Buy, Sell, or Hold These Three Struggling Starters?

Welcome to my debut article, covering three starting pitchers struggling in 2026. Beyond the typical luck and skill factors, we want to examine a pitcher’s pitch mix, movement profile, location, and more to figure out whether these starting pitchers are likely to improve as the season progresses. Two of the struggling pitchers are veterans with a lengthy track record; the third is a less established arm who has teased us with his skills.

Will these starting pitchers perform better soon? Or should we consider dropping or selling low in some formats?

Luis Castillo, SP, Mariners

There’s no denying Luis Castillo has been brutal in 2026, with a career worst in ERA (6.57), WHIP (1.62), and K-BB% (13%). Castillo has been unlucky, posting a .364 BABIP and 59% strand rate. For context, Castillo hasn’t had a BABIP above .300 or a strand rate under 75% since 2021 with the Reds. Unsurprisingly, there have been talks about Castillo piggybacking with Bryce Miller once he is ready to return, given three other established starting pitchers and the breakout from Emerson Hancock.

The visual below shows the brutal luck factors for Castillo, specifically how his strand rate has impacted his outcomes.

Castillo probably heard the news about pairing up with Miller for an outing, given his season-high 18 whiffs on Saturday, with a 21.4% swinging-strike rate. Like another pitcher we’ll discuss later, Castillo had one of his better outings this week as the fantasy community questions his role, skill, and sustainability. He was peppering his four-seamer in the upper parts of the zone, and added one inch of induced vertical break to the pitch. Castillo’s four-seamer elicited a 27% swinging-strike rate on Saturday against the White Sox, more than double his rates over the past few seasons (11-13%). 

As good as that four-seamer was Saturday, right-handed hitters have been blasting the pitch in 2026. Castillo has allowed a .399 wOBA (.342 xwOBA) to righties on the four-seamer, compared to .215 (.248 xwOBA) in 2025 and .207 (.248 xwOBA) in 2024. The four-seamer locations against right-handed hitters have been relatively the same. However, you can see from the visual below that Castillo is throwing his four-seamer more towards the outer third of the plate to righties this year compared to last.

That said, Castillo has a 117 Location+ on the pitch this year, 12 points above his career average (105).  There is a chance that Castillo’s batted ball results can improve if he continues to locate them in the upper third. Since Castillo’s four-seam Location+ numbers look good, but his heaters are being crushed, we’ll need the stuff numbers to improve, or the command metrics may fall. 

Castillo’s slider also stood out in a recent outing, eliciting a 22.7% swinging-strike rate, nine points higher than his season-long average (13.7%), and significantly better than his career rate on the pitch (16.7%). Over the past few seasons, he has thrown a gyro-like slider, which possesses 36-38 inches of downward movement and one inch of horizontal sweep. Castillo consistently threw his slider low and away from righties or low and toward his glove side against left-handed hitters. Those consistent slider locations align with Castillo’s career-best 113 slider Location+ in 2026, but the heatmaps beg to differ. 

The heatmaps below show Castillo’s slider locations in 2026 and 2025.

It appears that Castillo has been throwing his slider slightly more over the plate in the lower parts of the zone in 2026, instead of burying them low and below in 2025. Historically, Castillo’s slider had allowed weak contact against right-handed hitters in 2025 (.224 wOBA, .246 xwOBA) and 2024 (.218 wOBA, .272 xwOBA). However, right-handed hitters have been hitting the slider harder, evidenced by a .355 wOBA (.326 xwOBA), which makes sense after seeing the heatmaps.

The visual below shows the slider’s batted ball results against right-handed hitters.

It would be logical for Castillo to locate the slider lower and below the zone, especially against right-handed hitters. Castillo’s track record says to hold, but we can’t blame fantasy managers who want to cut and stream in shallower formats. Though Castillo’s expected ERA, skills, and underlying metrics don’t support better outcomes, I’m stubbornly holding because of the track record and trust in the Mariners’ organization. 

Castillo has shown high-end command, with potentially better results from his slider locations. A pitcher like Castillo with average to above-average stuff and high-end command deserves a spot on most fantasy rosters.

