Pitchers generally get less effective the deeper into a game they go, as the forces of fatigue and familiarity work against them. We've seen this change how pitching works at a fundamental level, as most starters are no longer expected to go deep into games. Both the Dodgers and Astros used this idea smartly in 2017, and teams like the Mets and Rays have indicated they'll do the same in '18.
The game has simply changed, so much so that starters who don't go deep are no longer a liability -- they're a strategy. Earlier this week, we looked into which starters might be best off if they weren't asked to turn a lineup over three times, avoiding their danger zone. Today, let's do the opposite: Which starters have been able to maintain steady production as they go deep?
It's a complicated question. This isn't so much "showing an innate talent to go deep," because no one's really better the third time through the order. It's more a look at what's happened over the past two years. While there's no such thing as a one-size-fits-all rule in baseball, no one's immune from the effect, not even the great Clayton Kershaw.
In order to find these pitchers, we first gathered the 125 pitchers over the past two seasons who faced at least 50 batters the first, second and third time through the lineup as starters. Then, we weighted 2017 performance twice as much as '16 numbers to give more credit for recent performance (this didn't actually change things that much) and calculated the average deviation for each pitcher to find the steadiest performers each time through. To be clear: We're looking for pitchers who maintained their performance the third time through, even if that performance was merely average along the way.
That sounds like a lot of math, but the outcome is simple. We want a way to show that a steady performer like Carlos Carrasco (a weighted .289 wOBA the first time through, then .298, then .311 the third time and beyond) maintains his performance better than someone like Lance McCullers (.279, then up to .322 and then .376).
wOBA is just like OBP, except it gives increasing value to extra-base hits rather than treating each time on base equally. The 2017 Major League average wOBA was .327.

The top 10 starters who have held the most steady through the lineup over the past two seasons are:
Matt Moore, Rangers
Eduardo Rodriguez, Red Sox
Jacob deGrom, Mets
Michael Fulmer, Tigers
Jose Quintana, Cubs
Joe Ross, Nationals
Jeff Samardzija, Giants
Ricky Nolasco, free agent
Kyle Hendricks, Cubs
Jameson Taillon, Pirates
That's a pretty interesting list. deGrom is a bona fide ace, and Hendricks has looked like one at times. MLB.com's Andrew Simon, using a completely separate method, just named Quintana as baseball's "most consistent pitcher." Ross, Rodriguez, Taillon and Fulmer are talented young pitchers at various stages of development, while Moore, Nolasco and Samardzija all just had below-average years. (Moore and Samardzija could at least point in part to the Giants' defense behind them.)

It should tell you that there's no one particular type. For example, Moore, who was recently traded to Texas, followed up an average 2016 (.304 wOBA, 4.08 ERA) with a down '17 (.354, 5.52). But using our weighted stats, he was remarkably steady as he went through the order -- .346, then .342, then .339. He just started from a higher place to do it.
Hendricks is a fascinating case, in part because his game is not about velocity. Only knuckleballer R.A. Dickey had a softer average fastball velocity than Hendricks's 85.9 mph, but the Cubs righty maintained it as he went through the lineup, averaging 87.9 mph the first time through, then 87.8, then 87.9. The average starter drops a full mph from the first time to the second.

This isn't intended to be predictive going forward, and to some extent, several of these guys were saved from more late damage by not being allowed to frequently pitch the third time through. For example, Nolasco faced 585 hitters the first time through the order the past two years, 578 the second time, but then only 441 beyond that. (This is why the idea tossed around in October that "Rich Hill was better the third time through" was a misreading of the numbers; he faced fewer than half as many hitters there.)
At the other end of the list, there are names like Mike Pelfrey (.374 wOBA the first time, then .342, then .475) and Michael Wacha (a stellar .268, then .352, then .379, which fuels some calls for him to be moved to relief). Obviously, there's different ways to succeed. In this case, our list of 10 names above have at least shown their managers that whatever they're doing to start, they can maintain some semblance of it to finish.