Eduardo Rodríguez has had a strange season

In some respects, this season was even better than last for Eduardo Rodríguez. His 3.67 RA9 is the lowest of his career, and for the first time he has made 32 starts. His 191 ¹⁄₃ IP blows away his previous high of 137 ¹⁄₃ IP. His 6.2 bWAR is more than double his previous high last season, and it is tied with Patrick Corbin for fifth-best in baseball. To really put that into context, Cy Young candidate Gerrit Cole has 6.1 bWAR! So has Rodríguez finally become an ace? Well...

Rodríguez’s peripherals are unremarkable for someone with a high bWAR. His walk rate is mediocre, and even though his 24.8 K% is better than the league average, it is a far cry from, say, Cole’s record-breaking 39.1 K%. To his credit, however, he has done a great job at limiting the long ball before you even start factoring in his ballpark.

Looking at the versions of WAR at FanGraphs and Baseball Prospectus, E-Rod’s season does not look nearly as impressive: 3.4 fWAR and 3.2 WARP. That is still pretty good, of course, but it does not make him look like a top-10 pitcher in the league. So why do these WAR values differ so much from the one at Baseball Reference? It can be really confusing for someone who goes on FanGraphs and sees that his 3.53 ERA and 3.91 FIP do not differ by too, too much.

Let’s get the biggest differences out of the way first for those who might not be familiar. B-Ref uses RA9 as the base for its WAR calculations, not ERA, because it makes separate adjustments for defense. FanGraphs uses FIP, and BP uses the most advanced pitching metric in the public sphere, DRA. Let’s set DRA-based WAR to the side for the moment and go into the differences between bWAR and fWAR in more detail.

Part of B-Ref’s WAR calculation is taking the pitcher’s RA9 and making adjustments for quality of opponent and quality of the defense. FanGraphs does not do either of that. I can’t speak to why they do not adjust for quality of opponent, but they do not do the latter because the whole point of FIP is to ignore defense, as it is based in DIPS theory.

It is really important to point out that the FIP value you see on any of FanGraphs’ stat pages is not the exact value that is used to calculate fWAR. They use something called FIPR9. First they modify FIP to count infield flyballs as strikeouts. Then, since FIP is on the ERA scale, they adjust it to the RA9 scale, which is pretty easy to do. You just take the league difference in RA9 and ERA and add in the difference.

Because I am a masochist, I calculated Rodríguez’s FIPR9 at 4.34, which is 0.43 runs higher than his FIP. Remember that his RA9 is 3.67. His RA9-FIPR9 is 0.67, compared to his ERA-FIP of 0.38. So how much of a difference does that make? It roughly brings Rodríguez’s 6.2 bWAR to 4.8, so it’s a lot but it is still over a win off from his 3.4 fWAR. We would have to get more into the guts of how these WAR models differ in order to cover that difference, and I don’t want to do that anymore than you want to read about it.

So what does this all mean in terms of how we evaluate Rodríguez’s season? It is tough to say. He did not benefit from any BABIP luck or high strand rates. His groundball percentage did shoot up, but his infield defense is not great, and there were no major changes in his pitch usage.

As for DRA, it is a bit of a black box, so it is harder to break that down, but it is interesting to see that it is nearly identical to Rodríguez’s 4.34 FIPR9. It also leads to a 3.2 WARP that is nearly identical to his fWAR, too. We can see that DRA determined that he really lucked out with how few runs he gave up given the quality of contact he allowed. One would think he was getting lucky in high leverage situations and RISP position, but just the opposite was true!

At worst, Rodríguez was just as good as last year, but other than that, it is hard to conclude that he has become the ace that his bWAR would indicate. His peripherals have not changed, and he is not doing anything differently. It will be interesting to see how he does next season.