Last year, the Astros had baseball's best pitching staff, leading the Majors in ERA (3.11), strikeout rate (28.5 percent), and Wins Above Replacement (29.4). They were so good that we hadn't even made it through May before we started talking about if they could be the best pitching staff in Major League history, and they came very close to pulling it off.

They followed it up with a winter of subtraction, not addition. Charlie Morton, who had a 3.36 ERA in two years with Houston, signed with Tampa Bay. Two-time All-Star and 2015 Cy Young winner Dallas Keuchel entered free agency and is still there. Lance McCullers Jr., an All-Star in 2017 who had pitched some of the team's biggest playoff innings, was lost to Tommy John surgery.

That's 500 innings out the door, and the only meaningful addition to replace them was a mere one-year deal for journeyman Wade Miley, who had posted a 4.26 ERA in parts of eight years for five teams -- and had been bought out by the Orioles, of all teams, after 2017.

The 2019 Astros pitching may be different, but they're not worse. Even after allowing 14 runs to Seattle on Wednesday, they might actually be better.

Despite losing an undeniably talented trio of arms, this year's Astros are challenging their own records. If we start with the simplest question -- do they prevent runners from reaching base? -- then no team since the dead ball ended in 1920 has done a better job.

Lowest on-base percentage allowed, team pitchers, 1920-pres.

.274 -- 2019 Astros
.278 -- 2019 Dodgers
.282 -- 2018 Astros
.282 -- 1972 Orioles (featuring Jim Palmer)
.282 -- 1968 Giants (featuring Juan Marichal)
.282 -- 1964 and 1966 White Sox

It would be the best in at least a century, which is how far back data is available, and possibly ever. Looking at Statcast's underlying metrics, their expected OBP is all of .282. They're earning this. They're earning being the most difficult staff to simply get on base against.

That the 2019 Dodgers are second on the list tells you a little something about partial-season stats and perhaps a little more about how dominant both of these clubs are (hope you enjoyed the 2017 World Series, because we're on track for a rematch in October). But this list also tells you that it's not a fluke, because last year's Astros were already tied for the lowest-ever OBP allowed.

(On that last point: Including the Major League team, the top six Astros affiliates, from Triple-A down to Low-A, led their league in strikeout rate in 2018. It's not quite so impressive in 2019, but taking the organization as a whole, they're still No. 1.)

Put another way, the .210/.274/.387 line the Astros pitchers are currently allowing means that every hitter has been essentially the equivalent of Toronto's Billy McKinney, who's hitting .221/.272/.364. Or he was hitting .221/.272/.364, that is, because he got demoted to the Minor Leagues two weeks ago.

But wait: There's more. That .210 average the Astros have allowed is the second-lowest in the last century, behind only the .206 that the 1968 Indians surrendered. But 1968 was "the Year of the Pitcher," that historically pitcher-dominant season when the Major League batting average was just .237 (lower than this year's .249) and led to MLB lowering the mound. So when looking at it in terms of being better than the average that year, the 1968 Indians allowed an average 10 percent better than the rest of the bigs. The 2019 Astros? Well ...

Lowest batting average allowed vs. MLB average, team pitchers, 1901-pres.

16 percent better (.212 vs .255) -- 2016 Cubs
15 percent better (.210 vs .249) -- 2019 Astros
14 percent better (.206 vs .247) -- 1906 Cubs
13 percent better (.217 vs .248) -- 2018 Astros
13 percent better (.225 vs .258) -- 1975 Dodgers
13 percent better (.241 vs .275) -- 1939 Yankees

There's the 2019 Astros again at the top, just behind the historic defense of the 2016 champion Cubs, and again just ahead of last year's Astros. It's not terribly hard to see why this is all happening. The Astros lead baseball in 2019 in highest strikeout rate, have the third-lowest walk rate, and the fourth-lowest slugging allowed. They've allowed the lowest Batting Average on Balls In Play this year, which makes sense, because they have the 13th-lowest year-adjusted BABIP ever too.

If you're more of a visual person, we can put all 2,222 team seasons since 1920 on a chart showing lowest average and OBP allowed, and it should make it clear what we're seeing here. (We can talk about the 1930 Phillies another day.)

When you strike a ton of hitters out, don't walk anyone, don't allow power, and do it all in front of one of baseball's best defenses, it's not hard to see how you end up with a possibly historic pitching staff.

But, again: They lost three good pitchers. They added one in Miley who is interesting, but is probably less talented than any of Morton, McCullers or Keuchel. Top prospect Forrest Whitley, who was expected to contribute, has struggled badly in the Minors. And remember, not everything at the big league level has worked. Collin McHugh was promoted from relief to the rotation, posted a 6.37 ERA in eight starts, got sent back to relief, then got hurt. Corbin Martin made five starts with a 5.59 ERA, which earned him a demotion to Triple-A.

So how, with all of that, how are they managing to make this -- and let's focus here on "keeping runners off the bases" -- happen? We'll point to three reasons.

