Using xwOBA to Identify Breakout Pitchers

January 15, 2020

Statcast metrics like xwOBA can help fantasy baseball owners identify breakout candidates at pitcher. Eric Samulski reviews some underperformers based on expected stats from 2019 in order to help with draft prep.

Heading into a new baseball season, fantasy owners are looking for any leg up on our competition. We turn over stones looking for any stat or training report that might suggest a breakout that few others can see coming. I’m not here to promise some groundbreaking stat, but I do believe that Expected Weighted On-base Average (xwOBA) can be useful when looking for hints as to how a player might perform.

xwOBA is calculated using exit velocity, launch angle and, Sprint Speed to evaluate the quality of contact that a batter makes or a pitcher gives up. While, as Craig Edwards effectively points out, it is not a predictive stat, it “can help explain how a pitcher has arrived at his runs-allowed total.” More specifically, when comparing xwOBA with wOBA, we can start to see if a pitcher earned the batted ball results he gave up or if the results were due to factors outside of his control. Essentially, did a pitcher deserve better numbers than he wound up with.

The following article will look at some of the pitchers who underperformed their xwOBA. By using the simple equation of [wOBA – xwOBA], we can find pitchers who had an xwOBA that was lower than his wOBA, suggesting that he should have statistically performed better based on the quality of contact made against him.

What Do You Expect?

As a point of comparison, some of the best starting pitchers in terms of their xwOBA last year were Gerrit Cole (.238 xwOBA), Justin Verlander (.249) Jacob deGrom (.253), and Max Scherzer (.254). For relievers, some leaders last year were Emilio Pagan (.221 xwOBA), Kirby Yates (.224), and Liam Hendriks (.229).

Below is a table of pitchers who faced at least 100 batters and are intriguing based on how they underperformed their expected results. Some of them are high-end arms who have the potential for a better 2020 season, some are players who have strong statistical numbers that could benefit from a new role, and others are players who could we are simply rolling the dice on in hopes that they become more fantasy relevant.

Blake Snell, Tampa Bay Rays

Locked-in SP1

I don’t think there are many people who are shying away from Snell in fantasy leagues, but after his Cy Young Award-winning season, injuries derailed his 2019 follow-up. While his overall numbers may not have been in line with last year’s leading fantasy aces, Snell’s expected stats suggest that he underperformed his overall ability. In fact, his .264 xwOBA is just behind Scherzer and was the 6^th best number of all pitchers with at least 350 batters faced (Snell only faced 441 due to injury). His xBA was third-best at .205, as was his xSLG at .327, a whopping .064 points lower than the actual slugging percentage that he gave up last year.

We mentioned that expected stats aren’t predictive, but they are reliable year-to-year, so Snell’s consistency between 2019 and 2018 (.203 xBA, .340xSLG, and .273 xwOBA) suggests that he pitched just as well, if not better, last year and we should see something similar in 2020. With the Rays still putting a strong offense behind him, Snell should remain a safe SP1 in fantasy drafts, and I would draft him over guys like Shane Bieber and Stephen Strasburg, who are currently going ahead of him.

Edwin Diaz, New York Mets

Top-Five closer

Edwin Diaz was always likely to be a volatile fantasy asset; however, nobody could have expected last year's meltdown. A 5.59 ERA with a 4.51 FIP and a 1.38 WHIP caused many fantasy owners, and Mets fans, to freak out. Despite his BB% rising and his SwStr% decreasing, his final numbers were still well within his career range and the range of most strong relievers. His 17.8 SwStr% was a drop of 1.1 from 2018, but would have put him 4th in MLB if he had enough innings to qualify, and his 8.7% walk rate was a 2.6% increase, but still well below his 2017 numbers.

For all intents and purposes, it seems like Diaz just had a bad and unlucky season. His home run per nine innings rate was 2.33, which was one of the worst rates among relief pitchers and a ridiculous jump from his 0.61 rate in 2018. He also added to that a BABIP of .377, which was the second-worst in the league. All of that, plus the .067 difference in xwOBA and wOBA tells me that even in this nadir, he still was pitching better than his results. I fully expect a bounce-back to a top-five closer who I'd gladly take over Liam Hendricks, Ken Giles, and Will Smith - all who I've seen go above him in recent mock drafts.

Noah Syndergaard, New York Mets

Solid SP2 with high-end SP2 upside

Thor seems to pull people in every year with those long, flowing locks and triple-digit fastball. However, he rarely matched expectations when it’s all said and done. That’s something you can take advantage of. As early ADP indicates, the fantasy public is souring on Thor as the failed expectations become part of the larger narrative. Yet, his underlying skills remain strong. He’s .280 xwOBA was 16^th among pitchers who faced a minimum of 350 batters, and he also ranked in the top 20 in the difference between actual slugging percentage and expected slugging percentage, with an XSLG of .366 what would have also put him 16^th last season. Part of the explanation for Syndergaard’s poor performance in relation to his expected stats is that the New York Mets ranked second-worst in DRS and seventh-worst in UZR, which means he got no help from his defense.

