Was the Kake a lie?

Meandering thoughts' on Nick Markakis and his 2015 season. As was already said, he was pretty much as advertised, though the ways he got there were possibly not what people expected.

By Ivan the Great Oct 14, 2015, 4:00pm EDT

Disclosure: I've had a post on Nick Markakis sitting in the editor since mid-June. It never got done. It was going to have little images of cakes. I think a targeted discussion of most players' seasons warrants more time and attention than I've been able to give it so far, but maybe I'll actually get this post done at some point before 2016.

For me, any discussion of Nick Markakis and his value, either backwards-looking to the 2015 season or forwards-looking to the remainder of his Braves tenure inevitably comes back to two points:

Hey, he's shown an awesome OBP skill. Yay!
Hey, he has pretty much no other effective skills. Damn!

Really, that's how this discussion could be left. So for once, the "tl;dr" component is right up front. If you don't want to read the rest of this long post, you can jump right to the comments and argue about these two points. In the end, though, that's what his 2015 comes down to, and possibly how we'll continue to think about him going forward. There'll be more on those things below, but really, that's all it is.

ZiPS projected Markakis for 1.4 fWAR/600 with a 105 wRC+. Steamer had him at 0.9 fWAR/600 with a 101 wRC+. I was in the same ballpark, 1.6 fWAR/600 and a 104 wRC+. In the end, he hit a little better and did everything else a little worse, and ended up at 1.4 fWAR/600 witha 107 wRC+. This is about as unsurprising as a player can be, in the end. But of course, the devil's in the details. ZiPS and Steamer figured him for a .340 and .331 OBP, respectively. If I had to guess based on his pre-2015 numbers, I'd say something like .342. Of course, his .370 2015 OBP blew those out of the water.

Here's some context about how great his OBP was: it was 18th overall among qualified batters, 24th among players with 200 or more PAs, and, it was the same as Freddie Freeman posted this season. In Fangraphs' glossary, they refer to a .370 OBP as "great." Nick Markakis was great at getting on base, and that was great for the Braves, because avoiding outs is pretty much the most important thing a position player can do. Well done, Nick.

But, of course, the second of my two points is what causes this bit of a rosy perspective inexorably crash upon the rocky shore of humbug-esque player evaluation: he does the most important thing well, but does he do anything else at all? Well, not so much:

He's was only minimally above average offensively despite that great OBP, based on a weak .376 SLG and an .080 ISO, combining for a 107 wRC+. This is a bit below average for full-time regulars (because wRC+ = 100 is dragged down by non-regulars who hit worse than regulars): 110.
Going back to 2001, no player (with 200+ PAs) has posted (in aggregate) an OBP of .370 and a wRC+ of below 109 except Tim Raines and his 200 PAs across 2001-2002. In the 5,000+ player-seasons since 2001, only 28 players have posted a wRC+ below his 107 with an OBP of .370 or greater. So I guess he's unique, but he's unique in the Jamey Caroll vein. The average wRC+ for a player who can OBP .370 is 137 (!!). The average ISO is .206. The average fWAR is 4.1, and that doesn't adjust upward for all those players who had partial seasons and would've posted higher figures. Markakis' 2015 season put him in the top 15% of all OBPs accrued since 2000, and yet...
His baserunning was not great. He's never been a superb baserunner and is marginally below average (about a run below average per 600 PAs) on the bases, but put up a similar baserunning score this year (by BsR) to Buster Posey and Wilson Ramos (about 2 runs below average). Again, it's two runs, it doesn't matter much. But it's a way in which his skillset again punted on delivering value.
He was unable to recreate his anomalous positive defensive ratings from 2014, which should make us feel good about UZR/DRS in general, but not so much about his defensive acumen. DRS was harsher to his defense (-6) than UZR (-4) but either way, playing a corner outfield spot a handful of runs below average costs him about a win overall.

So there you have it. Again, this is the bird's-eye view of Markakis and his 2015, and it's perfectly possible to be content there. He was a bit above average offensively, hit for zero power, and ended up somewhat below average for a major league regular. With that said, these are things that people mostly knew, and the remainder of this piece will try to shed light on some things that might be of more specific interest as a postmortem than the summary above.

Contact, Power, and Recovery

When looking at Markakis' offensive profile in 2015 relative to earlier seasons, a few questions come to mind immediately. First, how did he manage to knock his OBP that high, its highest level since 2010? Second, why did he post a career-low ISO? Third, what effect did his offseason neck surgery and lack of a normal offseason routine have on his performance. The answers to those can be gleaned from his batted ball and quality of contact stats, as well as looking at some of his seasonal trends.

