Analytics pay off in big way for Astros pitchers

The Astros are coming off a regular season with what can legitimately be called one of the best pitching staffs in decades, a lofty title that is in no way hyperbolic. Houston's 3.11 ERA wasn't just the best in baseball, it was (on a park-adjusted basis) 24 percent better than the rest of the American League. Dating back to World War II, that's tied with the 2016 Cubs, 1997 Braves, and the '54 Giants for the third-best pitching performance against that year's league average.
But when it came time to set their American League Division Series roster to face Cleveland, something interesting became clear. Of the 11 pitchers they carried, only three were homegrown: Dallas Keuchel, Lance McCullers Jr. and Josh James. (Keuchel was drafted long before the current front office was in place, and James didn't even appear in the ALDS.) That collection of mostly acquired pitchers held Cleveland bats to a mere .144/.196/.222 slashline, one of the most dominating performances in postseason history.
That construction stands in stark contrast to the hitters they carried on the ALDS roster, of whom nearly two-thirds -- including stars Alex Bregman, Jose Altuve, George Springer and Carlos Correa -- were drafted and developed by the team. Houston's strategy here seems clear: Draft hitters, find pitchers.

:: ALCS schedule and results ::
But it's not as simple as going out to find good pitchers, really. Anyone can do that. It's about the famously analytical Astros identifying talented players on other clubs who may not be performing to the best of their abilities and identifying ways to make them better. The defending champs use high-speed cameras and data analysis in ways that make other clubs seem like they're miles behind, and their pitching staff is full of stories of talented arms who came to Houston and improved.
We can quantify that, somewhat. Seventeen pitchers threw at least 20 innings for the Astros this year. Of the ones that remain, eight appeared first for other clubs and made their first Houston appearances in 2015 or later (Justin Verlander, Gerrit Cole, Charlie Morton, Roberto Osuna, Ryan Pressly, Héctor Rondón, Will Harris and Joe Smith).
• Gear up for the ALCS
As a group, in the five seasons since 2014, here's how they performed with Houston compared to the seven other teams they were variously on.
With Houston
2.78 ERA, .206/.270/.342, .612 OPS, 30 percent strikeout rate
With other teams
3.57 ERA, .238/.300/.378, .678 OPS, 23 percent strikeout rate
That's an enormous difference, and it doesn't even tell the entire story. (We didn't include Collin McHugh, Brad Peacock and Tony Sipp, who appeared for Houston before 2015. They had a combined 4.44 career ERA for other teams and a 3.68 mark in Houston.)
The ability to make pitchers better might be the defining trait of this Astros era. Here's how they've done it.
Justin Verlander
It's not like Verlander wasn't successful before he came to Houston: he won the Rookie of the Year, Cy Young, and Most Valuable Player Awards with Detroit, and when he ends up in Cooperstown someday, it will be with a Tigers logo on his hat. But he was more good than great late in his tenure in Motown, especially compared to what he's done since arriving in Texas.
With DET: 2016-17: 3.38 ERA, .216/.282/.376, 26 percent strikeout rate
With HOU: 2017-18: 2.32 ERA, .192/.236/.349, 35 percent strikeout rate
It's a stunning difference for a pitcher who's now 35 years old, and it's thanks to a variety of changes, including fastball location, strike-throwing, and all but abandoning his changeup. But perhaps most importantly, it's about increasing the height of his release point, which had been steadily decreasing for several years.

That matters, because it's allowed him to gain separation between his fastball and slider, making life tougher on hitters. In 2016, for example, he had a vertical movement difference of about 11 inches between the two pitches. In 2018, it was 20 inches. We're oversimplifying, of course. There's a lot more to it than this. But it's this, too. He's throwing like vintage MVP Verlander again.
Gerrit Cole
After Cole's breakout 2015 (2.65 ERA, fourth in the National League Cy Young Award voting), his next two seasons were more than a little disappointing (4.12 ERA in 54 starts), which was part of the reason Pittsburgh fans were somewhat unexcited about the four players they received in return for Cole in January.
It wasn't hard to see what the Astros were going to do. They were going to have Cole throw more breaking pitches and fewer sinkers. It's what they do. It's what they did, and we could all see it coming.