Jack Flaherty, SP, Tigers

After a brutal season in 2025, with a 4.64 ERA and 1.28 WHIP, the expected ERA metrics suggested better outcomes for Jack Flaherty in 2026. Flaherty had a 3.69 xFIP, 3.97 xERA, and a 3.67 SIERA in 2025. Unfortunately, Flaherty has been torching fantasy ratios in 2026, and even his xFIP (5.19), xERA (4.77), and SIERA (4.75) are awful. Are there any metrics or skills to provide more optimism after Flaherty’s early-season struggles?

The visual below shows the concerning decline in Flaherty’s strikeout skills, which we’ll continue to discuss.

Flaherty used to have a duo of breaking pitches that induced whiffs at an above-average rate: a slider with a career 18.4% swinging-strike rate and a knuckle curve at 17%. Flaherty’s swinging-strike rate via the knuckle curve has dipped to 15.7% in 2026. Over the past few seasons, Flaherty’s knuckle curve generated an 18-20% swinging-strike rate, showing it can still elicit whiffs. Meanwhile, Flaherty’s slider has lost effectiveness year-over-year:  16.6% swinging-strike rate in 2024, 12.4% in 2025, and 11.8% in 2026. 

When we see a pitch gain or lose whiffs, we want to question the movement profile and stuff. Flaherty’s knuckle curve and slider have very similar movement this year to last – neither has shifted more than one inch in vertical or horizontal movement. And yet his Stuff+ on both pitches is down (from 92 to 85 on the slider and from 113 to 105 on the curve). Unsurprisingly, Flaherty’s knuckle curve grades well in Stuff+, as it is his best pitch for whiffs. 

Flaherty seems to command his slider and knuckle curveball well by throwing them low and below the zone. The zone rates for the slider and knuckle curve have fluctuated throughout recent seasons, but the current averages in 2026 align with the career norms. His knuckle curve has a 34.3% zone rate this year, vs. 33.1% for his career; the slider has a 42.7% zone rate, compared to 43.4% for his career. 

Opposing hitters have been chasing less often (career low 24% O-swing rate) and perhaps they are waiting for Flaherty to hang a slider or knuckle curve in the zone to hit hard. For context, Flaherty’s 27.5% chase rate via the slider and the knuckle curve’s 28.4% chase rate are 7-8 percentage points below the career norms. Previously, Flaherty’s knuckle curve and slider allowed weaker contact. The knuckle curve (.293 wOBA, .223 xwOBA) and slider (.292 wOBA, .289 xwOBA) had wOBAs under .300 in 2025 against right-handed hitters. Unfortunately, right-handed hitters have been hitting the slider (.461 wOBA, .325 xwOBA) in 2026. Results on the curve haven’t been as bad, but the jump in xwOBA  (.284 wOBA, .330 xwOBA this year) is concerning. 

The visual below shows the results against Flaherty’s pitches from left-handed hitters.

Left-handed hitters have been feasting on all his pitches, including the four-seam (.394 wOBA, .361 xwOBA), slider (.400 wOBA, .365 xwOBA), and knuckle curve (.375 wOBA, .391 xwOBA) in 2026. Flaherty’s only pitch not being destroyed by left-handed hitters was the sinker (.000 wOBA, .038 xwOB), though he throws them sparingly, just 7.5% of the time. It’s worth highlighting that left-handed hitters have given Flaherty’s four-seam issues in recent seasons, yet his breaking pitches could elicit whiffs and weak contact. 

Though Flaherty’s four-seamer grades well in the stuff models (108 Stuff+), it possesses average induced vertical break (16.1 inches), similar to 2025 (15.8 inches) and 2024 (15.4 inches). However, Flaherty still generates above-average extension (67th percentile), potentially helping to boost the stuff grade. He pumps the four-seam in the zone 59% of the time, and hitters tend to tee off, especially against left-handed hitters (.394 wOBA, .361 xwOBA). 

Unfortunately, Flaherty was lit up and inefficient on Tuesday, allowing six hits, three earned runs, three walks, and two strikeouts across 3.2 innings and 97 pitches against the Mets. For context, the Mets ranked last in wRC+ (81) overall and 29th (82) against right-handed pitchers before Tuesday. Though there’s potential in the Mets’ lineup, Flaherty couldn’t survive against a struggling offense. 

This was a make-or-break week to cut or hold Flaherty because it’s hard to stomach the awful ratios when the skills, movement profiles, stuff grades, and command don’t suggest a likely bounce back. Tuesday was probably the last straw if fantasy managers were holding onto Flaherty. Unless the slider and knuckle curve whiffs bounce back, it’s best to stay away.  