A picture is worth a thousand words, and I'll try not to write a thousand words about the following:

The main things we care about here are center stage: OBP and ISO. OBP was better than in recent history in both the first and second halves, though the first half was the real zenith for Markakis. ISO is the verse: it's lower than in recent history, and there's a deep valley there for the first half. This suggests, superficially, that Markakis was basically trading ISO for OBP, whether deliberately or incidentally.
His BABIP was higher than he's been able to sustain in both halves of 2015. This in and of itself doesn't explain his OBP spike nor his ISO plunge, though it is a contributing factor (usually, higher BABIP = lower ISO unless the batter is killing the ball, as more luck on balls in play means more singles relative to hard-hit extra-base hits). The variation in infield hit rate helps to explain some of this BABIP fluctuation.
In the first half, Markakis kept the ball on a pretty low exit angle, with a very good line drive rate and a grounderball rate over 50%. His fly ball rate, meanwhile, was very low. In the second half, something fairly strange happened: he stopped hitting liners and increased his fly ball rate, but his grounder rate also went higher.

Looking just at the left half of the chart, it seems like one possible explanatory narrative here is something like: Markakis did not get in a full offseason of strength-building, and decided to focus on hitting line drives and grounders to reach base. This resulted in a higher OBP at the expense of a lower ISO. In the second half, he attempted to add more fly balls to his results to reclaim some of that lost power, but ended up rolling over pitches and harming his liner rate while increasing both his fly ball and grounder rates. That's not the only explanation, but it might make some inherent sense.

However, what's a little weird is the spray chart and quality of contact information. If we were sticking with the narrative from the previous paragraph, we'd expect him to pull more balls in the second half, and perhaps make more hard contact. But instead, we see kind of the opposite:

The spray chart change between the halves is more balls hit the other way. And this is pretty crazy if you think about it. Nothing exemplifies Markakis' first half more than the chopper between third and short for a base hit, well-placed but also not hit softly enough for either fielder to make a play on it. Yet he pulled the ball far less in the second half.
Markakis was already making a very high degree of medium contact, and doubled down on that even further in the second half, He did not start making harder contact in the second half even as he (deliberately or not) hit more balls in the air.

So what we have in Markakis' 2015 is a guy who, for whatever reason, starts the season with a lot of medium liners sprayed up the middle and to the left side and posts a BABIP higher than his career average and much higher than his recent performance indicates. He then shifts towards more grounders and more fly balls at the expense of liners, but is still largely hitting balls up the middle and the other way. And as he does so, he makes less soft contact but also less hard contact. Anyone's guess is as good as mine as to what this specifically means, but it's interesting to think about.

One thing that I've heard said more than once is that Markakis should return to a better level of offensive performance when he puts his neck surgery behind him and has a full offseason to do whatever it is he normally does. The question I'd pose in response is: in what manner of peripheral will this improvement occur, and how does it intersect with his two halves of 2015? In other words, here are his two halves:

First half: .345 BABIP, .060 ISO, .381 OBP, 107 wRC+
Second half: .329 BABIP, .104 ISO, .357 OBP, 108 wRC+

Given these, how much more can he push those numbers up? He's not likely to be able to combine an ISO over .100 with a BABIP over .340 - these seem to be at odds with one another as ISO-depressing medium liners and their substitution for more fly balls decreased BABIP and OBP but also decreased ISO in the second half. So even if Markakis is healthier in 2016 and beyond, what will those seasons look like? If he returns to pre-2015, with a lower BABIP, higher ISO, more fly balls, less grounders, and more hard contact, that profile still looks like a 106ish wRC+, because his OBP will decline relative to 2015 even as his ISO increases. Essentially, what it boils down to is whether Markakis will be able to increase both his OBP and his ISO, and I'm not sure that there's much indication he'll be able to do so.

Plate Discipline

Of course, what's being glossed over in the discussion above is that Markakis put more balls in play in the second half. In the first half, three-quarters of his plate appearances resulted in a ball in play; in the second half, this increased to 81%. This is important because league-wide, only about 72 percent of PAs result in a ball in play, so Markakis took his contact to an extreme level in the second half. Maybe this depressed his BABIP, and maybe he will be more selective going forward, which will push his walk rate up and increase his OBP to compensate for any slugging-increase-based declines in BABIP.

