Starting Pitcher Potential Improvers Based on OOPSY Projections

Ret of season projections are a great way to identify players likely to perform better moving forward. The table below shows all the starting pitchers whose rest of season OOPSY-projected ERA is 3.80 or better (as compared to a league average of 4.19) and at lest 0.20 better than what they have shown so far this year (through June 22).

We’ll likely follow this up with another version looking at potential fallers based on the OOPSY projections. Today, we’ll highlight four starting pitchers that should perform better based on their skills and stuff for the rest of the season.

George Kirby, SP, Mariners

George Kirby projects to have significantly better ratios moving forward, but things have not been good so far. Kirby’s 1.31 WHIP is unusually high for him. Kirby has boasted a K-BB% 20% or higher throughout his career, but this year it has fallen to a career-low 15.5%. His 10.5% swinging-strike rate is around one percentage point lower than the two most recent seasons – not a massive decline, but not great.

And a big chunk of that decline seems to be tied to his  knuckle curve. The swinging-strike rate on that pitch has dropped more than 10 percentage points, from 19% in 2025 to 8.6% this year. Kirby’s curveball has also been allowing more hits and hard contact against left-handed hitters. Lefties have a .418 wOBA off the pitch, compared to .221 last year. The stuff models seem not to like the knuckle curveball as much, with a 105 Stuff+ after 113 in 2025. Furthermore, Kirby’s 103 Location+ via the knuckle curveball has waned compared to a 109 Location+ in his career. However, Kirby’s knuckle curve has a similar movement profile with an additional inch of drop (52.7 inches) and an identical amount of glove-side sweep (9.6 inches). And those rough results against lefties aren’t really backed up by xwOBA (.296 this year vs. .254 last year).

Kirby’s knuckle curve heatmaps show slightly more consistency in 2025 against lefties, though he is throwing them more in the zone this year. The curveball zone rate was 34.8% in 2025, up to 43.9% in 2026. Specifically, opposing hitters have been making contact on the knuckle curve when thrown in the shadow parts of the zone this season (88.5% contact rate this year vs. 55.7% last). This explains why Kirby’s curveball is generating fewer whiffs and allowing more balls in play this season. 

Besides Kirby’s knuckle curve losing whiffs, he seems to have ditched the splitter for a changeup. That’s notable because Kirby’s splitter generated a 17.9% swinging-strike last season, with the changeup at 10.6% this season. After Kirby threw a splitter with an above-average downward movement profile, dropping 35-37 inches, his new changeup looks like a kick-change. Last year, David Laurila of FanGraphs wrote about supinators learning to throw a kick-change. Sometimes we find pitchers with a supinator bias develop a kick-change, since they typically lean on their cutters, sweepers, and curveballs.

That can be somewhat evident when looking at a pitcher’s changeup movement profile. In Kirby’s case, the changeup drops 30 inches, but averages 14.6 inches of arm-side fade. Kirby’s changeup is similar to Hunter Brown, Gerrit Cole, and Max Meyer’s changeup in terms of velocity and movement profile. Maybe it’s a moot point since Kirby only throws his changeup 10.7% of the time, primarily to left-handed hitters, but the results have been decent. 

Kirby struggled with consistency with the splitter to left-handed hitters in 2025 (.468 wOBA, .448 xwOBA) and 2024 (.318 wOBA, .315 xwOBA). This season, Kirby’s changeup allows a .328 wOBA (.420 xwOBA) to lefties. As mentioned earlier, the changeup might not matter because of its 83 Stuff+ and 95 Location+ this season. Though Kirby’s changeup has a 104 Location+ in his career, he hardly throws them. 

Kirby continues to have near-elite command of the fastballs, with a 115 Location+ from the four-seam and a 110 Location+ via the sinker. His sweeping slider also boasts a high-end 115 Location+. Like the OOPSY Rest of Season projections, I believe in Kirby’s skills, stuff, and command to support better outcomes. I would be buying low on Kirby if possible. 