Tanner Bibee, SP, Guardians

Before Saturday’s quality outing against the Twins (6 IP, 1 H, 1 ER, 2 BB, 9 SO), Tanner Bibee’s results have been brutal. Then Saturday was the first time Bibee elicited 20 whiffs in a game since June 2024. Before May 9, Bibee had a 4.58 ERA, 1.47 WHIP, 11.6% K-BB%, 37% ball rate, and a 12% swinging-strike rate. Home runs have been an issue for Bibee, so his 4.11 xFIP and 4.12 xERA give us the most optimism. 

Bibee threw more changeups to left-handed hitters on Saturday against the Twins (34%, up from 28.6% on the season). With more changeups, Bibee threw fewer curveballs against lefties (4%, down from 12%). He has been gradually throwing more changeups to left-handed hitters over the past few seasons. That’s been helpful because Bibee’s changeup elicits an 18.5% swinging-strike rate, three percentage points higher than the league average. 

Besides Bibee’s changeup generating more whiffs, it allows weak contact, given his .154 wOBA (.166 xwOBA), particularly against left-handed hitters. Interestingly, the pitch doesn’t have an above-average movement profile, yet it leads his arsenal from a Stuff+ standpoint. For context, Bibee’s changeup has a 117 Stuff+ in 2026, similar to his career average (116). Among right-handed pitchers, Bibee’s changeup Stuff+ ranks third behind Michael Wacha and Cristopher Sánchez. Wacha might be a good changeup comparison for Bibee because it potentially tunnels well with his sinker, given their induced movement profiles. 

Like Bibee’s changeup, we saw him increase his cutter usage to 42% on Saturday vs. 34.6% in 2026 and 27.2% in 2025 against right-handed hitters. Unfortunately, right-handed hitters have been crushing the cutter, with a .435 wOBA (.406 xwOBA) in 2026. Bibee’s cutter results against righties have been worse than the outcomes from 2025 (.313 wOBA, .244 xwOBA). The cutter has an above-average horizontal movement profile, which theoretically should help with weak contact. 

Bibee tends to throw his cutter low and away from right-handed hitters, which looks like a pitch he can command relatively well. Besides the changeup, Bibee’s cutter boasts the second-highest Stuff+ grade with a decent 102 Location+ in 2026, supporting the data via his movement profile and heatmaps against right-handed hitters. Given the cutter data, especially its 19.4% swinging-strike rate, I would expect the batted ball results to improve. 

Bibee raised his arm angle this year to 59 degrees 2026 from 54 degrees in 2025 and 51 degrees in 2024. He maintained his vertical release point, but brought his horizontal release point over two inches closer to his midline. Interestingly, Bibee’s new arm angle and horizontal release point led to over one inch more glove- or arm-side movement across his entire arsenal. On a pitch-level, Bibee’s curveball added over two inches of downward movement and over four inches of glove-side sweep in 2026. 

The curveball has been helpful to both sides of the plate, specifically right-handed hitters. Bibee threw his curveball 13.7% of the time to righties, a career best, leading to a .134 wOBA (.105 xwOBA) in 2026. Weirdly, Bibee’s additional downward movement and sweep via the curveball coincided with a drop in the Stuff+ (99) in 2026, down from a 102 Stuff+ across his career. However, Bibee consistently locates the curveball low and away from right-handed hitters, giving him one pitch that moves across the vertical and horizontal planes.

One narrative to highlight involves who caught Bibee’s games. Bo Naylor caught Bibee in his first eight starts of 2026; on May 9, Austin Hedges was behind the plate. Throughout Hedges’s career, he has been a strong defender, evidenced by his 14 Defensive WAR per season compared to Naylor’s 5.6 Defensive WAR per year. Hedges also boasts an 8.9 Statcast Fielding Run Value in runs above average per season, which accounts for a catcher’s defense and framing. For context, Naylor averaged 5 Statcast Fielding Run Value per season.

Hopefully, this is a launching point for Bibee to build upon. Bibee’s two pitches with an above-average swinging-strike rate, stuff, and pitch mix adjustment support potential improvements in his results moving forward. Think of Bibee as the right-handed version of MacKenzie Gore, where it could be a bumpy WHIP, yet the ERA estimators say better results are coming soon.