Pitch Model Breakout Candidates To Watch in 2025

Spring training is in full swing, which means everyone and their mother has already written an article identifying “The Top Breakout Candidates To Watch” in 2025.

I don’t mean that in a bad way. Heck, we’ve already done it here at Just Baseball. Picking breakout players is great!

Whether you’re full of optimism about players on your favorite team, trying to find the best sleeper picks to win your fantasy league, or you just want the bragging rights, it’s fun to try to guess which players will have surprisingly strong seasons in the upcoming year.

Still, I wanted to do things a little differently to make this particular “breakout picks” piece stand out. So, I decided to let the underlying numbers identify my breakout candidates for me. Specifically, I’ll be using data from the two pitch modeling systems available on FanGraphs: PitchingBot and Pitching+.

You can read more about PitchingBot and Pitching+ in the FanGraphs Library, but the gist of both systems is that they model pitch quality rather than pitch results.

In other words, they’re concerned with every aspect of a pitch except for its outcome. That means they look at variables like release point, velocity, and spin rate, but not whiff rate, exit velocity, or BABIP.

That’s precisely why pitch models can be so useful for identifying breakout candidates. PitchingBot and Pitching+ point us toward which pitchers are throwing great pitches, even if they have yet to see great results.

To identify my breakout candidates, I made a list of all 402 pitchers who threw at least 40 innings in 2024. Then, I removed anyone whose botERA (PitchingBot expected run value converted to the ERA scale) or Pitching+ was worse than the league average. That left me with 184 pitchers.

Next, I removed every pitcher from the sample whose actual ERA was better than his botERA, or whose actual ERA- was better than his Pitching+. After all, I’m trying to find breakout candidates. I’m not interested in guys who succeeded last year. That left me with a pool of 71 pitchers.

Finally, I ranked those remaining pitcher in four categories: (1) botERA, (2) Pitching+, (3) the difference between their botERA and actual ERA, and (4) the difference between their Pitching+ and ERA-. Then, I calculated each pitcher’s average ranking.

These were the top names on the list:

The seven pitchers listed above excelled last season – according to both PitchingBot and Pitching+. However, they had much worse ERAs than the pitch models would have expected.

So, we have our breakout candidates! Well… not quite yet.

Needless to say, five-time All-Star Josh Hader has already broken out. His stuff is just so nasty that his perfectly respectable 3.80 ERA looks mediocre in comparison.

Caleb Thielbar isn’t nearly as accomplished as Hader, but at 38 years old with eight MLB seasons under his belt, I’m not going to call him a breakout candidate either. The veteran lefty has a 3.38 ERA in close to 350 career games.

The other five names, however, are all unheralded and relatively inexperienced pitchers. Will that change in 2025? Let’s discuss.

Slade Cecconi

• RHP, Cleveland Guardians
• 2024 Stats: 20 G (13 GS), 77.0 IP, 6.66 ERA, 0.2 fWAR
• 2024 Pitch Modeling: 3.19 botERA, 105 Pitching+

Over 27 career outings (17 starts) at the MLB level, Slade Cecconi has gone 2-8 with an ERA over 6.00. His FIP, xFIP, xERA, and SIERA are all significantly better, but they’re still in the mid-4.00s.

However, Cecconi’s PitchingBot ERA in 2024 was 3.19. His Pitching+ was 105, 5% better than league average. According to both models, he has below-average stuff but makes up for it with phenomenal command. In particular, he locates his fastball with remarkable precision.

Over the past two years, Cecconi ranks seventh among all pitchers (min. 100 IP) in four-seam fastball command, according to PitchingBot. Meanwhile, his four-seam fastball ranks second in Location+.

It’s hardly surprising that Cecconi is a control artist, considering his career 4.7% walk rate, but his pitch model grades are a promising sign that he can use that control to become an above-average pitcher.

Cecconi is stretching out as a potential starter for the Guardians this spring, but if that experiment fails, he could make for a nice bullpen piece. His average four-seam velocity ticks up to 95 mph out of the ‘pen, and perhaps he could get it even higher in a full-time relief role. A well-commanded fastball with above-average velocity is a strong weapon for the late innings.