Jesús Luzardo, SP, Phillies

We go through this often with Jesús Luzardo, where we’re soaking in the upside and hurting from some blowup outings. Luzardo has five outings with five or more earned runs and seven quality starts with two or fewer earned runs out of 15 games started. He has one of the largest projected differences in ERA for the rest of the system, according to the OOPSY projections. It’s partly because Luzardo seems to be allowing a higher BABIP this season, nearly 30 points higher than his career norm. Over the past two seasons, the Phillies allowed the sixth- (2025) and second-highest (2026) BABIP in baseball, so his teammates don’t appear to be helping him out.

However, Luzardo’s strikeout skills (19.2% K-BB% and 13.4% swinging-strike rate) have been consistent with the career averages. Both are above average. Interestingly, Luzardo’s groundball rate is a career best (47.6%), yet he continues to allow a 12.9% HR/FB rate. Though Luzardo’s HR/FB rate is close to his career averages, a groundball rate six points above his career norm has helped him keep down his HR/9. Those additional grounders are thanks to his  changeup generating more groundballs (5-6 points), while his sinker has been inducing a ridiculously high 73.3% groundball rate. 

Luzardo has been commanding his sinker well, with 113 Location+. He throws his sinker 10.5% of the time to right-handed hitters, but they hit them hard, with a .425 wOBA (.388 xwOBA). Meanwhile, the weaker contact and groundballs tend to come against lefties. Luzardo throws sinkers 35.6% of the time, up 10 points from 2025 versus lefties. The sinker allows a .323 wOBA (.257 xwOBA) and a -17 degree launch angle against left-handed hitters. 

Luzardo’s sinker added over 1.5 inches of drop and nearly one inch of arm-side fade against lefties. Specifically, against left-handed hitters, Luzardo has been locating his sinker over 4.3 inches lower in the zone and over one inch more toward his arm side (toward the inside of the plate to left-handed hitters). Some of Luzardo’s pitches tend to be hit hard, so the sinker could help significantly by inducing weak contact, assuming the groundball rate sustains.

Luzardo’s changeup has been dropping 35.1 inches, nearly five points more than in 2025. The changeup’s downward movement is the highest amount in Luzardo’s career while maintaining his 15-16 inches of arm-side fade. More downward movement on the changeup aligns with Luzardo’s 104 changeup Stuff+, a career high. That’s notable because Luzardo already commands the changeup well (105 Location+), which he peppers low and away from right-handed hitters. 

Like Kirby, I would be buying low on Luzardo if possible. 

Kyle Bradish, SP, Orioles

We’ve been on the Kyle Bradish roller coaster ride. Earlier in the season, some questioned whether Bradish was droppable. Then Bradish struck out 10 against the Athletics in early May, with four quality starts in the next five games. After that, Bradish allowed 10 earned runs across eight innings in two starts before setting a season-high with 12 punchouts against the Mariners. One of the issues has been Bradish’s control, with a 38.9% ball rate and 11.7% walk rate. Throughout Bradish’s career, he has posted a 37% ball rate and an 8.6% walk rate. 

Maybe Bradish’s walk rate regresses closer to his career norm if the ball rate hasn’t changed significantly. So far, it appears that Bradish’s 14.6% swinging-strike rate from last season may be the outlier, compared to 10.8% this season (similar to his career average). Bradish’s slider (21.7%) and curveball (23.4%) swinging-strike rates were 5-7 points higher last season than his career norms. The slider is eliciting a 15.8% swinging-strike rate with the curveball at 19.2% this season, so there’s hope for more whiffs if he can maintain those numbers on the breaking pitches.

Bradish’s command has taken a significant hit with a 90 Location+ after 105 throughout his career. The most notable command changes came from his sinker (102 Location+) and slider (90 Location+), both of which were 12-14 points below his career norm. Bradish has been throwing his sinker nearly three inches (2.9) lower in the zone against right-handed hitters, though they’re still thrown toward the heart of the plate. 

