Are The Cubs Pitchers Getting Unlucky?

Last season the Cubs pitching had a great BABIP, .255, the best in MLB by quite a margin.  This has been attributed to superior defense and great pitching. Both of these seem like good enough explanations. This season, with roughly the same team, their BABIP has dropped to .286, effectively league average.

Now, there is a lot you may assume from these numbers. You may think, of course, BABIP is so unpredictable. Of course a team that was above average one season would return to average the following season.  Fair enough, you could say that. Some argue their outfield defense has taken a step back. That might be true.

These are all guesses and assumptions, but fortunately we have a few more tools for evaluating quality of contact, so let’s see what they may tell us about this Cubs pitching staff. We can use Statcast to evaluate quality of contact. When we do so, it paints a bit of a different picture.

Cubs Pitching xStats
2015 .234 .290 .365 .296 .318 .286
2016 .233 .303 .373 .289 .314 .294
2017 .247 .320 .395 .305 .330 .311

A Few Consistent Stats

First, let’s talk about a few aspects of the pitching staff that have remained relatively stable: strikeouts , poorly hit balls, flyball rate (note, I am using my own definition of flyballs, which are a subset of flyballs you’d find on Fangraphs read more.)

Alright, that may seem like a short list, but it tells us something interesting. These pitchers, on average, have retained their ability to suppress the frequency of plate appearances where the batter has a chance to succeed. That’s not nothing. In fact, it is certainly something. Last season 43.9% of plate appearances against Cubs pitching ended in either a strikeout or a poorly hit ball.  This season, it is… also 43.9%. The same. This is a good thing.

As for flyballs, this may be more of a coincidence than a meaningful stat. The Cubs have averaged 13.4% flyballs in 2015, 2016, and 2017. That’s weird. Normally these numbers fluctuate, if even only a small amount. But, hey, weird things happen sometimes. This 13.4% figure isn’t particularly noteworthy, 13.8% is the league average, give or take. But, hey, the number has remained the same so it falls in this category.

Lots of Changing Stats!

Okay, so three things have remained constant, that means just about everything has varied. Some of these variations are the random fluctuations you’d expect from year to year stats. Things like a slight juggling of exit velocity between batted ball types, or variations in BABIP or batting average. Other things, well, they are a little more troubling.  Like this steady increase in walk rate, from 7.7% in 2015 to 8.2% in 2016, to 8.8% in 2017. Although, even then, that walk rate isn’t particularly impressive. Not ideal, sure, but nothing to write home about.

But no, the real troubling aspect stems from a jump in expected slugging percentage, fueled by a bump in Value Hit rate, which itself is fueled by a big tick up in High Drive launch angles (HD%). Yay jargon, let’s dig deeper.

First, take a look at exit velocity. I want to start here because it is misleading.  You might look at the overall average exit velocity and think, well, it is down, that’s a good thing, right?  Well, maybe.  Probably.  I mean, it is always better for exit velocity to go down rather than up, but what does average velocity tell us?

In this case, the velocity is getting dragged down by what I call DB, the weak ground balls. These types of balls are generally outs in the first place, registering a hit less than 20% of the time.  Is it really a benefit to reduce exit velocity on these balls?  Probably not. I’m not sure. It might help. But how much worse can these balls get? You might be eliminating singles, that’s about it. Singles are rarely your threat, doubles and home runs are where we want to focus our attention.

In terms of doubles and home runs, you want to look towards LD, HD, and FB.  In this case, each of LD, HD, and FB are within their normal exit velocity bounds. Not really significantly higher or lower than previous seasons or the league average. So, yes, exit velocity is down for these pitchers.  But it is down in the wrong area, and that isn’t helpful.

How about launch angle frequency? The chart below shows the frequency of balls in the air over the past few seasons. As I stated before, the FB category has been oddly stable.  That’s a weird coincidence. Pop ups have remained somewhat stable, as have Line Drives. High Drives, though.  They are up. They are up to 10.6%, note that 9.5% is the league average. Also note that HD have a wOBA of .730 and a slugging of 1.252.

So, what may be driving this uptick in High Drives? Well, delving into the pitch stats this season makes me a bit uneasy, as I’m not quite sure how to interpret things like pitch movement in the new world of Trackman, especially when comparing to old world stats measured by PitchFX.

One thing of note, the pitch velocity is down for the Cubs. This has sparked some debate around baseball between fans, and I am sure even at the executive level. Tom Tango has written a bit about this, and has a tangentially (very tangentially) related post on his blog.

