talk about how Arrieta had raised his delivery slot and that was part of the explanation for his graduation to excellence. there’s a good article on it at fivethirtyeight. Has it dropped back down
pretty hard to imagine that it isn’t related to the steroid use
It doesn’t look to me like he’s doing anything different. It looks like the batters are doing something different.
I noticed that his ability to throw strikes in the last 8 games compared to his first 9 games is down from 65% to 61%. His Pitches per BF comp is 3.9 to 4.2. This raises an entirely different subject for me. As much as I am a baseball traditionalist, I really think we need to move to automated ball/strike calls. Has Arrieta really lost some of his control or is he not getting all the calls he was getting before? I have no idea, but by automating the process you eliminate:
-Umpire-to-umpire strike zone variations
-Flat out missed calls
-Late inning strike zone changes
-Heat of the moment calls
-Catcher slight of hand
-Batter’s position in the box
-Height of Batter
-RH vs. LH Batters variation
-etc.
With today’s technology we can easily assign a strike zone range for every player in the league. This ensures that Jose Altuve doesn’t have the same strike zone as Freddie Freeman. No more Tom Glavine 6″ off the plate calls at the end of games. Everyone is on equal footing at every point of the game. Overall this is better for the Batter, Pitcher, analysts, scouts and fans. This isn’t putting the home plate umpire out of business, they can just focus on other things and get those calls correct. This way we could determine if Arrieta has lost some of his control or whether is was getting calls he isn’t now.
I think the steroid talk is silly and nonsense
I think we have to look at the huge spike in workload/innings pitched
in 2013 he pitched 80.4 innings
In 2014 he went up to 156 innings
Then last year a large boost to 229 innings plus the 20 or so playoff innings
and this year so far he has pitched 108 innings
so from 2014 till now his arm has pitched roughly 513 innings
I think what were is seeing decreased control and effectiveness because of the large amount of wear on his arm and hopefully for Cubs fans it’s just a dead arm period and no long term effect on his health/arm.
He just isn’t locating consistently. Last year he could spot the slider to both sides of the plate anytime. This year he is throwing fastballs in counts where he would be throwing the slider. Curveball looks the same as it’s less of a command pitch, but the slider isn’t as sharp.
Overall his zone rate is the same as last year, 44.1% vs 44.2%. He’s throwing as many pitches in the strike zone. What’s changed radically is his o-swing%, it’s dropped from 34.2% to 29.2%. He’s isn’t getting batters to chase the balls out of the strike zone as last year.
Okay, I just looked at the heatmaps comparing the second half of last year with the last six weeks and there are a few things going on. First of all, as you said, the control of the slider is way off. In terms of overall zone% numbers, this is presumably being balanced out by him using the slider less and the fastball more. But also the batters are focusing their swings on the slider much more than they were last year–last year they were swinging sliders that were high, low, inside, outside, they really didn’t know where the pitch was going, now they’re basically focusing their swings on when the pitch is in the strike zone much better. The batters seem to be able to figure out much better where the pitch is going (which seems sort of strange because Arrieta seems to be less able to control it). The batters are also swinging at low sinkers out of the strike zone less now (as I said, this was what was most obvious to me in the games I watched). There’s also a change in the curve. He’s throwing fewer curves for strikes, but as you said it isn’t a control pitch anyway, in any case last year didn’t throw very many curves for strikes, he threw a huge number in the dirt and hitters swung anyway. Now they’re swinging way less on low curves out of the strike zone. It may be that they’ve figured out that most of his curves will be out of the strike zone, but this may also be affected by the fact that in fact he is throwing even fewer curves in the strike zone than he was last year.
Swapping the 4 seamer for a 2 seamer, reducing offspeed usage. Isn’t that similar to what Chris Sale is doing as part of his ill-conceived pitch to contact plan?
Players, and frequently managers, still seem to channel Crash Davis.
It isn’t the same thing. Arrieta’s fastball was never as much of a strikeout weapon as Sale’s. And Arrieta is(apparently) reducing the offspeed usage because he isn’t throwing the offspeed pitches for strikes, not as part of a general plan to change his approach to one which will produce better results. His contact numbers were very good before.