The three teams that threw the fewest fastballs last year were the Indians, Yankees, and Dodgers, who all made the postseason. The World Series champion Astros threw the sixth-fewest fastballs. The D-backs, one of the final four teams in the National League, threw the 10th-fewest. Meanwhile, the sport as a whole has been throwing fewer and fewer fastballs every year for the last decade. It's not about "establishing your fastball" anymore, not in the age of launch angle and record-setting home run totals.

We all saw Lance McCullers throw 24 straight curves at the Yankees in the American League Championship Series, the same year he had the highest curve percentage we've ever seen in the pitch-tracking era (back to 2008). We've seen Rich Hill resurrect his career by throwing his curve nearly half the time. We saw Corey Kluber dominate throwing his fastball less than half the time, and we saw Clayton Kershaw finally fall below 50 percent fastballs after years of pushing in that direction.

So, who's next? Who might be a good candidate to improve by taking a good secondary pitch and simply throwing it more? It's one thing to find an ace like Kluber and make him slightly better; it's another to find a pitcher who has struggled to find consistency or success and make him valuable.

Let's do exactly that. Let's find pitchers who have good secondary pitches and below-average fastballs, and suggest that they change their pitch mix. Take the good pitch and throw it more. Take the less effective pitch and throw it less. It's as simple as that. As Hill said of his curve in 2016, "It is my best pitch and if I can utilize that pitch to the best of my ability, I'm increasing my percentages of success."

It's not as simple as that, of course. There's a great big world of pitch sequencing and theory about how a decent fastball can set a hitter up for a deadly breaking pitch, but we're not going to get into that here.

We found our pitchers by using four rules.

1) A minimum of 1,500 pitches thrown in 2017. That's 150 pitchers.

2) We're not trying to fix Chris Sale, so we'll exclude pitchers who were already good last year, which we'll call "at least 5 percent above average" based on the Statcast™ quality-of-contact expected outcomes.. We're down to 119 pitchers.

3) We don't need pitchers who already rarely throw fastballs, so we'll exclude 25 pitchers who had a fastball percentage of 50 percent or under. We don't need pitchers with effective fastballs, so we'll exclude 36 more who had better-than-average results on their fastballs, again via the Statcast™ quality-of-contact metric. Toss out a handful more who are unsigned, injured, retired or call Coors Field home, and we've just cut over 70 more names.

4) Finally, this only works if you have a good secondary pitch in the first place. So, we're looking at only those pitchers with offspeed pitches that were thrown at least 10 percent of the time and allowed an expected outcome of .300 or below, which is very good.

This still leaves us with about three dozen pitchers, but with apologies to Patrick Corbin (slider), Blake Snell (curve), and Padres pitchers Dinelson Lamet and Luis Perdomo (slider), we're not going list every single one. Instead, we're going to focus on just five.

Kyle Gibson, slider

Gibson has had back-to-back seasons of a 5.07 ERA, and at 30 years old with a career 4.70 ERA, it would seem to be nearing "now or never" time for the Twins righty. That's in part because his 92 mph fastball has never been that impressive; it actually had the second-worst results of any regular four-seamer in baseball last year, and his sinker was only mildly better.

But over his final eight starts, he had a 2.92 ERA, with a 46/10 strikeout-to-walk ratio, in part because he began throwing the slider more (6 percent in May, and 22 percent in September, for example), moving his position on the rubber, and in part because it's a good pitch. Last year, among all pitchers who got 200 swings on their slider, his swing-and-miss rate was Top 15. The year before, it was Top 8, similar to Noah Syndergaard.

We know the fastball hasn't worked. The slider might -- and given his lack of successful track record, he's the perfect candidate to lean into it.

Jordan Zimmermann, slider

You can say basically the same thing for Zimmermann as we did for Gibson, with the exception being that Zimmermann, who made two All-Star teams for Washington, actually was once a very good pitcher before falling on hard times in Detroit.

That's partially due to health, and velocity. Zimmermann's fastball was once 94-95 mph, and last year it was 92.7 mph. It was never really an elite pitch to begin with, and over the last two years, it's been crushed -- allowing a .339 average in 2016, and a .342 in 2017.

But the slider still showed promise, even in a down year, as hitters had just a .227 average against it, and no slider in baseball generated more swings. When he does it right, as he did earlier this month against Aaron Judge, he can throw it in the zone and still get misses on it.

A good slider, thrown with more frequency, could even help improve the fastball, if hitters have to be more aware of the breaking pitch. Zimmermann likely isn't getting back to his former stature, but at this point, Detroit would simply take a league-average starter. This is how.

Gerrit Cole, curveball and slider

Chad Kuhl, slider

We're lumping these two together because they're similar in many ways, as they're both talented hard-throwing starters who were Pittsburgh teammates last year. (Among starters who threw 500 fastballs, Cole had the fifth-hardest velocity, and Kuhl the 13th-hardest.)

They each also had essentially league-average seasons, as Cole had a 4.26 ERA and Kuhl had a 4.35 mark. Why? In part, because for all their velocity, their fastballs haven't had a great deal of success. Kuhl's sinker induced one of the lowest ground-ball rates, while Cole's four-seamer is a straight pitch with league-average spin and unimpressive outcomes. His sinker, meanwhile, allowed a .515 slugging percentage and eight homers last year.

They both have potentially plus secondary offerings, like Kuhl's slider, which has a 38 percent strikeout rate in his short career and was on display earlier this week.

Here's where they differ: Cole was traded to Houston, the home of McCullers and Charlie Morton. The Astros threw the sixth-fewest fastballs; the Pirates threw the most. It's almost a guarantee that we'll see a pitch mix change from Cole -- and likely for the better.

Tyler Chatwood, curveball

Chatwood's fastballs were crushed last year. Batters hit .323/.471/.561 off his four-seamer and .296/.380/.543 off his sinker, and that's just a lot of damage, no matter how you slice it.

Some of that should improve by simply leaving Coors Field, but there's more to it than that. As we investigated before he signed with the Cubs, there are a few reasons to really like Chatwood, and one of those reasons is his high-spin curveball, which he threw 270 times and allowed just four hits off of.

But Chatwood only threw that curve 10 percent of the time last year, and here's where we have to point back to his home park, as many pitchers are hesitant to throw their curve at altitude. His fastball has only been OK, but his curveball might be great. Now that he's back at sea level, it's the perfect time to throw it more and more -- as we're already seeing him do.

"That's definitely the plan," Chatwood said to The Athletic last month. "Use a lot more curveballs and changeups this spring."

Of course it's the plan. The Cubs certainly can see the same thing we do. More curves should help, because it's a good pitch -- and it might help the fastball, which averaged 95 mph last year, play up too.