1) Justin Verlander is still baseball's best -- and Gerrit Cole isn't far behind

For all the talk about the changing usage of pitching staffs in today's game, this all begins with two old-school starters. Verlander, who turned 36 in February, is currently sporting a career-best 2.27 ERA, but that's not really the number that stands out here. It's that he's allowed a .205 OBP, which is another way of saying that 20% of hitters reach, and 80% do not.

Verlander starts Thursday in Seattle, and he's thrown 87 1/3 innings so far. Let's set 85 innings as the minimum, go back through the entirety of baseball history to find the 10,528 seasons where a starter has thrown at least 85 innings, sort by lowest OBP ever, and ...

.204 -- Clayton Kershaw, 2016
.205 -- Verlander, 2019
.213 -- Pedro Martinez, 2000
.224 -- Greg Maddux, 1995
.227 -- Sandy Koufax, 1965

... that could hardly be a more impressive list of names.

Now, Martinez and Maddux each threw over 200 innings, and Koufax got up to 335 2/3, so it's not that Verlander's 87 1/3 innings so far are a guarantee he'll end the season here. This is a combination of obvious skill -- he's the extremely clear front-runner for the AL Cy Young this year, which would be his second, and his ninth top-five finish, so his ticket to Cooperstown seems nearly assured -- and fortunate outcomes, because his .161 BABIP would be the lowest from a starter ever.

For Cole's part, forget about his ERA, which is almost entirely the product of two bad games. (He allowed 14 earned runs between a April 20 start against the Rangers and a May 22 start against the White Sox, and has a 2.64 ERA in his remaining 11 games.) He's leading the Majors in strikeout rate (37.4 percent), which would essentially be tied for the highest rate of all time, albeit in a high-strikeout era. His .261 OBP is 10th-best among starters who have faced 200 hitters.

Between these two: 165 innings of 3.05 ERA, allowing a mere .178/.232/.363 line. We should also mention that Miley has lowered his OBP since joining Houston (.285 this year after .309 last year, though he's surrendered more slugging) and that Brad Peacock has seamlessly moved back into the rotation, cutting his OBP from .290 last year to .272 this year..

2) Because the top of the bullpen has been unhittable.

Let's go back to our OBP list, and find all pitchers who have faced at least 100 batters in 2019. That gives us 273 results. The top five are:

.176 -- Ryan Pressly, Astros
.190 -- Roberto Osuna, Astros
.194 -- Josh Hader, Brewers
.200 -- John Gant, Cardinals
.205 -- Verlander, Astros

There's Verlander again, and closer Osuna, but the focus here is really on Pressly, who has quickly entered his name in the conversation of baseball's most dominant relievers since being traded to Houston last summer. (You may remember the record-setting 40 straight appearances without allowing a run, and all the talk about how the Astros harnessed his high spin rates into dominance.)

Pressly, in 52 games as an Astro, has allowed four earned runs, striking out 64 while issuing only five walks. He's thrown 28 1/3 innings this year, and if we go back through the entirety of baseball history, finding seasons of at least 28 innings -- of which there's nearly 28,000 -- Pressly's current .176 OBP allowed would be fifth best. Among the four ahead of him, you'll find Dennis Eckersley's 1989 and '90 seasons where he revolutionized the idea of a reliever, posted a 1.03 ERA, and had back-to-back top-six MVP seasons.

If we were to drop that minimum to 85 batters faced, we'd have another Astro in the top 25 -- Will Harris, who owns a 1.23 ERA and a .247 OBP against. Harris merely owns the fourth-lowest ERA in Astros history, among those who have thrown 200 innings.

3) Because even the back end of the pitching staff has been effective enough.

Perhaps the best way to show what's happening here is by splitting the staff into two.

If we take these seven names -- Verlander, Cole, Miley, Peacock in the rotation, and Pressly, Osuna, and Harris in the bullpen, you get otherworldly numbers: 381 2/3 innings, 2.77 ERA, and a .192/.245/.345 line. That's a .590 OPS; for context, Boston's Jackie Bradley Jr. carried a .589 OPS into Wednesday's game, and he's hitting .186/.281/.308.

If we take the other eight Houston pitchers this year -- solid ones like Hector Rondon and Framber Valdez as well as struggling ones like McHugh, Martin, and Brady Rodgers, who allowed most of Wednesday's damage -- that group has combined for 174 1/3 innings of a 5.20 ERA. That's a lot less impressive, though just Wednesday's game inflated that mark up from 4.70. Still, even that group, the secondary group, has given up a line of just .247/.331/.473, an OPS of .804. That's Brian Goodwin, or Domingo Santana, or Kole Calhoun -- decent hitters, slightly better than average, but hardly world-beaters.

If your good pitchers can dominate, that's a great start. If your secondary pitchers behind them can be basically league average, that's how you keep things afloat when your stars aren't there. Morton, McCullers and Keuchel are all talented pitchers, and surely the Astros could have found use for them this year. They haven't really missed them, though. They've barely noticed at all.