Similarly, Bradish is throwing sinkers 2.6 inches lower in the zone after locating sinkers high and away (toward his arm side) against left-handed hitters in 2024 and 2025. Though the high sinkers don’t generate many groundballs (16-17 degree launch angle in 2024 and 2025), the lower sinkers have been generating more groundballs (7 degree launch angle) and allowing more balls in play (.302 BABIP) this season against left-handed hitters. We could speculate on Bradish’s sinker command being related to his lowering his arm angle to 53 degrees, down from 56 (2025) and 58 (2024).

Let’s close with examining Bradish’s slider command. The Location+ numbers indicate Bradish is struggling to command his slider. Bradish’s slider heatmaps and zone rates have been relatively similar against right-handed hitters. However, the average slider pitch height dropped by 3.5 inches this season against right-handed hitters. That makes his gyro-like slider more low and away instead of along the outside corner against righties. 

However, Bradish’s slider is allowing more hard contact (.405 wOBA, .356 xwOBA) and balls in play (which hurts given a .370 BABIP) against righties. For context, Bradish’s slider allowed a .212 wOBA (.249 xwOBA) and .249 BABIP against right-handed hitters before 2026. Opposing hitters have been making contact 94.5% of the time (up nearly 10 points) when sliders have been thrown in the heart of the zone. Meanwhile, the 74.7% contact rate on the slider in the shadow parts of the zone jumped by nearly 20 points compared to 2025. It seems like opposing right-handed hitters have been attacking the slider if he leaves them in the heart of the zone (.838 wOBA) and shadow parts (.373 wOBA).

The OOPSY rest of the season projections seem the most optimistic, besides STEAMER. That likely indicates they’re weighing the recent skills, stuff, and command numbers more heavily than the current season. Bradish’s small samples in 2024 and 2025 might be the outlier, but there’s still an above-average starting pitcher within his range of outcomes. Hate to be boring, but I’m looking to acquire Bradish because I’m still optimistic.

Framber Valdez, SP, Tigers

After spending eight seasons with the Astros, Framber Valdez joined the Tigers. Some fantasy managers choose to be skeptical of players joining new teams, and while there isn’t much research to back that up, so far it looks like they would have been right to be out on Valdez. Like Bradish, Valdez’s 9.5% K-BB% is far below his career 14.4% K-BB%. Valdez’s hasn’t changed much, but the whiffs took a significant hit. He has been eliciting an 8.8% swinging-strike rate after 11.2% or higher in the previous four seasons (2022-2025). Valdez typically paired above-average whiffs with plenty of groundball; both are down this year.

Valdez’s groundball rate has plummeted to 50.8% this season, 10 points below his career average. Unsurprisingly, nearly all of Valdez’s pitches have been generating a lower groundball rate this season. The most notable pitches with a lower groundball rate are his sinker (50.9%) and changeup (52.4%), both dipping by 9-10 percentage points. Specifically, Valdez’s sinker has a 100 Location+, three points below his career norm. He is throwing his sinker 2.4 inches higher in the zone against right-handed hitters. Valdez has also been cutting his sinker more toward his glove side, meaning it’s losing 1.4 inches of arm-side run against righties. 

Unfortunately, Valdez’s curveball hasn’t been generating many whiffs (12.7% swinging-strike rate), down over five points from his career norm. Meanwhile, Valdez’s curveball groundball rate dipped by over 7 points this season. The curveball’s pitch shape looks similar to recent seasons, though Valdez’s 117 curveball Stuff+ dropped by five points compared to his career average. It’s still one of his best pitches from a stuff standpoint, so that’s probably not significant if the movement profiles have been nearly identical. Though Valdez’s slider has a 123 Stuff+, he has thrown them under 5% of the time over the past three seasons (2024-2026).

This could be a lesson in patience with Valdez, given his extensive track record. However, it’s a slimmer margin for error if Valdez doesn’t regain the whiffs and groundballs because he turns into more of a mediocre than an above-average starting pitcher. I’m not as optimistic about Valdez turning it around, but fantasy managers probably need to ride it out.