As you know, the Cardinals went into the season with a lot more depth than usual. The skilled infield positions were particularly crowded. The plan going into the year was for Jhonny Peralta, Matt Carpenter, and Kolten Wong to be the assumed starters, with Jedd Gyorko and one of either Aledmys Diaz or Greg Garcia serving off the bench.
Of course, that plan went out the window when Jhonny Peralta got hurt in Spring Training, with Diaz taking over the starting duties in Peralta’s absence. Garcia was still mostly stuck in Memphis after the acquisition of Ruben Tejada, but after Tejada was sent packing, he’s stayed around.
The depth ended up being necessary, as Carpenter had his oblique problem and Peralta re-aggravated his thumb shorty after returning. Kolten Wong of course has struggled to the point of losing an assumed starting job. As a result, Gyorko and Garcia haven’t been buried on the depth chart, and have instead played more important roles than what was assumed going into the year. To start off, here’s the two’s core stats in 2016:
Garcia has had low power numbers and a below average BABIP. He’s had a high BABIP until a cold streak recently. That Carpenter-like walk rate gives him a very high floor, and when you add a lower than average K% and what was an above-average BABIP to the mix, well, that’s how you get a VEB nickname like "OBP-monster".
Gyorko’s power has been his calling card, with an ISO over .200. His walk and strikeout numbers are right around average, but he’s carrying an extremely low BABIP that brings his total line down to a 107 wRC+.
In such small sample sizes, those ISO and BABIP numbers are not very reliable though. Can Garcia get back to an above-average BABIP, and can Gyorko keep up that ISO, perhaps with some BABIP regression? That’s what I wanted to find out. Luckily, recently I have developed some stats using the Statcast data found at BaseballSavant.com to give me a better idea.
I introduced an xBABIP calculator three weeks ago, which used all recorded Statcast data available to find the average BABIP of each combination of Exit Velocity and Launch Angle, two batted ball stats now tracked by Statcast. That allowed me to peruse Matt Holliday’s batted ball profile, and come to the conclusion that he had been very unlucky on balls in play.
The following week, I did the same with homers, finding the average home run per batted ball rate for each unique combination of Exit Velocity and Launch Angle. Using this data, I was unfortunately very disappointed in Kolten Wong’s complete lack of home run power in 2016. I advise reading both articles to familiarize yourself with the terminology used here. Anyway, using the statcast stats I developed recently, here is how Garcia and Gyorko’s contact quality grades out. For convenience and better understanding, I also listed the league average for each stat. First, we’ll start with home run power:
In case you haven't read the article on Kolten Wong and homers, HRPBB stands for Home Run possible batted balls, which is any batted ball between 18 and 42 degrees, inclusive. Angle-based xHR/HRPBB% takes each of those batted balls in that range, and compares each batted ball's angle to the league average home run rate for balls hit at that angle. Velocity-based HR/HRPBB% is calculated the same way, but using the velocity of each Home Run possible batted ball. Total xHR/HRPBB takes the specific angle/velocity combination (or vector) into account. I also track the percentage of Home Run possible batted balls that are below 90 MPH, because those batted balls only very rarely leave the park.
Gyorko shows a very above-average ability to hit homers, though not quite as good as he’s shown. That’s due to both hitting home-run possible batted balls hard, and by hitting them at more optimal angles than average. Garcia also hits at better angles for homers than average, but doesn’t hit them near as hard, leading to a much lower expected home run rate. The rate at which he hits home-run possible batted balls under 90 MPH is also concerning.
Now let’s look at BABIP:
Garcia manages an above-average total xBABIP due to fantastic angle-based xBABIP. Angle-based xBABIP is calculated by taking the angle of each player’s batted balls, and finding the average BABIP of every ball hit at that angle. velocity-based xBABIP is calculated the same way, but only using velocity. A Graph of BABIP by angle and by velocity is available in the Matt Holliday post linked above. His swing produces much weaker than average velocities, but they’re sent out at good enough angles to make up for it. Part of that is the fact that he has yet to hit a pop-up this season, which are as close to automatic outs as batted balls can get when only considering angle. For a closer inspection, I checked out his angle profile at Baseball Savant:
First, as a reminder, batted balls are classified by Statcast as follows: below 10 degrees are grounders, between 10 and 25 are line drives, 25 to 50 are fly balls, and above 50 are pop-ups. He has his highest spikes at 20 degrees, 5 to 10 degrees, and -5 to -10 degrees. The positive degrees listed there all support high BABIP’s, even without excellent velocity. The -5 to -10 degrees angles aren’t great, but they’re ok, and he has very few batted balls higher than 30 or lower than -10, both being ranges where BABIP is weak.