Cole cut his sinker usage in half, from 14 percent to 6 percent. He dumped his changeup, from 10 percent to 4 percent. He threw slightly more sliders, a lot more curves, and the four-seam fastballs he threw were located very, very differently.
Percentage of four-seamers thrown in the upper third of the zone or higher:
2017 -- 36
2018 -- 52
That went from below average to above average, as the 2018 Major League average was 46 percent. "Fastballs high and curveballs low" is the Houston way, and Cole's strikeout rate jumped from 23 percent to 34 percent -- the largest increase of any pitcher with 100 innings thrown in both years.
Ryan Pressly
Pressly is perhaps the perfect Houston pitching story, because he was a somewhat-undervalued Minnesota reliever who wasn't collecting wins or saves, but he was showing clear underlying talents thanks to his extremely high spin rate -- as we noted before he was traded in July, and again when he came to Houston and dominated.
"We've targeted him, we watched him progress," Astros GM Jeff Luhnow said. "He really made some big strides the past couple of years, and I don't know why he was not more of a household name because what he was doing in terms of his arsenal and getting guys to swing and miss. It was pretty impressive."

In 23 1/3 Houston innings after the trade, Pressly struck out 32 and walked just three. He was having a good year with the Twins, but the Astros took his best-in-baseball curveball spin and his Top 10 fastball spin, and made some very Astros changes. This is going to sound familiar: More curves, fewer fastballs, more high fastballs.
Fastball rate: 49 percent with Twins, 35 percent with Astros
Curveball rate: 25 percent with Twins, 37 percent with Astros
High fastball rate: 36 percent with Twins, 43 percent with Astros
There's also an argument that Pressly's slider, which he's throwing lower than last year, may be the best slider in the game. Each of his three main pitches is dominant in some way. He has three career saves. It doesn't matter. Don't say you weren't warned when he's getting big outs in huge situations.
Will Harris
Harris had had his moments with Arizona and Colorado before Houston claimed him on waivers after the 2014 season, but especially so over the past two seasons, when he has posted a fantastic 116-to-21 strikeout-to-walk ratio.  
It's not hard to see what changed. His curveball, his main secondary pitch, has gained nearly 4 mph of velocity over the years.

How does that happen? Here's an MLB.com article from 2015 on the subject.
Harris credited pitching coach Brent Strom with helping him change the grip of his breaking ball.
"I felt like I had a much harder, sharper breaking ball than I had last year, and we kind of worked on it one day and figured out something in my grip," Harris said. "That's definitely been the main thing. My curveball has been a little harder, a little sharper than it has been in the past."
He's also nearly doubled the horizontal movement on the pitch since his 2012. Sometimes, it's about technology and data. Sometimes, it's just a pitching coach offering a new grip.
Charlie Morton
After seeing Cole and Morton leave Pittsburgh to succeed in Houston, Pirates fans are surely wary of any future trades between the teams. A big part of Morton's story is merely about adding velocity, and that began in a brief stop in Philadelphia in 2016 for the simplest possible reason -- "I tried to throw the ball harder."
Even so, his two seasons in Houston have been stellar, with a 3.36 ERA and a 28 percent strikeout rate, compared to a 4.39 ERA and a 16 percent mark with the Pirates and Phillies. Morton has a high spin curveball too, but his transition from a groundball artist into a strikeout pitcher has coincided with fewer sinkers and more four-seamers.

This is the Astros way, really. They finished in the bottom 10 in sinker rate, mirroring a trend in baseball away from sinkers and toward breaking pitches and four-seam fastballs.
It's not that Houston is better than anyone else at finding good pitchers, necessarily. It's that the Astros might be better than anyone else at getting the most out of the pitchers they find. It's a huge part of the reason why they won it all in 2017, and why they may be the best-positioned team to win it again in 2018.