Nathan Eovaldi's 2014 season and fielding-independent pitching

Over the course of the second half of the season, it has become common fare among Miami Marlins fans to pick on Nathan Eovaldi, who has struggled to a 6.04 ERA over 67 innings. Perhaps as part of the team's efforts to watch their young pitchers' workloads, the Marlins are considering shutting down Eovaldi as a starter for the rest of the year. But much of the reason for this potential move is to get Eovaldi time to work on potential problems with his work, as evidenced by his poor second half ERA. Here is president of baseball operations Michael Hill on the subject of Eovaldi.

"A lot of times when you think of starters moving to the bullpen, it's when there are short glimpses with velo, and then it falls off," Hill said. "It's not a falling off situation for him. There are probably some mechanical issues with him that we need to address. He needs to make better pitches and improve the consistency of his secondary pitches to allow him to be effective. Right now, he's definitely a starter."

They can see that Eovaldi's velocity remains intact. Indeed, in the first half, Eovaldi averaged 95.5 mph on his fastball according to Pitch F/X data available on FanGraphs, while in the second half he has thrown it at a 95.6 mph mark. Other sources essentially corroborate the idea that his fastball velocity is unchanged.

Hill mentions the possibility of something mechanical. Earlier, there was discussion of something potentially mechanical that pitching coach Chuck Hernandez noted and presumably fixed to great success. But just a couple of starts later, Eovaldi returned to having problems and giving up bunches of hits, and now these concerns have crept up again.

But here's the interesting thing about his first and second halves: they only really look different by two measures.

He has definitely been a bit worse in the second half than the first. But the differences in terms of strikeouts, walks, homers, and batted ball types is very small. The difference in those FIPs is the difference between Zack Wheeler and Stephen Strasburg this season, which is to say they are different, but not dramatically so. Both have been above average pitchers this season, one more than the other, but both generally qualify as "good."

The difference between Eovaldi from the first half and the second half thus far has been the BABIP, which has shot up from .300 to .354. If you are a reader here at Fish Stripes, then you probably already know that the rate of hits on balls in play are a fickle thing. It has been well-researched that pitchers have a hard time repeating performances on balls in play from season to season, and that values from one year are not great at predicting the next. The reason is simple: there are a lot of factors that affect BABIP, from the quality of hitters faced to the pitcher's play to the quality of the defense behind him. It turns out BABIP is better correlated to defense year to year than it is to pitcher!

This is not to say that pitchers have no control over balls in play, or that a pitcher who performs poorly on balls in play is "just unlucky" and not necessarily pitching poorly. Pitchers could have pitched poorly, and they do indeed have control over balls in play. But at the Major League level, that skill has essentially leveled out, and professional baseball has essentially weeded out anyone who cannot allow an acceptable amount of hits on balls in play. So a player's performance year-to-year or half-to-half is not a great measure of future performance, because the skill level at the bigs is very close to equal across the board.

You can see this trend over the long run. The league average BABIP since 2011 is .293. I looked at all pitchers with at least 400 innings pitched since 2011, and separated the 30 best ERAs among that group and the 30 worst ERAs. The 30 best ERAs averaged an ERA of 3.15 during that time period, and their average BABIP was at .285, which is eight points lower than the league average. The 30 worst ERAs with at least 400 innings had an ERA of 4.47, significantly worse than the best ERA group. But they also had a .299 BABIP, which is only six points higher than the average.

This kind of matches up with the ERA changes, and it also is a biased study, because BABIP contributes to ERA directly. The important thing to note is that the difference between the performances of the 30 best and 30 worst pitchers who still pitched plenty in the majors was around 15 points of BABIP. The difference between Eovaldi in the first and second half was far too large to not have caused any change in any of his other skills. If it really was a skill change, as indicated by perhaps a mechanical problem, then we would expect to see changes in his strikeout or walk rates that were perhaps a little more egregious than the ones we saw. It seems unlikely that his mechanical flaw is only affecting the one thing that happens to be most volatile in pitchers.

This does not mean that Miami should not take a careful look at Eovaldi's mechanics and see if there is not some flaw in his most recent work. But here is where the numbers and scouting should work together: if the coaches and scouts do not see a difference in his mechanics or throwing motion, they should be encouraged to know that this could easily be random variation that is not indicative of a loss of skill. The coaching staff should not be hunting down a reason for Eovaldi's poor second half. If one is apparent, then they can fix it. But if there is not one, that is not necessarily a cause for concern either.