Felix Hernandez in Safeco Field
I am endlessly, helplessly fascinated by home/road player splits. All splits are are numbers, and it's easy to end up drowning in baseball numbers if you aren't careful, but what splits reflect are differences in playing environment, and their effects. Baseball is the only major sport where there are significant differences from stadium to stadium, and if you sit back and think about that for a few moments, it's wild. Why on earth is baseball like that? And because baseball is like that, how can we see what this means? Hello, home/road player splits. There's no better way of measuring what a certain environment does for or against a certain player, and every player is impacted differently.
I am not, however, fascinated by single-season home/road player splits. Or two-season, or even three-season home/road player splits. I love splits, but what splits do is cut samples in half, and we already have myriad problems with small sample sizes in baseball statistics. It's hard to tell much from small-sample player splits; better to collect a lot of data. Of course.
There are only so many opportunities to see what a park does to a player's numbers over a long period of time, because so many players move around and that does us no good. Best to analyze those who've been in one place for a while, so we can observe that, say, Michael Young has an .854 career OPS in Texas over 874 games, and a .737 career OPS on the road over 866 games. Clearly, Young has been given a significant boost by his home ballpark. We couldn't say that conclusively about, I don't know, Craig Gentry. Gentry would be helped by the park, but we can't say by how much.
Anyway, to bring this all back, home/road splits are interesting, provided you have enough data. What we've talked about recently is how much better Jason Vargas has pitched at home than on the road. That's not a surprise, because he's a left-handed flyball pitcher, and Safeco wraps those guys in a cozy warm blanket. But someone else who's been here for a long time is Felix Hernandez, and his career splits aren't nearly so distinct. I get the feeling like I might have written about this before, but, shit, I've written about everything before, and I've never written about this with as much data as we have today. Onward!
Felix debuted in 2005 and has made 220 starts since, throwing more than 700 innings in Seattle and more than 700 innings away from Seattle. The sample sizes aren't infinite, but they're large, and large enough for a high degree of comfort. Let's look at how theoverall have pitched at home and on the road since 2005:
This is sort of Felix's peer group. You can see substantial differences, which of course you expect to see. The home ERA is 85% of the road ERA. The home RA is 82% of the road RA. 90% for FIP, 96% for xFIP, 92% for wOBA, 95% for BABIP. You know what Safeco reduces? Runs, and the things that lead to them. Hits, extra-base hits, home runs, whatever, Safeco doesn't like them, and Mariners pitchers have predictably benefited over the years.
Now let's look at Felix since 2005.
Jason Vargas is a left-handed flyball pitcher, so he's well-equipped to take advantage of what Safeco has to offer for pitchers. Felix is a right-handed groundball pitcher, so he's less well-equipped. He faces more lefties and allows fewer balls in the air, and as a result of those things and probably other things, his splits aren't nearly so dramatic. I wouldn't even call them dramatic at all. You see better performance at home than on the road, but everybody performs better at home than on the road.
Since breaking in, Felix's home ERA is 98% of his road ERA. On average over the entire American League since 2005, home ERA is 90% of road ERA. Felix's home wOBA, maybe the best stat here, is 98% of his road wOBA, whereas for the league it's 96%. And Felix has actually allowed more hits per ball in play at home. Based on much of the evidence, Felix has taken less advantage of his home park than the average pitcher, and Felix has pitched in a more run-suppressing home environment than the average pitcher.
Wow, look at all of these numbers! If you've gotten this far, you are very dedicated to this blog post, and I appreciate it. We're almost done. What's the conclusion here? There's no one conclusion, and a conclusion was just reached in the previous paragraph. Felix has been better at home than on the road, but not greatly so, and you wouldn't think from looking at the home/road splits that he's played in a home park with dimensions that more and more people want changed with every passing day.
This isn't about Felix "underperforming" at Safeco or anything. If at any point while reading this you developed negative or critical thoughts, your neurons fired wrong. What this really shows is that one should never apply blanket park effects to all who play within it. One might look at Felix Hernandez's numbers and say, "they're good, but he pitches half the time in Safeco Field." That's true, he does do that. But he doesn't derive nearly the same benefit from Safeco as other guys do, so the "but he pitches half the time in Safeco Field" qualifier doesn't mean nearly as much. With Felix, it's safer to apply no park effect at all than to apply the standard Safeco factor. He does surely benefit in some ways, and I'm guessing he's had a handful of would-be dingers hauled in on the track, but Felix isn't Jason Vargas. How many times do I have to tell you to stop saying that Felix is Jason Vargas? They're completely different!