As a young pitcher with a ton of promise who has yet to reach his full potential, Cecconi couldn’t have wound up in a better place than Cleveland. It will be exciting to watch and see what additions and adjustments the team will have him make to get the most out of his promising arsenal.

Stephen Kolek

• RHP, San Diego Padres
• 2024 Stats: 42 G, 46.2 IP, 5.21 ERA, 0.3 fWAR
• 2024 Pitch Modeling: 2.75 botERA, 117 Pitching+

Don’t feel too bad if you don’t recognize the name Stephen Kolek. An 11th-round draft pick, the righty was never a highly ranked prospect.

Kolek began his professional career with the Dodgers, who would eventually flip him to the Mariners for cash considerations. Following three years in the Mariners organization, he was selected by the Padres in last offseason’s Rule 5 draft. He made his MLB debut weeks before his 27th birthday.

Over the first four months of his rookie season, Kolek pitched to a 5.21 ERA in 42 innings of mostly low-leverage work. He then spent the final two months on the injured list with tendinitis in his pitching arm.

Yet, there’s a reason the Padres did not return Kolek to the Mariners. And it’s the same reason why they’re giving him some run as a starter this spring, despite the fact that he has only topped 70 IP or 10 starts in a season twice in his pro career.

Kolek’s FIP, xFIP, and SIERA were all in the mid-3.00s last season, despite his mid-5.00s ERA.

Even better, his pitch modeling numbers were phenomenal. Only six NL pitchers (min. 40 IP) had a lower botERA, and all six had excellent seasons: Adrian Morejon, Tyler Rogers, Tanner Scott, Orion Kerkering, Ryan Helsley, and A.J. Puk.

Meanwhile, Kolek’s Pitching+ ranked ninth in the NL, trailing arms like Helsley, Morejon, Rogers, Puk, and Kerkering, as well as Jeremiah Estrada, Justin Martinez, and Edwin Díaz.

Simply put, that’s incredible company to be in. Eight of those nine pitchers had a relief ERA under 3.00 in 2024. The only one with an ERA higher than 3.00 was Díaz, whose career speaks for itself.

If the starting experiment doesn’t work out, Kolek has the skills to thrive in a big league bullpen.

Brandon Pfaadt

• RHP, Arizona Diamondbacks
• 2024 Stats: 32 GS, 181.2 IP, 4.71 ERA, 3.1 fWAR
• 2024 Pitch Modeling: 2.84 botERA, 106 Pitching+

It’s not so long ago that Brandon Pfaadt was one of Just Baseball’s top 25 prospects in the game. Yet, outside of a heroic performance in the 2023 postseason, he has yet to meet those lofty expectations.

In 51 career games (50 starts), the righty has a 5.06 ERA. His ERA estimators have been better, painting the picture of a mid-rotation arm. Indeed, thanks to a 3.61 FIP, he ranked among the top 20 NL pitchers by fWAR in 2024.

But if the pitch models are to be believed, we could be looking at a future ace.

PitchingBot, in particular, is sky-high on Pfaadt. His 2.84 botERA ranked first among all qualified pitchers in 2024, comfortably ahead of Zack Wheeler’s 2.97 botERA in second place.

Pfaadt also ranked among the top 10 qualified NL pitchers according to Pitching+, slotting in between Logan Webb and Shota Imanaga on the leaderboard.

His only pitch with outrageous stuff grades is his sweeper, which he uses as his primary weapon against right-handed batters and his tertiary offering against lefties. However, he had above-average command/location grades on almost all of his pitches.

Pfaadt is almost certainly the highest upside arm to watch from this list. He is currently the favorite to win the number five spot in Arizona’s rotation, although Jordan Montgomery and Ryne Nelson are in the running as well.

Nick Mears

• RHP, Milwaukee Brewers
• 2024 Stats: 54 G, 57.2 IP, 5.93 ERA, 0.8 fWAR
• 2024 Pitch Modeling: 3.65 botERA, 107 Pitching+

After parts of five seasons in MLB, Nick Mears has a 5.20 ERA (107.1 IP) to show for it. Yet, there’s a reason the Brewers – an organization with a strong reputation for developing pitchers – targeted him at last year’s trade deadline.