Can Garcia continue to hit a large majority of his balls in play into high-BABIP vectors? The tightly-packed nature of his angle profile would seem to indicate that he’s good at squaring the ball up, which seems like a repeatable skill to me. It also makes sense that a hitter could go on a hot or cold streak in terms of squaring the ball up though, and maybe Garcia is just on a hot streak. It surely stabilizes a lot quicker than BABIP though, so either way this is a good sign. The important thing is that it seems he’s been unlucky on balls in play..
Gyorko also looks to have been unlucky, with a BABIP of .249 compared to an xBABIP of .263 However, my calculation doesn’t take a player’s foot speed into account yet, and Gyorko is a below-average runner. He doesn’t get shifted against very often at least, at 9.8% compared to an average of 15.2%, so he could gain back a few points that he lost from being a weak runner. Gyorko’s BABIP might be expected to shift upwards, but not by much.
Jedd’s BABIP profile means he has to make up for it with power, and he does. Above-average angle optimization and velocity give him a strong score in total xHR/HRPBB, and he’s been above-average at hitting it in the range where homers are possible as well. Like with Greg, let’s look at Jedd’s angle profile:
Gyorko has a large spike right around the optimal angles for homers, 27 and 28 degrees. That fuels his higher than average angle-based xHR/HRPBB. He also has a large spike right around -5 degrees, which isn’t bad if he’s hitting them hard, which in general he is. However, he also has smaller but notable spikes at 45 and 60 degrees, which aren’t going to do anything for his power or average.
The slightly wider range of Gyorko’s angles could imply that he has less ability than Garcia at squaring the ball up, despite the fact that his average Exit Velocity beats Garcia, 87.3 MPH to 84.5 MPH. That could imply that Gyorko has some combination of more raw strength, and/or a swing that better taps into his strength than Garcia.
So, one has strong BABIP and weak power, the other vice-versa. Who has the better overall contact quality? Before we get into that, I’ll preface that it’s probably not all that important. I didn’t write this post to decide who was better, but because of the contrast in their styles. Garcia is a lefty, Gyorko is a righty, the decision of who to play should probably just depend on the pitcher’s handedness. Anyway, using the same Statcast data and the same methods, I calculated an on-contact xwOBA stat (on-contact expected weighted on-base average, to read about wOBA check here).
However, just looking at contact isn’t fair, because Garcia is clearly the better of the two in non-contact results, with a higher walk rate and lower strikeout rate. So using each player’s non-contact stats, as well as xwOBA on contact, I also calculated an xwOBA. Here is the results:
Gyorko has the better contact quality, but is slightly worse than average in terms of non-contact (average non-contact wOBA is .200). Garcia’s impressive walk rate and below-average strikeout rate allows him to take the lead, despite below average overall contact quality (aveage on-contact wOBA is .367). Jedd’s xwOBA very nearly matches his wOBA, whereas my system sees Garcia as being 24 points of wOBA better than he has been.
Garcia and Gyorko are quite the opposite hitters: one hits for BABIP, the other for homers. Gyorko has the better overall contact, but Garcia is much better in terms of drawing walks and limiting strikeouts. Garcia comes out to be the better hitter according to this analysis, and hey, he’s probably a better defender than Gyorko too.
Again though, this isn’t a competition. This is close enough that deciding who should start at third or second or who should pinch-hit should just depend on pitcher handedness, especially since I still don’t know how predictable these stats are yet. With league average wOBA at .318 this year, both have displayed contact quality that, with their K and BB numbers, make them look like above-average hitters who can play skilled infield positions. With Wong’s struggles, this analysis makes me hope that Matheny looks to one of these two more often, to see if they can keep it up.