Best offense ever? Astros have shot at history

In 2017, the Astros had the best offense in baseball, and that's not a particularly controversial statement. Houston scored the most runs (896), had the fewest strikeouts (1,087 -- 17 percent of its times to the plate), and had the second-most home runs (238). Like traditional stats? The Astros had the highest batting average (.282) and slugging percentage (.478). Prefer advanced metrics? They led in Weighted On-Base Average (.349) and Weighted Runs Created Plus (121, where 100 is "league average" for that year), too.
They were so good, in fact, that to call them "2017's best offense" actually undersells them considerably. Last year's Astros were either modern baseball's fourth-best offense ever (if you include pitchers batting) or the 10th-best offense ever (if you don't).
Either way, it was a historic performance, and in 2018, they're going to bring back more or less the same group, which leads to an important question. No, it's not "can they do it again." It's this: Can they be even better? Is there a path to another historic season here?
Yes -- if everything goes right, which it rarely does. But could it?
Before we get to their ceiling, let's start with the Astros' floor, which is considerable. Despite the presence of multiple superstars here, this isn't a collection that revolves around one or two players. Part of the reason they're so good is because they're so deep -- their primary No. 8 hitter last year, for example, was Alex Bregman. All he did was to hit .284/.352/.475 (122 wRC+) at 23 years old and homer off Clayton Kershaw, Kenley Jansen and Chris Sale (twice) in the playoffs.

Just over one year ago, we looked at the Steamer numbers and noted that the Astros were projected to have 2017's deepest lineup, with eight hitters expected to have at least 300 plate appearances and league-average or better production. As it turned out, they had nine. (Like everyone else on Earth, Steamer did not see Marwin Gonzalez coming.)
If we do the same exercise for 2018, we get nine projected to be average or better. No other team has more, and that makes sense. Veterans like George Springer, Brian McCann, Evan Gattis and Josh Reddick have found their above-average level and stayed there for the past few seasons. We're now four seasons into Jose Altuve being a star, and two years into him being a superstar. Correa has been a stud from the first day he set foot onto the diamond. Yuli Gurriel's first full season was more or less what his Cuban history suggested. There's not a huge risk of collapse from most of this group, barring injury. Almost no one had an unexpected breakout.
We know the Astros can be good, but how can they be great? Let's build out a best-case scenario with three reasons to think that the 2018 group could top the '17 version and challenge the best lineups in baseball history -- and, to be fair, one big reason they might not.
First, the three reasons they should improve:
Correa stays healthy
You won't see Correa on the leaderboard of baseball's top hitters, because he didn't qualify. Thanks to a thumb injury and ensuing surgery, Correa played in only 109 games and took 481 plate appearances, short of the necessary 502. If he had managed to get those 19 additional times up to the plate, he'd have tied with Freddie Freeman to be baseball's sixth-best hitter in terms of wRC+ -- and he only turned 23 in September.

You obviously can't just assume that Correa's same performance from last year will continue into next year, which is why we look to projections. By their nature, they're conservative, but both ZiPS (142 wRC+) and Steamer (140 wRC+) suggest a reasonable outcome is an elite-level season, with a chance for more. Correa might have missed something like 150 plate appearances last year; some of those went to J.D. Davis (.279 OBP), who filled in at third as Bregman pushed to short. The more Correa plate appearances, the better.
Fixing the problem spots
Speaking of shifting plate appearances to more productive hitters... Last year, 17 Astros hitters received at least 50 plate appearances. Of those 17, five produced at a below-average level:
Designated hitter Carlos Beltrán (509 PA, 76 wRC+)
Left fielder Norichika Aoki (224 PA, 88 wRC+)
Left fielder Derek Fisher (166 PA, 82 wRC+)
Center fielder Cameron Maybin (63 PA, 72 wRC+)
Center Juan Centeno (57 PA, 67 wRC+)
That's 1,019 plate appearances that came out to a total line of .234/.292/.378 (79 wRC+), which is basically what Jason Heyward (.243./315/.353, 78 wRC+) has done over 1,073 plate appearances for the Cubs in 2016-17. That can do a lot to drag down an offense.

But as things stand, four of those players will not be on the Opening Day roster, possibly all five. Beltran retired at the end of the year. Maybin is a free agent, unlikely to return. Aoki was traded to Toronto in July. Centeno was claimed off waivers by Texas in November. Fisher will probably see time with Houston, but he's also just 24 and could easily return to Triple-A to begin the season.
The Astros already had the fewest below-average hitters last year, and now this entire group might be replaced (or, in Fisher's case, retains hope of better performance). Maybe prospect Kyle Tucker steps up, or a healthy Jake Marisnick continues his 2017 improvement (.243/.319/.496, 117 wRC+), or a Lucas Duda type arrives to take some DH time. We don't know yet. But we do know that last year's limited problem areas aren't returning.
Bregman picks up where he left off
In the first half, Bregman was good, but not great, hitting .256/.338/.419 (105 wRC+). It made him the 93rd-best hitter of 166 qualified batters. In the second half, Bregman exploded, hitting hitting .315/.367/.536 (141 wRC+). That made him the 18th-best batter of 162 qualified hitters -- in the second half, he outhit José Ramírez, Anthony Rendon and Cody Bellinger.

There's reason to wonder how much of that is sustainable; Bregman didn't hit the ball harder, or in the air more, or even walk more. He mostly struck out less and pulled the ball more. That's why the projections are splitting the difference; Steamer (122 wRC+) and ZiPS (115 wRC+) are seeing basically what Bregman put up in 2017 (122 wRC+). That's probably fair, to think that he's better than his first half and not as good as his second half. But Bregman is still not even 24 yet, which makes his second-half breakout all the more appealing.
Now, other than injury, what might make us hesitant to expect greatness?
No one knows what to make of Gonzalez
Between his smashing breakout hitting year (.303/.377/.530, 144 wRC+) and his ability to make starts at five positions, Gonzalez got some down-ballot American League MVP Award support in 2017. But what is he? Over Gonzalez's previous five seasons, he hit just .257/.298/.389 (90 wRC+). Last year, only one other hitter had a larger gap between what his Statcast™ metrics indicated would be expected and what actually happened. Looking back at 2016's biggest overperformers -- like Trea Turner, Sandy León, Tyler Naquin and Aledmys Díaz -- very few maintained their performance in 2017.

It doesn't mean Gonzalez can't repeat, of course. But you understand why the projection systems are so conservative here, with Steamer seeing a 103 wRC+ and ZiPS at 106. How could you not be?
Ultimately, the argument is academic. Barring an injury disaster of epic proportions, this is going to be one of the best lineups in the baseball, probably the best, and whether it ranks in the top five or 15 or 30 of the all-time list is less important than how much damage the put on the rest of the AL in 2018. From the way things stand now, that looks to be considerable.