There were also many reports that he lost a feel for the slider, possibly due to the change in the ball. Thor's slider dropped from a 7.8 Pitch Value (pVAL) to a 1.7 pVAL pitch, which helped his swinging strike percentage (SwStrk%) also drop 1.1 points to 12.5%. However, the ball is reportedly being switched back and could lead to improved results in Thor’s slider, which would then be paired with a change-up that saw a pVAL jump to up 6.8 from 1.4 in 2018. Thor is by no means a sure thing, but his upside is still high enough that I would take him over Yu Darvish, Zack Greinke, and Tyler Glasnow, who are all going before him.

Matthew Boyd, Detroit Tigers

Low-End SP3

Boyd’s .297 xwOBA would have put him 33^rd in the league, which gives him some room for improvement; however, I’m not as bullish on Boyd as I am on some of the other guys on this list. Boyd’s success and failure last year was heavily tied to his fastball velocity, which jumped from 91.1 MPH on average in 2018 to 92.4 in 2019. That, however, is still below average fastball velocity, and the pitch had a 1.9 pVAL; although that’s an improvement from the -0.5 the year before, it doesn’t give me confidence that he has newfound, consistent success with it.

There is some intrigue in the fact that he saw a jump in SwStrk% from 10.2 to 14.0, which can partially be tied to his improvement in getting hitters to chase outside of the zone. He had a 4.6% jump in O-Swing%, and batters swung at Boyd’s 1550 pitches and missed on 484 of them (31.2%) which is above league average (24.9%). Although his slider was markedly worse in pVAL than it was in 2018, it still has above-average results and movement. Boyd’s slider moves, on average, five inches towards a right-handed batter and drops 46 inches, while the league average horizontal movement is six inches and 39 inches drop. Improvement in 2020 from Boyd is supported by a 3.88 xFIP and those aforementioned jumps from last year; however, I would be cautious of expecting consistent production throughout the whole season, which makes Boyd more of a high floor, low-end SP3 for me.

Josh James, Houston Astros

SP3 (if he gets a rotation spot)

Nothing beats a post-hype sleeper. In the middle of last Spring Training, everybody was all over Josh James. He seemed like the next stud to grace the mound for the Houston Astros. Then he suffered a quad injury in spring and began the year in the bullpen, featuring dynamic raw stuff and the potential upside of a dominant arm, only for him to get hurt again in July and be placed on the IL with a shoulder strain. His numbers dipped a bit in July while pitching through the injury, but his overall underlying metrics tell us that his year was likely better than many think. James' .264 xwOBA would put him right in line with Snell, Mike Clevinger, and Stephen Strasburg. He also limited contact to a .171 xBA and a .286 xSLG, despite registering a .374 SLG on the season. His 16.2% SwStr% was near elite and his xFIP of 3.77 paints a much rosier picture than his 4.70 actual ERA.

When you match the numbers under the hood with a fastball that he can run up over 100 MPH, and three positive pVAL pitches, including a slider that has elite spin and above-average horizontal movement and drop, all the pieces are there for a Josh James breakout. With Jose Urquidy and Brad Peacock currently slotted into the number four and five spots in the Astros' rotation, James is not an unrealistic option to slide ahead of them. Urquidy has no Major League success and Peacock's best years came in the bullpen, so if James is able to snag a spot from one of them, we could see the true breakout we all wanted last year.

Mitch Keller, Pittsburgh Pirates

SP4 with SP3 upside

As it stands right now, Keller appears set to begin the 2020 season in the Pirates rotation. Despite mediocre results in his MLB debut, Keller proved to be a dynamic prospect as he worked his way up through the Pittsburgh’s minor league system. He possesses a strong fastball with elite spin, which tops out at 98, and has elite spin on his curve, which leads to two inches more drop than the league average. He also throws a slider that recorded a 27% SwStrk% and a 50.5 O-Swing%. Since we know he possesses good raw stuff, the fact that no pitcher in baseball underperformed their xwOBA more than Keller immediately jumps out.

When you look closer, you can see a few major culprits. For starters, he allowed a .475 BABIP, despite finishing with a .324 mark in AAA and .366 in AAA the year before and also had a sub-60% LOB rate. Both of these indicate that Keller was particularly unlucky or hurt by poor defense, a scenario that’s supported by his 3.19 FIP and 3.47 xFIP last year. I’d expect an ERA in the high three's with good strikeout numbers, which makes him an intriguing arm to breakout without having to draft him all that high.

