Williams: What If There Were No Starting Pitchers?

When Franklin Johnson was a senior in high school, he started getting the attention of professional scouts.

Johnson grew eight inches in height between sixth and seventh grade, and by the time he entered high school, he had learned how to leverage this new height into a fastball that could strike out most hitters in the Fransernando Valley Junior Varsity Baseball League.

By the time he was a senior in high school, Johnson had figured out how to leverage a 95-97 MPH fastball out of his now 6-foot-5, 187 pound frame. He led the varsity league in strikeouts, and scouts were looking at him as a potential first round pick in that year’s MLB draft, with a major university commitment and scholarship as a backup.

Johnson was taken in the first round. His new team cited his upper 90s velocity, the plus potential from the spin on his breaking pitch, and his willingness to practice a changeup beyond his already two unhittable pitches — mostly because those two pitches also couldn’t be thrown for strikes consistently.

Most importantly, his new general manager said these key words: “We believe Franklin Johnson can become a starter in the majors one day.”

From that day forward, Franklin Johnson’s career was changed, perhaps for the worst.

*****

I feel like middle relievers are the easiest position to fill in baseball.

They exist in No Man’s Land, inside of a world filled with pre-determined roles.

If Franklin Johnson were a real player, the moment his general manager deemed him a starter, he would be on a mission: Pitch six or more innings every five days with your best stuff.

If he can’t do all six innings, can he do one inning really well? I’m sure the prospect analysts in that universe would have projected Johnson as a potential reliever at minimum, since he was already throwing upper 90s with plus breaking potential. Even though Franklin Johnson has control issues, ridiculous stuff can make hitters make ridiculous decisions.

So we dream, but we dream of the extremes.

Franklin Johnson needs to either throw six innings every five days, or one inning as often as possible and strategically dictated.

And if he can only go one inning, it’s kind of seen as a failure, unless he can consistently get the best hitters out on command.

That mentality leaves a lot of pitchers in No Man’s Land.

Real pitchers.

Who have real MLB talent.

Real pitchers like Wil Crowe, who most would have cut before the season, after being one of the worst pitchers in the majors last year. Crowe had a 5.48 ERA in 116.2 innings for a -0.3 WAR in 2021. That’s a mixed bag, because he was good enough to make 26 starts and not get completely hammered, but the Pirates added his arm in exchange for Josh Bell, so they want more than one inning from him.

So far this year, we’ve seen Crowe make six appearances with 13.1 shutout innings. His strikeouts are up, his control is better, and his confidence looks outwardly improved.

Looking deeper at his stuff, Crowe is attacking hitters with more off-speed stuff this year, which is generating more chases out of the zone. Crowe has thrown his fastball just 35% of the time this year, with the big benefactor in the fastball decrease being his changeup increasing to 33% usage, becoming his second most used pitch over his slider.

The changeup was his third-most used pitch last year, and generated the most swings and misses. If you divide his fastballs into individual pitches, his changeup becomes his most shown pitch to opposing hitters, and has maintained an 18.3% swinging strike rate.

Aside from the results, there are two ways we can see the effectiveness of the new pitch mix. The first is with Crowe’s chase rate. He’s getting hitters to swing at 35.9% of pitches out of the zone, up from 28.9% last year. The out-of-zone contact is up from 60.6% to 66.1%, but you could argue that more out of the zone contact leads to less hard contact. That’s been backed up by the numbers, with Crowe seeing a big reduction in hard contact (down to 20%), but a big increase in medium contact (70%).

What’s encouraging is that Crowe is seeing a drop in contact inside the zone. That number is down from 86.9% to 82.1%. The league average is 84.3%, which means this is the first time that Crowe has seen above-average swing and miss in the zone. His swinging strike rate is at 11.8%, up from 10.4% last year. His overall strikes are up from 25.4% to 28.9%.

We’ve also seen Dillon Peters thriving in a similar role.

The Pirates added Peters last summer in a waiver claim, and he pitched fairly well, putting up a 3.71 ERA in 26.2 innings over six starts. The advanced metrics showed that he was throwing more like a 4.50 ERA pitcher, and his career numbers indicated that could still be an improvement.

Peters isn’t much different of a pitcher this year on the surface. His strikeouts are down slightly, his walks are up slightly, and he’s yet to give up a home run in 10.1 scoreless innings over five appearances.

Just like Crowe, Peters is seeing a massive decrease in hard hits, down to just 14.3%. He’s getting much more soft contact this year. He added his slider back this year, giving him a four pitch mix.

