2015 Visualized: Second Base

2015 Visualized: First Base
2015 Visualized: Catcher

* * *

For the next few weeks, the RotoGraphs staff will devote an entire week to each defensive position, including spotlights on particular players as well as trends throughout the 2015 season. This week, we’re highlighting second basemen.

I don’t claim to be a Tableau (or data visualization) whiz by any means, but I thought it would be cool to visually represent the second base landscape in 2015 — with some analysis sprinkled in.

Steamer and ZiPS represent premier player projection systems; FanGraphs’ Depth Charts combines the two, and the writing staff allocate playing time accordingly. The playing time part is less important relative to the combined projections, as aggregated projections tend to perform better than standalones.

I compared projected wOBA (weighted on-base average) from the preseason to actual wOBA (1) by team and (2) by player within team. Unlike WAR, wOBA is a rate metric, so it does not need to be scaled according to playing time.

First: the difference between a team’s wOBA generated by second basemen and their projected wOBA. Blue represents the highest expected wOBA; yellow represents lowest expected wOBA. The size of the bar above (below) the line represents actual wOBA above (below) expectations.

The measurements aren’t perfect — in an attempt to simplify matters, I included only players who recorded at least 50 plate appearances officially designated as a second baseman. Moreover, many second basemen played multiple positions. Accordingly, I isolated their production at second base and excluded everything else.

As a group, second basemen outperformed their projected wOBA by two points (.311 actual to .309 expected). A few teams who performed better than projected were carried mostly, if not entirely, by one player (Joe Panik of the Giants, Jason Kipnis of the Indians, Dee Gordon of the Marlins). The Mariners (aka Robinson Cano) and Pirates managed to still generate above-average production despite falling short of projections.

And if you want your World Series hot take: Mets second basemen posted baseball’s second-best wOBA, whereas Royals second basemen — or singular second baseman Omar Infante — posted the second-worst wOBA.

Second: the difference between each player’s actual wOBA and his projected wOBA, categorized by team, represented by the distance of each dot from zero. Colors are scaled by actual wOBA, and dot magnitude represents sample size.

A couple of rookies who, per ZiPS and Steamer, weren’t expected to play in 2015, accumulated a decent number of PAs with mixed results:

• Daniel Castro, ATL — .293 wOBA. Exactly matched the playing time threshold of 50 PAs. Young and defensively inclined, Kiley McDaniel thinks he’ll find a role as a utility infielder sometime, somewhere.
• Kelby Tomlinson, SFG — .335 wOBA. If there’s any reason to know his name, it’s because he stole 57 bases across two Minor League levels in 2014. He didn’t really do a lot of that in 2015, though, and his two home runs and .303 batting average in 193 PAs are largely a mirage.

Jonathan Herrera and Ivan De Jesus also made their obligatory underwhelming appearances in 2015, posting wOBAs of .222 and .298, respectively.

For a second straight year, Dee Gordon paid dividends on his average draft position (ADP), doing everything Billy Hamilton owners wanted and much more about eight picks later. It’s great that Jason Kipnis posted baseball’s best WAR for a second baseman (5.2), but his production was fairly underwhelming (nine home runs, 12 stolen bases in 641 PAs) outside a .303 batting average fueled by a .356 batting average on balls in play (BABIP). Logan Forsythe quietly posted a solid, repeatable offensive season for the lowly Rays.

Moreover, a bunch of full-timers — Jose Altuve, Ian Kinsler, Brian Dozier, Brandon Phillips, Kolten Wong — all performed to or slightly overperformed compared to projections.

You can’t talk about second base without dropping Robinson Cano’s name. His power rebounded, sure, but only nominally. The seven-homer increase over 2014’s total — a partial rebound to 2013-and-prior levels — mask a slew of troubling trends, including:

• a career-worst strikeout rate (15.9 K%), coupled with his worst walk rate since 2010 (6.4 BB%);
• a ground ball rate (GB%) that remained north of 50 percent;
• a career-worst contact rate (Contact%), fueled by dips in both in-zone and out-of-zone contact;
• a hard-hit rate that matches his career rate but falls far short of his 2012-13 peak.

As for other “disappointments,” Neil Walker underperformed expectations, but he basically matched his career numbers and fell short of a projection buoyed by a 2014 spike. With a .251 wOBA across more than 400 PAs, Chris Owings experienced the dreaded sophomore slump and has betrayed his former top-prospect status. He simply doesn’t make enough hard contact to be a power threat, and it’s some semblance of power that Owings needs to offset is atrocious plate discipline. The speed is modest but not special at middle infield.