But what we see is a little different. The main difference is mostly just that Markakis swung more in the second half. A lot more. More at balls (this will probably decrease BABIP and pretty much any good offensive stat, as you don't want to swing at balls), more at strikes, more in general. His rate of swinging at balls increased more than his contact rate fell on those balls, meaning that he was putting more non-strikes in play. He was also putting many more strikes in play. Given all this, I'm still not entirely sure what Markakis was trying to do in the second half. Perhaps he wasn't trying to do anything specific, and this is just how the chips fell. But in any case, what we see is a somewhat selective, but contact-oriented first half resulting in a lot of opposite field and up the middle medium-contact liners, and a lot less selectivity combined with more fly balls in the second half. Perhaps the elevated o-swing% resulted in the increase in grounders in the second half, in which case the question morphs to: can Markakis combine selectivity and increased power output going forward? I don't know the answer, but I still revert to what I indicated earlier: if he's going to improve all-around, I just don't know how exactly that's going to happen. We've seen a few different variations on the same theme of hitter in 2015 from him, and none of them are really all that different in the end, or all that valuable overall, given his lack of value on the bases or on defense.

Defense

Speaking of which, defense. To be honest, this was already mostly covered here, by BravesRays, to the extent warranting consideration, so I'll just add a few things.

First, holy moly, that arm score. -7 by DRS, and -6 by UZR. That's a consensus that his arm was more than half a win below average. Now, I'm tempted to say this is an anomaly just because it's so far out of line with anything he's done to date (his arm has been almost always positive to some extent), so it'll be interesting to see what happens with this going forward. My own personal eye test indicates that yes, his arm was an issue, but that's fairly meaningless. I also know that arm ratings can take on a spiral effect: players with worse arms will result in more extra bases being taken on them as word gets out, which then further depresses their arm rating as they are unable to throw out runners. Still, I can't fathom Markakis will post a similarly bad arm score going forward, but if he does, that might be grounds to really move him to left field, which generally relies less on arm strength and accuracy.

Second, his range this season was honestly not that bad. Now, it was still pretty much average by UZR which is pretty feh for a corner outfielder, but he did have a +4 from DRS in the category. In both cases, the metrics' measures of his range were the best since his crazy-good defensive season in 2008 (his only net positive defensive season to date, and a 6 fWAR season overall), so it's hard to say that his range was brutalizing his value, even if his overall defensive play was.

A minor note of interest here is that DRS includes a subjective catchall category of "good fielding plays," which are basically runs or partial runs given or taken away from fielders for various actions that do not easily fall into either the "prevented/didn't prevent runners from taking extra bases" category or the "caught/didn't catch a ball in a zone" category. For outfielders, these things might include things like playing caroms well, cutting off balls that might otherwise be extra base hits, and so on. The noteworthy thing here is that Markakis is known to be a solid fielder that doesn't mess around and really does things the right way, whether or not he has the footspeed and instincts to make plays that more naturally-gifted outfielders do. But, for the first time, he got a negative score by DRS in this metric. Now, this may not mean anything other than some anomalous flubbed plays, but I do wonder a bit about it. No real takeaways, but maybe interesting to think about.

Going forward, there's some reason to be optimistic about his defense if you believe that his arm score is an anomaly and he'll continue to be a solid fielder with slightly below average range, as he has been in 2014 and 2015. However, if you believe either that the arm score is an anomaly, or that aging will impact him and start pulling down on his range and placing it back to where it was pre-2014, he might start putting up even uglier overall values.

Looking Forward

Overall, I don't know if my estimation of Markakis was really altered by his 2015, even if the various component pieces were interesting to look at. He did pretty much what I (and lots of others) expected him to, which is be fairly solid and unspectacular, with one great skill (which is the greatest skill of them all) and little else in the way of value other than an ability to stay on the field. So no, the Kake was not a lie. It's just still debatable whether or not the Braves needed any Kake to begin with, which has always been the main sticking point about his signing.

Going forward, I'd guess he puts up something like 1.3 fWAR/600 next year with a 102 ish wRC+, just slight declines off his 2015 performance due to aging. If you're optimistic, I can see you maybe thinking he'll come in somewhere around 2 fWAR/600, with average fielding and a wRC+ closer to 110, but in racking my brain I don't think I'd feel comfortable guessing he ends up anywhere above that.