Allow me to explain the situation as I understand it:

Trackman radar appears to be very good at measuring ball speed, and that is it’s primary function. Whether it is pitch speed or batted ball speed. It can do so with very good accuracy, greater accuracy than we’ve seen with prior camera based systems like PitchFX.  Generally speaking, you should trust the velocity readings. That said, there could be bugs or calibration problems. Strong emphasis on *could*. No such problem, to my knowledge, has been identified. As a result, we should trust the velocity numbers for the Cubs pitchers.

Having said that, these numbers do not inspire confidence. The velocities are down 1+ mph from last year, on average. If we break it into individual pitchers:

Cubs Fastball Velocity
2016 FB 2017 FB Delta
Hendricks 87.5 85.5 -2
Lackey 91.6 89.9 -1.7
Arrieta 93.8 92.2 -1.6
Lester 91 90.5 -0.5

You don’t need to worry about Lester, but the other three starters are troubling. Especially Hendricks, who has fallen well below league average in terms of velocity, and while he has been able to make it work to date, his scFIP has ballooned from 2.66 last season to 4.11 this season.

Cubs Pitching xOBA
2015 2016 2017
Jake Arrieta .249 .294 .303
John Lackey .305 .309 .336
Jon Lester .282 .272 .283
Kyle Hendricks .288 .260 .316

In terms of overall batted ball quality, Lester remains an above average pitcher. Kyle Hendricks was well above average last season, but has dropped towards average. Arrieta was well above average two years ago, but hasn’t been able to find his mojo since. Lackey has fallen from average to below average.

Is It Luck?  Is It Defense?

Short answer, no.  

Longer answer:

Arrieta is teetering on the edge between above average and average. Since his heyday in 2015, which had a 45% DB rate, Arrieta has given up increasing numbers of fly balls, line drives, and pop ups. Obviously the pop ups are no big deal, but those line drives and flyballs are leaving a big mark on his ability to limit damage. He could still return to elite level pitching, if he can either induce more pop ups or drop a few of those line drives into lower launch angle categories. But his dip in velocity may be hurting him on that front. While I am very uncomfortable making this comparison, he’s also lost a ton of movement, both horizontal and vertical, on pretty much all of his pitches.

Hendricks is getting hit significantly harder this year than he had in years prior. There have been many times in the past few years where people have questioned his ability to get outs, and he has proven his doubters wrong at every turn. However, at that point in time he was inducing a ton of weak contact and perhaps getting unlucky with sequencing or low probability hits.  This year it is the opposite, he is getting hit very hard and has largely escaped damage with lucky sequencing or perhaps some good defensive plays.  If he doesn’t figure this out soon, his luck could change in a heartbeat, and his ERA could land somewhere north of 4.

Lackey has been trending downhill for three years now. This isn’t unexpected, he is 38 years old afterall. Soon to be 39. His problems appear to stem from a regular aging curve, so there isn’t so much to say about him here.  He’s shedding velocity and movement, and you probably shouldn’t expect much more from him from here on out.  

Jon Lester is the bright spot in this rotation.  He’s been hit a bit harder this year than last.  His velocity is down a tick. But this is general aging, and nothing to be concerned about.  He’s still been elite, by every metric I have available. He’s allowing even fewer High Drive balls this season with lower exit velocity. He’s giving up fewer line drives and more weak ground balls. He’s been all around great.

The Cubs pitching has been getting hit harder. Especially Lackey and Hendricks. Good defense might be able to limit some of the damage, but the real problem here stems from pitching.  Not luck. Not defense. This Cubs rotation is fighting through real, measurable problems.

What does this mean for fantasy?

First and foremost, you might want to trade Kyle Hendricks. He has pitched reasonably well over the past few weeks, and still has an above average ERA and well above average BABIP. However, these warning signs shouldn’t be ignored, and you might want to flip him for another piece while you have the chance.

Arrieta should bounce back. His start against the Rockies was absolutely brutal. His starts against the Redsox and Reds weren’t too much better. He should still be an above average pitcher, though. Maybe not the ace you thought you were getting, but above average nonetheless.

Lackey is sunk costs.

Andrew Perpetua is the creator of and, and plays around with Statcast data for fun. Follow him on Twitter @AndrewPerpetua.

Newest Most Voted
Inline Feedbacks
View all comments
gaius marius
5 years ago

this is interesting, but it would be more useful if we had context on variance in Statcast-derived expected results. is it normal or unusual for a pitcher (eg, Hendricks) to randomly vary by 40 bips around some xOBA mean — particularly over just 9 starts and 52+ innings, rather than 33 and 200? another way of saying it: is there anything to suggest that his last 50 innings are more or less predictive of his next 50 than his last 400 are, and if so to what degree?

too often the analyses that we generate in sabermetrics are lacking in rigorous formulations of variance and error. that stuff is boring, of course, and often undermines the narratives we wish to create. but it’s badly needed context for meaning.