Pebble Hunting: Jordan Walden and Small Samples

In 2011, the Angels began the season with Fernando Rodney as their closer. Oh, man, was Fernando Rodney bad at baseball a year ago. Rodney was the Angels’ closer, and he was also one of the worst relievers in baseball. He converted his first save, and he blew his second save, and he was replaced by Jordan Walden. Jordan Walden made the All-Star team. The Angels didn’t add a closer in the offseason. The Angels didn’t suggest any sort of closer controversy was brewing. The Angels didn't leave the issue of the ninth inning open-ended at all. Jordan Walden spent his winter chopping wood, shoveling snow, and quietly being the Angels’ closer. “What do you do?” people would ask him at parties. “Awwwww,” he would say, trying to be humble, because nobody likes a boaster, “I’m involved in recreation.” Pressed, he would acknowledge that he closed baseball games for the Angels. Women would casually touch his arm.

He saved his first game, and he blew his second game, and he was replaced by Scott Downs. Fernando Rodney is a closer, and Jordan Walden no longer is. That was very fast! One blown save. Four and a third total innings, and nine baserunners. If his season were a start, it would be Clayton Kershaw’s April 15 start. Very, very fast.

Oh, man, I want to condemn this as a small-sample judgment about Walden. Problem is this: everything is a small-sample judgment about Jordan Walden. There is simply no way to get a reasonable sample for him. It doesn’t exist yet.

Jordan Walden’s career:

1. Twelfth-round pick suddenly becomes top prospect as a starter. Arm troubles end that. (Three years)
2. Mediocre reliever in Double-A. Just not good at all. (50 innings)
3. Suddenly very successful major-league reliever. (80 innings)

It’s not clear how relevant the first step is to the third, or even the second step to the third, or really even the third step to the third. PECOTA hasn’t known what to do with Walden; it forecast a 4.83 ERA and 6.6 Ks per nine for him in 2011 and pegged him for just 0.2 WARP in 2012. In three or four years, Mike Scioscia’s decision might be easier, but for now, Jordan Walden is a man without a past, like Rango. Like somebody in a tougher movie than Rango, but I just saw Rango.

So we have small samples. Lots and lots of small samples. Let’s just jump in and look at some of these small samples to see if they tell us anything about Jordan Walden. (Note: They don’t, really.)

Smallest sample: 2-2 pitch to Brandon Allen, April 26.
Jordan Walden threw a fastball to Brandon Allen with a one-run lead, and Allen hit it out. If Allen had lined it to the first baseman, Jordan Walden would still be the closer. This is the pitch:

It was not where the target was, but I would say it was also not a terrible pitch. It was down in the zone. It was 97 mph. Walden missed his target by perhaps eight inches, the wrong eight inches, but pitchers miss by more than eight inches all the time. Not a great pitch, and also not a pitch that has to be a home run.

Second-smallest sample: The rest of that April 26 appearance.
The tying run was already on base when Allen came up, because Walden allowed a single to B.J. Upton. Upton’s hit was a groundball between the first and second basemen, though, so that’s not really something to hold against Walden.

Mike Scioscia has said that Walden needs to work on his secondary pitches, including his slider, before he’ll get his job back. He threw two sliders to Upton, both for balls, one of which he spiked into the ground. He didn’t throw an off-speed pitch to Allen at all. While the fastball Allen hit wasn’t a terrible fastball, perhaps it was Walden’s overreliance on the fastball that let Allen catch up to it. The less frequently Walden has thrown his slider in his career, the lower the whiff rate on his fastball. And this all might actually be the most important conclusion to draw from this game: Walden’s secondary stuff is getting worse, not better, and he is throwing it less, not more.

First half of 2011: 18 percent sliders, 54 percent strikes, 27 percent whiff rate
Second half of 2011: 13 percent sliders, 42 percent strikes, 21 percent whiff rate
2012: 11 percent sliders, 18 percent strikes, 18 percent whiff rate

I know these are all small samples, but let me stress that the 2012 count is actually absurdly low, like so low that I'm ashamed to have put it there and should probably delete it rather than risk anybody being misled. Walden has thrown about 110 pitches this season, and his slider rate is not even worth observing, and there is no conclusion to be drawn, and there is no suggestion to be made. It’s merely something to watch that Walden has actually thrown more changeups than sliders this year. His changeup is not a good pitch.

Third-smallest sample: Walden’s final save chance of 2011.
I'm pretty sure Jordan Walden would still be the closer if he hadn’t blown a save against Oakland in late September, with the Angels barely in the Wild Card race. That outing (and a poor outing in a meaningless game 162) lifted Walden’s ERA from 2.45 to 2.98, earned him the league lead in blown saves, and hung around stinking for much of the offseason, with Angels fans and rumors reporters identifying “closer” as a place for the Angels to upgrade.

That September blown save also wasn’t as much about Walden’s ability to pitch and more about his ability to field. After putting the tying runs on—weak single, hard single—Walden looked like he'd gotten the game-ending double-play he needed, on a comebacker.  He turned and threw wildly into center field. Still protecting the lead, he threw a two-strike fastball out of the zone away to Kurt Suzuki, and Suzuki slapped it perhaps two feet fair down the right field line. That was the Angels’ season.

Fourth-smallest sample: A little more than a year as a closer.
As noted, Walden blew a lot of saves last year, the most in the American League. Jordan Walden is, therefore, in that strange purgatory that usually holds two or three major-league relievers: lots of saves, not a Proven Closer. Jordan Walden is hanging out there with Frank Francisco and Jon Rauch.

Is that really fair? Probably not, as most Proven Closer discussions aren’t. Walden blew 10 saves. But he also held batters to this line in situations classified by Baseball-Reference as High Leverage:

.222/.313/.319 (170 PA)

Mariano Rivera, in his career, has held batters to a .228/.279/.317 line in high leverage. Even if you believe that closer stress destroys otherwise competent pitchers, Jordan Walden seems to pitch perfectly well under closer stress. He blew 10 saves, eight of them in one-run games, one of them on a throwing error, two of them on inherited runners in the eighth inning. It was too many blown saves, and if it had been two fewer then he would have had as many as Craig Kimbrel.

What is known about Walden is that he throws hard enough, probably strikes out enough batters, and probably gets enough outs for high-leverage innings. It’s possible that this decision to remove him from those innings will cost the Angels. (On Friday, with the Angels up two runs in the seventh inning, Scioscia went to Hisanori Takahashi and Kevin Jepsen because he didn’t consider Walden a high-leverage option. They blew the lead, and the Angels lost.) More likely, though, Walden is one or two good outings from being back in high leverage, and maybe one or two good outings from being back in the ninth inning. It was a small sample to cost him his job. It’ll be a small sample that gets it back. With most relievers, small samples are all there are. For Walden, that’s especially so. Also for Rango.