Similarly, there’s a reason they’re keeping him around for 2025 despite his poor performance down the stretch in 2024; he gave up 10 earned runs in 12.1 innings of work for the Crew.

Mears struck out 29.3% of batters he faced last season while walking just 9.4%. Both of those numbers were career-bests. What’s more, his strikeout-to-walk ratio was even better in Milwaukee. He struck out 18 hitters while walking just three.

If you need another reason to believe that Mears took a genuine step forward with the Brewers, look no further than the pitch models.

Mears improved his botERA from 4.02 with the Rockies to 2.07 with the Brewers. Similarly, he increased his Pitching+ from 103 to 126. These are obviously small sample sizes we’re working with, but that’s precisely why we’re looking at pitch models – they stabilize far more quickly than most other metrics.

Both models agree that his stuff and command improved exponentially in Milwaukee. In particular, his fastball took a big step forward. According to both PitchingBot and Pitching+, it was the best four-seam fastball in the NL (min. 10 IP) from the day Mears joined the Brewers to the end of the regular season.

Interestingly, he wasn’t throwing the pitch any harder, but he added two inches of rise. In addition, he was locating his fastball in the strike zone significantly more often, and he was locating it throughout the zone rather than focusing too frequently on one particular area.

Mears also started throwing more sliders and fewer changeups, and the models agree that his slider is a much stronger pitch. It will be fascinating to see what his arsenal looks like in 2025 after a full spring in camp with the Brewers.

Andrew Nardi

• LHP, Miami Marlins
• 2024 Stats: 59 G, 49.2 IP, 5.07 ERA, 0.7 fWAR
• 2024 Pitch Modeling: 2.79 botERA, 109 Pitching+

Andrew Nardi was phenomenal for the Marlins in 2023, going 8-1 with three saves, 17 holds, and a 2.67 ERA. By Win Probability Added, he was a top-10 reliever in the National League. His role in getting the Marlins to a surprise Wild Card berth cannot be overstated.

Unfortunately, Nardi sandwiched that 2023 season between a 9.82 ERA performance in 2022 and a 5.07 ERA campaign in 2024. His career 12.35 K/9 ranks third among NL relievers (min. 120 IP) over the last three years, but his 4.51 ERA puts him just outside the bottom five.

So, although you could argue that Nardi already had his breakout season in 2023, I’m still considering him a breakout candidate for this piece. Besides, it’s not as if Andrew Nardi is a household name.

And yet, the underlying numbers suggest we’re looking at a future bullpen stud. Indeed, his ERA estimators were actually better in 2024 than they were in 2023, despite his bloated ERA.

Even better, Nardi put up a 2.79 botERA and 109 Pitching+ over 49.2 innings this past year. And if you want a larger sample size, just look at his 2.88 botERA and 107 Pitching+ in 121.2 innings dating back to 2022.

Over the past three seasons, only three NL relievers (min. 120 IP) have a lower botERA than Nardi: Brusdar Graterol, Ryan Helsley, and Devin Williams. Names like Helsley and Williams speak for themselves, while Graterol has quietly pitched to a 2.10 ERA for the Dodgers in that time. In other words, we’re talking about excellent company for Nardi to be in.

Pitching+ isn’t quite as high on Nardi, but that’s only because PitchingBot likes him so much. His 107 Pitching+ from 2022-24 ranks among the top 20 NL relievers (min. 120 IP) in that time.

Nardi doesn’t throw his four-seam fastball particularly hard. Indeed, it really doesn’t seem all that exceptional by any one metric. Yet, pitch models serve as a good reminder of how many different factors go into making a successful pitch. A great pitch is often more than the sum of its parts.

Nardi only really throws one other pitch – a sweeping slider – but the models love his breaking ball, too. What’s more, the north-south relationship between the two pitches is a big part of his game, and it might explain why his overall pitch model numbers (i.e. botERA and Pitching+) are better than his individual stuff and command/location grades.

My only cause for concern when it comes to Nardi is his health. The southpaw has yet to pitch this spring as he deals with tightness in his back. It’s unclear how serious the issue will prove to be, but even if Nardi were back to full strength tomorrow, it’s hard to think he’d have enough time to ramp up for Opening Day. An early-season IL stint seems likely.