What’s interesting here is that Peters has arguably the worst control of his career by some measure. He’s in the zone 41.2% of the time, which is a career low. He’s starting off in a 1-0 count against 61% of hitters.

Maybe it’s intentional that he works outside of the zone.

Opposing hitters are still chasing at a 35% rate (read above about the hard contact implications, which matches with Peters, who has a 14% rate). Peters is working outside of the zone more than ever, and has only seen a drop in total strikes of less than one percentage point.

In both of these cases, the potential value as starters is low, and prone to error. After a first look, opposing hitters might figure out that changeup, or learn not to chase with Peters. The off-speed heavy, high chase approach with Crowe might not work either.

But who says pitchers need to pitch multiple times through an order?

*****

The Pirates have been using their middle relievers strategically, in combination with a reduced workload on starting pitchers. Jason Mackey of the Post Gazette wrote a good summary of this recently, detailing how the Pirates were utilizing matchups and keeping outings shorter.

This is working especially well for Peters. Last year, Peters had his best result yet in his first time through the order, holding opponents to a .191 wOBA and a 3.76 xFIP. Things quickly went downhill after that. He had a .402 and 5.63 combo in his second time through the order, and .521 and 4.58 the third time. Peters was effective up until hitters saw him a second time.

This year, Peters has a .112 wOBA and a 3.58 xFIP in his first time through the order. Both of those are improved over last year’s totals, but you could say they’re similar ranges. Peters has yet to pitch a second time through the order.

Crowe is an interesting case, because he wasn’t good the first time through the order last year. Crowe has made massive improvements in those numbers this year. Alex Stumpf had a great breakdown of some of the contributing factors that might be leading to those improvements.

The thing that stuck out to me the most was Crowe’s confidence increase this year.

I think there’s something to telling a pitcher to just go get outs, with no further expectations than the hitter in front of him.

Remember our friend Franklin Johnson?

What if his general manager never set an expectation for him to fill a role that expected him to pitch six innings each night?

What if his general manager instead simply touted up the quality of pitches and said that he has the base potential to get MLB hitters out one day?

From there, the only question we’d have to ask is: How many hitters can Franklin Johnson get out in a single night?

Three minimum is required by any standard to maintain a long-term career.

Anything more than that requires pitching through a break, and maintaining the same intensity and focus, while the other team starts sharing notes and building a live report on what they are seeing.

Then, Franklin Johnson turns the lineup over and faces someone for the second time, giving his first chance at attacking someone who has recent memory on what is coming.

When we deem that prospects like the fictional Franklin Johnson are future starters, what we’re saying is that they can get outs the first time, second time, and maybe third time through the order.

Most of the time we say this without even knowing if they can eventually get one single hitter out in the majors.

Shouldn’t that be the focus, before projecting out the possibility of a second and third tour through the lineup?

That’s where the scouting aspect comes in.

Roansy Contreras, for example, has the stuff to make it highly probable that he can get hitters out, even when they have a good idea of what he’s throwing that night. More importantly in an age where it seems every pitcher has “the stuff” on paper, Contreras knows how to use his stuff effectively to get outs deep into his outings.

But does that mean we have to use the term “starter” and “reliever”?

What if every pitcher was graded on their talent and pitches, and every night they were told to go and get outs for as long as possible?

What if we used “times through the order” as the metric to divide pitchers, rather than a wide split of one inning or six-plus innings?

Every pitcher could potentially make it once through the order. Some might not have the ability to maintain their performance through a break.

Some pitchers will eventually show that they can make it twice through an order, which would equate to a starter on the good nights where that pitcher wasn’t having issues the first time he saw the opposing team.

Some pitchers will show that the opposing hitters stand no chance, no matter how many chances they get.

Those guys will never be denied their “starting” roles.

Every other pitcher might see an improvement if they’re not required to spread their stuff out over X amount of innings, but instead just told to go out and give their best stuff for as long as possible.

The question with all of this is: Can it work over a full season?

I can’t think of a better year for the Pirates to experiment than a season where they’re universally expected to lose, and only surprisingly at .500 at the current moment.

And since they’re carrying a similar approach to the minors, it indicates that this might be a long-term plan, similar to what other small market teams like the Tampa Bay Rays have been doing for a few seasons now.

For that reason, this week our Prospect Roundtable group article looks at pitchers in the current system who could benefit from the “long-relief” approach. Ryan Palencer also has a feature on a guy down in Indianapolis who might be able to help in a multi-inning, once through the order role.