PITTSBURGH -- A year ago, the Pirates viewed their starting pitching as the backbone of their team. Steven Brault called their collection of arms “scary,” in a good way, and MLB.com suggested they might have MLB’s most underrated rotation.

Instead of being a strength, the Pirates’ rotation was a glaring weakness during their 93-loss season. Beset by injuries and ineffectiveness, Pittsburgh’s starters combined to record a 5.40 ERA last year, the fifth-highest mark in the Majors.

Yet the Pirates haven’t acquired a starter this offseason, which means most of the main characters from that cast -- minus Jameson Taillon, who on Monday began his throwing progression after Tommy John surgery -- are set to reprise their roles in 2020. They’ve been a little busier behind the scenes, though.

New general manager Ben Cherington and manager Derek Shelton have emphasized a collaborative culture in pursuit of “player improvement.” They hired a data-driven pitching coach, Oscar Marin, who believes in maximizing pitchers’ individual abilities. They retained catcher Jacob Stallings and signed the similarly defensive-minded Luke Maile, betting that improved game-calling and framing will help their pitchers.

All of their activity, or lack thereof, seems to be indicative of a club hoping its existing options can improve organically.

“I think we see potential gains, but the proof is in the pudding,” Cherington said last month. “You have to go do it.”

Let’s take a quick look at how the Pirates’ top six starters can make those potential gains this year.

Chris Archer

Archer began this process last summer by shelving his two-seam fastball and throwing his best pitches -- his four-seamer and wipeout slider -- more often. He reintroduced his windup less than a year after ditching it. He kept throwing his changeup and curveball, which add variety to his arsenal.

Archer still posted a 4.68 ERA in his final 10 starts, so it’s not like he totally turned the corner. But he was harder to hit. He struck out 69 batters in 50 innings over that stretch, and opponents slashed just .244/.318/.409 against him -- much more like his 2017 season with the Rays (.246/.303/.407) than his tenure in Pittsburgh to that point.

Better health is paramount, as the typically durable Archer has been banged up throughout his time with the Pirates. So far, so good on that front.

Joe Musgrove

Last year, Musgrove proved he can handle both a starter’s workload (31 starts, 170 1/3 innings) and the six-pitch mix he once considered trimming.

Musgrove throws a four-seamer, sinker, cutter, slider, curveball and changeup. Half of those pitches -- the slider, curveball and changeup -- produced a swinging-strike rate above 30 percent and a wOBA below .300 last season. Opponents fared better against his fastballs, but they were key to getting the strike-throwing Musgrove ahead in the count before he could unleash his offspeed stuff.

Evaluators still see untapped upside in Musgrove because of his athleticism and promising peripheral numbers, including a FIP that's been half a run lower than his ERA over the last three years. Maybe another year of experience and a modified pitch mix, with fewer fastballs and more offspeed stuff, will unlock that potential.

Trevor Williams

This isn’t necessarily about Williams reproducing his 2018 performance, though nobody would complain about a 3.11 ERA in 31 starts. It’s about staying healthy and repeating his delivery, two key aspects of his best season to date.

Williams didn’t overpower batters while recording a 1.38 ERA in the second half of ’18. He just executed pitches and attacked hitters’ weaknesses with pinpoint command, which came naturally with his delivery in sync. He was off to a good start last year, too, posting a 3.33 ERA while holding opponents to a .678 OPS in his first nine starts.

Then came an injury to Williams’ right side, which sidelined him for a month, and he simply wasn’t the same when he returned. Most of his issues stemmed from an inconsistent delivery, which may have been the result of a surprisingly short stay on the injured list. In his next 15 starts, Williams logged a 7.12 ERA while giving up 19 homers -- four more than he allowed in all of ’18.

As long as he stays healthy and repeats his delivery, it’s probably fair to expect more of the 2017-18 Williams than the ’19 version.

Mitch Keller

The top-line numbers weren’t pretty: a 1-5 record, a 7.13 ERA and only 48 innings pitched in 11 starts. But if you don’t sweat the surface-level stats, there were encouraging elements within Keller’s debut.

The former top prospect struck out 65 and walked only 16 of the 227 batters he faced. While teams shattered home run records, Keller only gave up six in 48 innings. That led to a 3.19 FIP, which is typically more predictive of future success than ERA.

Opponents also batted a mind-blowing .478 when putting the ball in play against Keller. For reference, the league-average batting average on balls in play (BABIP) is typically around .300. According to Baseball-Reference data, only 15 pitchers have worked at least 45 innings while being victimized by a BABIP of at least .400 in one season -- and Keller’s .478 mark was easily the highest.

Statcast data backs up the notion that Keller was unlucky, particularly when throwing his fastball. Opponents hit .461 with a .719 slugging percentage against his fastball, far better than his expected marks of .324 and .476.

Keller throws hard with a lot of spin. He offers two quality breaking balls, which he didn’t use enough last year. He commanded his pitches well in the Minors. With more experience, better game planning and some luck that’s not historically bad, he should be fine.

Steven Brault

If there’s a trend for most Pirates starters, it’s too many fastballs and not enough offspeed. But some pitchers, like Brault, can thrive with a fastball-first approach.

From May 20-Sept. 1, Brault put together a 2.87 ERA and held opponents to a .666 OPS. His success was driven by well-located fastballs -- a lot of them. His Sept. 1 outing at Coors Field was overshadowed by the 441-foot homer he hit, but it was the purest example of that approach. Seventy-seven of the 82 pitches he threw over 6 1/3 innings were fastballs. Brault isn’t going to out-stuff hitters, but when he commands his fastball like that, he can be effective.

Chad Kuhl

This one isn’t too complicated: He just needs to pitch. Kuhl hasn’t scaled the mound since June 26, 2018, but he’s back to full health after recovering from Tommy John surgery. At the very least, Kuhl is another Major League-ready option for the rotation -- something Pittsburgh lacked last season. But if he stays healthy and continues to refine his stuff, he still could be more than that.

Kuhl has the raw stuff. In 2018, Kuhl’s average fastball clocked in at 95.3 mph. His curveball spin rate (2,901 rpm) ranked in the Majors’ 96th percentile. His slider produced a .229 xwOBA. It’s still not clear if Kuhl will return as a starter or reliever, but his high-octane stuff makes him an interesting addition either way.