Finding the real Eduardo Rodriguez

With David Price avoiding major elbow surgery but also expected to start the year on the disabled list, the competition for the back end of the Red Sox rotation received a temporary stay. Drew Pomeranz, Steven Wright, and Eduardo Rodriguez were battling for the fourth and fifth sports, but now it seems they’ll all have a role until Price is able to return.

In his first two seasons of major league action, the 23-year-old Rodriguez has eclipsed both 100 innings pitched and 20 games started. That’s not bad considering how slowly some teams take the development of their young pitchers. Now entering what will be his third season in the bigs, Red Sox skipper John Farrell is bullish on his ability and had this to say him, from Christopher L. Gasper in the Boston Globe:

“He burst onto the scene where I think his talents really allowed him to have the success he did,” said Farrell. “But much like every new guy that comes into the league, the league catches up to you, so it’s going to come down to consistent execution, to pitching to a scouting report . . . So, there are no secrets. He is into his third year in the big leagues. Guys know what to expect. He is a known commodity. But he has got as good of stuff as anybody in our rotation.”

Here’s the thing: While it’s true that hitters may have a consistent approach in how to attack his stuff, with regards to evaluating him as a pitcher, Rodriguez is most certainly NOT a known commodity. His year one and year two statistics have been very different, both by process and outcome.

First, the outcomes. Depending on which statistics you cite, Rodriguez was a very different pitcher in 2015 than he was in 2016.

Using both traditional ERA and FIP, Rodriguez was demonstrably worse in his second year. There was barely a difference in the year one numbers, and year two saw a gap but not a huge chasm. Based on ERA-FIP his 2016 ERA might have been a little out of his control, but he took a step back regardless of any bad luck.

Navigating over to Baseball Prospectus and looking at their proprietary statistic DRA (Deserved Run Average) tells a different version of what happened. This story is one of a pitcher who got significantly better from year one to year two. Whichever narrative you prefer can be validated by the various iterations of Wins Above Replacement produced by FanGraphs, Baseball Reference, and Baseball Prospectus.

The FIP-based WAR of FanGraphs and runs allowed based version of Baseball Reference had Rodriguez as a more valuable pitcher in 2015. Baseball Prospectus’ WARP, which incorporates DRA, echoes that statistic’s valuation that Rodriguez was a better pitcher in 2016.

Now we turn to the process -the components of how he pitched- to see how Rodriguez changed and saw his outcomes experience such huge fluctuation. Let’s look at the important rate stats first:

Whoa, did you see that?! Here, just in case you missed it:

Data via FanGraphs

The league average ground-ball rate in 2015 was 45.3 percent, so you couldn’t really classify Rodriguez as a “ground-ball pitcher,” but he at least induced grounders around a league average rate in year one. Boy did that change. In 2016 his ground-ball rate fell 11.4 percentage points and his fly-ball rate rose 12.6 percentage points. While you couldn’t definitively call him a ground-ball pitcher in 2015, you could absolutely call him a fly-ball pitcher in 2016. His fly-ball rate was 11.4 percentage points above league average.

The newfound fly-ball profile was evidenced in how he attacked hitters with the four-seam fastball. Here’s his fastball pitch percentage heatmap with 2015 on the left and 2016 on the right:

While overall Rodriguez threw fewer four-seam fastballs, instead mixing in a higher percentage of two-seamers, he made a clear effort to work the four-seamer up in the zone. His fly-ball percentage on the pitch made a huge jump from 27.5 percent to 52.4 percent, as did his swinging strike rate, which was up to 9.7 percent from 5.6 percent.

The concern in this strategy would be giving up more long balls. Rodriguez saw his HR/FB ratio drop by almost a little more than a percentage point on the pitch, but since he allowed so many more fly balls overall, he did give up more home runs per four-seamer thrown.

He also changed his approach with the slider. Here’s that pitch percentage heatmap (again 2015 on the left, 2016 on the right):

In his first year, Rodriguez used the slider in a typical fashion, low and away to the glove side. In 2016 it seems he started to do two other things with the pitch; throw it up in the zone like he did with the four-seamer and try to back door right-handers. His usage didn’t change the swinging strike rate on the pitch much, but hitters swung at the pitch far less often in 2016.

While he did vary where he threw the slider in the zone, batters refused to bite when he tried to back door the pitch. He changed his location and the pitch generated less contact, but it also generated fewer swings. The slider itself didn’t give up an increased fly-ball rate but it seems that hitters just paid less attention to it and waited for that four-seam fastball, which had a 6.2 percentage point increase in swing rate.

So Rodriguez’s new tendency to work up in the zone caused his reinvention as a fly-ball pitcher. But as it turns out, the effect was only seen against right-handers. The new plan of attack actually caused his fly-ball rate against lefties to drop 3.9 percentage points, whereas it rose 17.1 percentage points against righties. The change evened out his reverse platoon split overall -from a wOBA 58 points higher against left-handers in 2015 to just six points in 2016- but made him more vulnerable to right-handed power. That difference is illustrated by the kind of the home runs he allowed in each season.

In 2016, 10 of the 16 home runs Rodriguez gave up were at home, all but one was to the left side, and all but two were to hitters batting right-handed. In 2015, nine of the 13 home runs he allowed were at home, but eight were to center or right field and seven were hit by lefties. This is important because his home park is Fenway, a place were right handed home run hitters thrive. Here’s the visual of his home runs allowed at Fenway in each season compared to a more neutral park (in this case Kansas City):

via Baseball Savant

via Baseball Savant

Pitching in Fenway might have cost Rodriguez four home runs in 2015 but eight or nine in 2016. Surely it’s just scratching the surface of why his DRA improved so much and moved in such a drastically different direction from his ERA and FIP, but it’s a reminder that context is incredibly important. That’s DRA’s aim, to take context into account and quantify.

The increase in ERA and FIP coincided with Rodriguez becoming a totally different pitcher. DRA would argue that when you parse the circumstances of his pitching he actually became a better pitcher. The problem is that his home ballpark will continue to be Fenway Park and this new approach as a fly-ball pitcher, while certainly with it’s merits, may be detrimental because of his environment.

Through two seasons we’ve seen two different pitchers. It will be interesting to see on if he continues to elevate his fastball to maintain 2016’s fly-ball approach or if he tries to recapture some of the lost ground-balls from his first year. Who is the real Eduardo Rodriguez? We’ll see.