Rebuilt Yankees 'pen is game's scariest unit

New York's relief corps' 2.41 ERA past 30 days is best in Majors

August 18th, 2017

A month ago today, the Yankees made the trade everyone expected, upgrading their third-base situation by acquiring from the White Sox. As the biggest name in the deal, and with third base being New York's biggest trouble spot, it was "the Todd Frazier trade." Even here, we understandably relegated the other pieces headed to the Bronx to the "New York also receives …" section.
A month later, it's not the Frazier trade any longer, if it even was in the first place. It's the "Tommy Kahnle and trade," and it's one of a few reasons that the Yankees have turned a bullpen group that was in disarray in June into what may be baseball's most dominating unit down the stretch. We learned last year that bullpens are what matter in October, even more than starters. If the Yanks can hold onto their current 3 1/2-game American League Wild Card lead and make it to October, this is going to matter, a lot.
New York was expected to have a good bullpen this year, of course. Any unit fronted by and was going to get off to a good start, and it did. Over the first two months of the season, Yankees relievers had baseball's fourth-best ERA (3.08), and baseball's best wOBA (.257; wOBA is just like OBP, except it gives more credit to extra-base hits rather than treating every time on base equally).

But the month-long stretch from June 15 to July 15 was, to put it charitably, not great. That ERA shot up to 4.81, the 10th worst in that span. The wOBA jumped to .323, tied for 11th worst. Betances (7.84 ERA) ran into massive control issues, walking 14, the most of any reliever; allowed 62 percent of the runs he gave up as a Yankee in the span of just seven games. The bullpen wasn't the only reason the Yanks went a Major League-worst 8-17 in that stretch, but it certainly helped.
Three days later, the Kahnle/Robertson deal was made. Let's check in on the performance of the bullpen since, and you might say things have turned around.

Over the past 30 days, the Yankees' bullpen has a 2.41 ERA, easily the best in the game, and if you're the best at run prevention, it means you're the best at a lot of things.
Yankees' bullpen ranks over the past 30 days
1st in ERA (2.41)
1st in innings pitched (100 2/3, tied with Seattle)
1st in strikeout rate (32.1 percent)
1st in average against (.181)
8th in walk rate (7.6 percent)
2nd in xWOBA on contact (.321)
That final one sounds complicated, but it's a funny acronym for an important concept. Sure, a bullpen with Chapman, Betances, Kahnle, Robertson and is going to miss a ton of bats. But it's also valuable to know what happens when contact is made. Are they allowing dangerous contact when they aren't getting whiffs? Or can hitters only manage poor contact when they're lucky enough to make contact at all?

It's the latter. For every batted ball, the combination of exit velocity and launch angle provides an expected, defense-free outcome (that is, if , for example, allows a weak bloop that turns into a hit just 20 percent of the time, we'll credit him for that contact skill, regardless of whether or makes the catch, because that's out of Warren's control). What that number is saying is that when contact has been made over the last month, it's the second least-dangerous contact of any bullpen. Combine that with the best strikeout rate, and you can see why things are going so well.
So what changed? In part, the personnel. Here's what the Yankees' Opening Day bullpen looked like:
LHP -- Chapman, , Tommy Layne
RHP -- Betances, Warren, Clippard, ,
Clippard went to Chicago in the Robertson deal, then later to Houston. Layne was cut loose in June, and was released from the Dodgers' Triple-A team earlier this month. Holder has pitched in only two games for the Yankees since June 26; he's currently on the Triple-A disabled list. Mitchell and Shreve are both active now, though each has spent plenty of time in the Minors this year.

Here's what the Yankees' bullpen looks like currently:
LHP -- Chapman, , Shreve
RHP -- Betances, Robertson, Kahnle, Green, Warren
Robertson, since returning to New York, has been fantastic, with a 1.42 ERA and a 21/2 K/BB in 14 2/3 innings. Kahnle, the best reliever no one seems to know, has a 1.73 ERA and a career-best 25 percent strikeout rate. And Betances overcame his loss of command in a big way, allowing one earned run in 14 2/3 innings, striking out 20.
Then there's Green. Acquired along with in the 2015 trade that sent to Detroit, Green's 2016 was debut wasn't that notable, posting a 4.73 ERA in 12 games (8 starts), and suffering an elbow injury cut his season short. You'll notice he wasn't on that Opening Day roster; he worked as a starter in Triple-A for the first month.

Green made a spot start in June, but otherwise he's been a reliever, and to say he's been a success story undersells it by a lot. Let's take a look at the 158 qualified relievers this year, display the top 10 by strikeout rate, and you'll notice a lot of pinstripes here.
Top strikeout rates by a reliever in 2017
50 percent --
42 percent --
41 percent --
40 percent -- Green/Betances/Kahnle
38 percent -- /
37 percent -- /Robertson
That's four Yankees in the top 10, plus two 2016 Yankees -- seeing Miller and, shockingly, Yates, offers a fascinating what-might-have-been -- and the completely unheralded Green is up there with all of them. Where did this come from?
Unlike many starter-to-reliever conversions, Green hasn't added velocity. But he did start throwing his sinker, slider and cutter less, choosing to focus more on a four-seamer that he features 62 percent of the time, up from 40 percent last year, and that's a good choice. It has very high spin (high fastball spin is positively correlated to strikeouts), the 16th highest of the 240 relievers who have thrown 100, and no regular reliever has allowed a lower batting average on his four-seamer.

This isn't a fluke, is a point. No team outside of maybe the Dodgers can come close to the depth and talent we're seeing here, and remember: By his standards, Chapman hasn't actually been that good (which may say more about Chapman's standards than his performance, but still, this isn't the Chapman the Yankees expected).
The focus of the Yankees' moves was on Frazier, on , on even . All are useful. Yet at this team heads towards October, the real strength is in the bullpen. With Chapman struggling, with Miller in Cleveland, with Betances having command issues, you couldn't possibly have expected the Yanks' 2017 bullpen to exceed last year's group. The scary thing is, that might be exactly what's happening.