Joe Nathan saved his 34th game of the season on Monday, and that's good. It makes him fourth in the league in saves, behind Mariano Rivera (who is the best of ever), Craig Kimbrel (who is the nastiest, disgusting, filthiest pitcher on the planet) and Jim Johnson (I'm told he cooks some really good casseroles).
I'm getting this saves thing out in the open early because I'm not going to talk about them so much anymore. I'm of the opinion (and it's not a rare one) that the save, as a statistic, has done more damage to baseball than any other. They've changed the way baseball is managed, they've changed careers, and they've changed the economics of relievers. What's worse, people who don't like stats ascribe some kind of magic to them.
Besides, they obfuscate the point. In this case, the point is a question. That question is, is 2013 Joe Nathan's best season?
Nathan's had a very long, very celebrated and very, very good career. This is his 13th season (not counting 2010, which just didn't happen for Joe). He's accumulated 25.4 bWAR over those 13 years. His career ERA is 2.79, and a big chunk of his career was spent in a very generous run-scoring environment.
Of course, this year, Nathan's holding up a 1.54 ERA. He's only had a better ERA once, in 2008, when he put up a 1.33. ERA is a limited, but I actually like it for relievers like Nathan who tend to enter an inning clean, with no runners on base. At some point, the only job a late-inning reliever really has is “Don't let these guys score” and ERA pretty much tells you how a guy does at that. Other than the objectivity in what's an error vs. what's a hit and how it still depends largely on defense even outside of errors. And it's not context-neutral like adjusted stats are but other than that, I like it ok.
Anyway, the point is his ERA is the second best of his career so far.
Strikeouts haven't been as kind to Joe this year as they have been before. He's striking out 9.26 per nine innings pitched, which works out to 26.7% of batters faced. It's impressive, but 35 qualified relievers have higher rates. The league has gone strikeout crazy for relievers. It's actually the 8th best season Joe has had- his worst since 2003. His best K-rate was 2006, when he struck out 36.3% of batters that stood in against him.
That would put him 7th in the league this year in K-rate, that 36.3 would. Aroldis Chapman, Greg Holland, and Craig Kimbrel are all striking out over 40% of opposing batters. Craig Kimbrel is a disgusting beast-man. He is so good at baseball.
There's also the issue of walks. This season Joe is holding at 9.4% walks, which works out to 3.28 walks per nine innings pitched. It matches his career average, actually, but that obscures the fact that- no, you're not having deja vu- it's his worst rate since 2003. Just last year, Nathan walked only 5.1% of opposing batters. If you noticed that's pretty close to half of what he's doing this year, then you're pretty observant.
So, if Nathan is striking out less guys (and strikeouts are a very important part of Nathan's game) and he's walking more guys (and avoiding walks is very important), how is he keeping teams off the scoreboard the way he is?
The answer seems to be weak contact. Hitters are only hitting 154 against Nathan this year, which is even better than what teams are hitting against Craig Kimbrel, who is some kind of monstrous demon relief monster who is not human. That's good, because limiting hits is part of the point of pitching. The troubling thing, though, is Nathan's BABiP- Batting Average on Balls in Play, which simply looks at how often hitters making contact results in a hit versus an out- is 205. His career average is 252,and league average is normally around 300.
This is one situation where a low number is worrisome, because there's two monsters we all fear- regression, and Craig Kimbrel. If your BABiP is low, it's safe to bet that it's going to work it's way up; which means more hits fall, which means more runs come in.
Maybe it's a skill, but Craig Kimbrel, whose arm was actually created in a top-secret DARPA weapons lab, sports a 280 BABiP. Aroldis Chapman's is 284. Greg Holland's is 303. Those are pretty much the three most filthy, nasty, most evil relievers in baseball, and they aren't displaying a skill for holding a low BABiP.
There's more to BABiP than the raw number, though - you can normally break it down into groundball, flyball, and line drive rates to explain it. Line drives fall in for hits a drastic majority of the time; groundballs turn into hits a little bit more than average, but they make up for that by being singles most of the time. Flyballs turn into outs a lot, but they also turn into home runs a lot, and home runs are really bad. The point is, there's a rule of thumb out there, that you can take a pitcher's line rate, LD%, add 12 to it, and have a pretty good approximation of what his BABiP 'should' be.
Nathan's LD% this year is 22.3%, which means you could expect his BABiP to be around 350. That would essentially double the amount of hits Nathan has allowed, and then add another another 15 or soon top. It's readily apparent that kind of regression would be disastrous.
Left with all that, there's pretty much no defense for saying this is Joe Nathan's best year ever. He's had better raw results (2008 jumps off the page, when he pitched 67.2 innings at a 1.33 ERA) and better peripherals (in 2006, he struck out more guys, while walking less, with a 276 BABiP).
That doesn't mean we can't enjoy it, and no matter what we think of closers in general, I think everyone can agree to be happy we have Joe and thankful for the year he's given us.
And we can all be afraid of Craig Kimbrel.