On Jordy Mercer and Jung-Ho Kang

It would be a tremendous understatement to say that Jordy Mercer's 2015 hasn't gone as planned. This isn't the first time the former Oklahoma State Cowboys closer has struggled to open a season--his batting average was sub-Mendozan for much of the early part of 2014. When one's hitting performance elicits non-negligible nostalgia for Clint Barmes, something has gone very wrong.

Last year, the Pirates' alternatives to Mercer at shortstop were limited--as demonstrated by the depressing late-season cameos of such luminaries as Jayson Nix (currently hitting a robust .190/.203/.302 for the Phillies' AAA affiliate) and Michael Martinez, the Mercer/Barmes duo was pretty much the only show in town. And despite Clint Hurdle's well-chronicled affection for the slick-fielding Barmes (all the Clints gotta stick together), he kept the faith with Mercer and was ultimately rewarded with a solid year.

Unfortunately for Mercer, the emergence of Korean shortstop Jung-Ho Kang has made his slow start in 2015 quite a bit more costly. This wasn't an unforeseen conflict--as soon as the Pirates unexpectedly posted the winning bid for Kang, questions existed regarding how he might fit into the left side of the Pirates' infield.

The Mercer/Kang situation isn't a zero-sum game--both Kang's ability to play third base and Josh Harrison's general versatility allow Mercer and Kang to play side-by-side--but it's clear that both can't really receive starter-level playing time simultaneously without significantly impacting Harrison, Gregory Polanco, or Neil Walker. And while having too many talented shortstops seems like baseball's equivalent of a first world problem, Hurdle does appear to have a decision on his hands when it comes to who should start at shortstop against right-handed pitchers.

The Pirates opened the season with Mercer as the starter, due presumably to some mixture of the advantages of incumbency, his strong performance in 2013 and 2014, and caution regarding Kang's need to adjust to American pitching.

Well, Kang's adjusted. He's fulfilled his stated desire to face Aroldis Chapman, doubling off of him on a pitch that hit 100 mph. He's done this to a 96 mph Jason Motte fastball. According to Pitch(fx), he's been the sixth-best hitter in the game on four-seam fastballs. So the bat speed is (emphatically) there. Which is why it's surprising that people keep grooving him fastballs:

The raw power's for real, too--there have been 272 hitters for whom Baseball Savant has at least 50 batted-ball velocity data points this year. Among these hitters, Kang's average FB/LD velocity of 95.74 MPH ranks 21st, surrounded by guys like Chris Carter, Jorge Soler, and our own Starling Marte. As you can see below, Kang (the red dot) might even have more in the tank in terms of power production when you consider how the ball flies off of his bat:

It makes you wish that he elevated the ball a bit more often--Kang hits 55.9% of his batted balls on the ground, good for 19th in the league (right behind Marte). This might be part of why his ISO is lower than you'd predict.

If you want to take a higher-level view, Kang's .280/.351/.409 batting line has produced a 115 wRC+, fourth among all shortstops. He's also tied with Jose Reyes and Danny Santana to contribute the most baserunning value (1.8 runs) of all shortstops. Put it all together, and he's ninth in WAR among shortstops, at 1.2. He's almost produced enough to justify his contract already.

Meanwhile, Mercer's been something less than stellar. He's hitting .225/.273/.294, and while his strikeout and walk rates are relatively unchanged his isolated power has halved since last year. He's been below replacement level, with -0.2 WAR.

So, there should be no contest about who starts at shortstop, right? Kang all the way--all of the various statistics I word-vomited out over past several paragraphs would seem to indicate that he's the superior choice.

Maybe it's not that simple. Here are their respective rest of season projections (blended ZiPS and Steamer), prorated to 600 plate appearances:

Maybe it is that simple.

You might say that projection systems aren't going to properly capture the defensive gap between the two players, except we don't have much reason to think Kang is anything less than an average defensive infielder. You might say that projection systems are going to be systematically less accurate for international players with short track records, but Kang has already pretty much eliminated the 'worst case scenario' portion of his projection--the one in which he lacked the bat speed to hit American pitching--and the projection above still includes significant regression.

The conclusion here is that Kang appears to be our best choice at shortstop moving forward. Mercer will still start against lefties, with Kang moving to third.

Does hope remain for Mercer, though? For one, I think that both the projections above are a bit bearish (at least in terms of WAR)--Mercer, in particular, has always rated much better using DRS than UZR, and if you believe he's a plus defender he might be closer to 1.5 WAR/600. Mercer's substantial platoon split also means that starting only against lefties could improve his batting line.

In short, he's a useful dude to have around.