Lucas Sims, Cincinnati Reds

SP4 (if he gets a rotation spot)

Damn the Reds and their solid addition of Wade Miley. The veteran lefty gives the Reds four locked-in starters: Miley, Trevor Bauer, Luis Castillo, and Sonny Gray, which means Sims could either compete for the 5^th spot in the rotation with oft-injured Anthony DeSclafani or be moved back to the bullpen. For the purposes of this endeavor, let’s assume he gets a crack at the starting job in the preseason; there is a decent amount to like. In addition to the noticeable difference in expected stats in the table above, Sims’ improvement last year was borne out by a 4.12 FIP in AAA and the highest K% of his career at 30% during Triple-A and 32.2% in the Majors. He also saw a 12% drop in hard contact rate from 2018 and a drop in his walk rate by 6%.

He has three solid offerings with a fastball that moves 10 inches towards a right-handed batter despite a league average horizontal movement of 7 inches, a slider that has a horizontal movement of 12 inches despite a league average of 6, and a curve that moves 2 inches more horizontally than major league average. The curve also had a pVAL of 5.5 last year, which would have been good for 15^th best had he qualified with the right amount of innings thrown. His 14.9% SwStr% would have put him 10^th in the league behind Lucas Giolito, which suggests that strikeout upside is real. If he wins the job, he could be dinged a bit by pitching in a hitter’s park, but I see a path to a sub-4.00 ERA with good strikeout numbers despite a high walk rate.

Justus Sheffield, Seattle Mariners

End-of-Draft Stash

Once an (over) hyped Yankees prospect, Sheffield posted a 5.50 ERA in 36 innings with Seattle last year but a more concerning 6.87 ERA in Triple-A. However, he dominated AA with a 2.19 ERA, which gave some people cause for optimism. After all, Sheffield is still only 23-years-old, and the large discrepancy in his expected OBA versus his wOBA gives us a reason to dive in again. While a .321 xwOBA seems high on the surface, it puts him in the same boat as Pablo Lopez, Caleb Smith, Robbie Ray, and Chase Anderson. His xBA last year was .245 which came in well below his actual BAA of .303, and the difference in his SLG allowed and xSLG was also a sizable .092. Part of that has to do with him only giving up 16.5% hard contact, but it also feels a little fluky.

Sheffield has strikeout upside with a strong 22% K% during his major league stint last year. He has a 93 MPH fastball that he can get up to 95, which is solid from the left side, and pairs with that a slider that drops two inches more than league average that he throws 35.7% of the time. However, his ceiling is currently limited by only being a two-pitch pitcher. He has some untapped upside to emerge as an SP4 in fantasy leagues, but that’s not the type of upside that you’ll be cursing yourself for missing out on if somebody swoops before the last rounds of the draft.

Elieser Hernandez, Miami Marlins

Waiver Wire Watch

The beauty of fantasy baseball is that something we find value in the unlikeliest of places. Coming into the year, many people thought the Miami Marlins would be a glorified minor league team, but we all took turns falling in love with Caleb Smith, Pablo Lopez, and Zac Gallen (until he was traded). Is it time to add Elieser Hernandez's name to that list? Not likely. However, his expected stats can put him on our list of pitchers to watch in the early weeks of the season.

He had an elite exit velocity against, which helped him to an xBA of only .209, sizably lower than his .242 final numbers. He also pitched to a .391 xSLG, which was a full .112 points below the actual slugging percentage he allowed. It was his age 24 season and only his second crack at the major, but he dropped walk % by 2.1 points and raised his K % by a whopping 8.2 points. His slider had a 7 pVAL and has almost double the league average horizontal movement. At the end of the day, he's not going to win you fantasy leagues, but he's a young pitcher with an above-average SwkStrk who induces soft fly balls in a pitcher's park. It's not going to be pretty, but there is a scenario where it is useful.

Darwinzon Hernandez, Boston Red Sox

Who Knows?

I have no idea where Hernandez pitches for the Red Sox this season, but his underlying metrics suggest that he needs to be mentioned on this list. He had the fifth-largest discrepancy between xwOBA and wOBA of any pitcher with over 100 plate appearances against, and his .159 xBA trailed only Josh Hader with the same qualifications. What's most impressive is that the average suppressing was done with both of his most frequently used pitches. His fastball registered a .154 xBA, .269 xwOBA, and.205 xSLG), which his slider finished with a .177 xBA,.240 xwOBA, and .277 xSLG. He had a solid 13.7 SwStrk% and saw his K% finish at 38.8% in his first big league stint.

He currently seems slotted to begin the year out of the bullpen, but the Red Sox are also trying to move David Price to clear salary cap, which could free up a rotation spot for the 23-year-old. If he stays in the bullpen, there is a chance that he could work himself into high-leverage situations and perhaps become useful in SV/HLD leagues. Spring Training will be big for him, but he has the tools to be a strong fantasy contributor.

More Sabermetrics & Statcast for Fantasy Baseball