How in the world can you explain a team like the Rays? There are a lot of strange and seemingly magical things going on there, but let’s focus on just their starters. They churn out top-of-the-line dudes like no one’s business. Shane McClanahan is nasty. Tyler Glasnow looks unhittable at times. Jeffrey Springs went from zero to hero and stayed there. Zach Eflin is suddenly dominant. They can’t seem to take a step without tripping over a great starter.
They’re also always hungry for more. Whether it’s bad luck, adverse selection, or something about their performance training methods, the Rays stack up pitching injuries like few teams in baseball history. Of that group I named up above, only Eflin hasn’t missed significant time in 2023, and both McClanahan and Springs are out for the rest of the year. The Rays not only have all these starters, but they also traded for Aaron Civale at the deadline, and they’re still short on arms.
They did what anyone would do: point at a random reliever in the bullpen and tell him he’s now an excellent starter. Wait, that’s not what anyone else would do? Only the Rays do that? You’re right, at least a little bit; surely you recall the Drew Rasmussen experiment from 2021. That one was a big hit until Rasmussen tore his UCL this year. Read the rest of this entry »
One of the strange things about projecting baseball players is that even results themselves are small samples. Full seasons result in specific numbers that have minimal predictive value, such as BABIP for pitchers. The predictive value isn’t literally zero — individual seasons form much of the basis of projections, whether math-y ones like ZiPS or simply our personal opinions on how good a player is — but we have to develop tools that improve our ability to explain some of these stats. It’s not enough to know that the number of home runs allowed by a pitcher is volatile; we need to know how and why pitchers allow homers beyond a general sense of pitching poorly or being Jordan Lyles.
Data like that which StatCast provides gives us the ability to get at what’s more elemental, such as exit velocities and launch angles and the like — things that are in themselves more predictive than their end products (the number of homers). StatCast has its own implementation of this kind of exercise in its various “x” stats. ZiPS uses slightly different models with a similar purpose, which I’ve dubbed zStats. (I’m going to make you guess what the z stands for!) The differences in the models can be significant. For example, when talking about grounders, balls hit directly toward the second base bag became singles 48.7% of the time from 2012 to ’19, with 51.0% outs and 0.2% doubles. But grounders hit 16 degrees to the “left” of the bag only became hits 10.6% of the time over the same stretch, and toward the second base side, it was 9.8%. ZiPS uses data like sprint speed when calculating hitter BABIP, because how fast a player is has an effect on BABIP and extra-base hits.
ZiPS doesn’t discard actual stats; the models all improve from knowing the actual numbers in addition to the zStats. You can read more on how zStats relate to actual stats here. For those curious about the r-squared values between zStats and real stats for the offensive components, it’s 0.59 for zBABIP, 0.86 for strikeouts, 0.83 for walks, and 0.78 for homers. Those relationships are what make these stats useful for predicting the future. If you can explain 78% of the variance in home run rate between hitters with no information about how many homers they actually hit, you’ve answered a lot of the riddle. All of these numbers correlate better than the actual numbers with future numbers, though a model that uses both zStats and actual ones, as the full model of ZiPS does, is superior to either by themselves. Read the rest of this entry »
Red Sox hitters will be facing an enigmatic fastball when they host the Royals at Fenway Park tonight. When Eric Longenhagen put together our Kansas City Royals Top Prospect list last month, he wrote that Alec Marsh’s fastball “doesn’t hop even though it has the approach angle and spin axis of a fastball that typically does; it has below-average vertical break and might be surprisingly hittable against big league bats.” Calling the pitch “weird,” our lead prospect analyst further described it as a “flat-angled, high-spin sinker,” adding that the 25-year-old right-hander, whom he assigned a 45 FV and a no. 3 ranking in the system, is no longer touching triple digits as he did in his initial seasons of pro ball.
Intrigued, I showed the 2019 second-round pick out of Arizona State University what my colleague had written, and asked him a simple question: How accurate is this?
The Chicago Cubs are white hot. Shortly after the All-Star break, the Cubbies were as many as 8 1/2 games out of first place in the NL Central and six games under .500. About three weeks later, they’re a little better than an even-money bet to grab a playoff spot.
So let’s talk about Ian Happ, who has been a key offensive player for the Cubs over that span, and the odd season he’s having. If you make a habit of checking the major league walk rate leaderboards regularly, as I’m sure we all do, you will no doubt have noticed that Happ is in the top five with a 16.1% walk rate. (All stats current through Tuesday’s action.)
Now, Happ has always been capable of drawing a walk; his career walk rate heading into this season was 11.2%, which is pretty high. But you wouldn’t think of him as one of the most discerning hitters in baseball — Juan Soto, Kyle Schwarber, Max Muncy, and so on. Until this season. Read the rest of this entry »
Reports to: Assistant Director, Foundational Research Additional reporting responsibilities to: Director, Integrative Baseball Performance Location: Philadelphia, PA or Remote
Position Overview:
Analysts with the Phillies are tasked with maximizing organizational impact, in pursuit of building a consistent World Series contender. As a Performance Analyst in the Foundational Research department, you tackle baseball’s most difficult problems in biomechanics, strength and conditioning, and athletic training, leveraging proprietary data sources to systematically improve player evaluation, development, performance, and health at scale.
You work closely with technical staff and performance specialists in Baseball Operations to outline a strategic vision for the future of performance research at the Phillies. You internalize cutting-edge insights from the Phillies’ Research & Development department in player evaluation and development, leveraging those insights in performance research that is effortlessly additive to our existing player acquisition and development paradigms.
Success in this role would involve improving our existing player evaluation tools, identifying novel and high-value avenues of high performance research, building systems to improve the efficiency and value of player assessment data collection, and seeing an integrative, best-in-class baseball performance process incorporated in organizational decision-making, spanning player acquisition and player development.
Responsibilities:
Conduct and oversee performance research projects and manage the integration of their outputs into our proprietary tools and applications, in direct support of player evaluation, acquisition, development, and performance maximization
Ensure projects conform to best practices for implementing, maintaining, and improving predictive models throughout their life cycles
Communicate with front office executives, scouts, coaches, and medical staff to design and interpret statistical studies
Assist the rest of the Foundational Research and Integrative Baseball Performance teams with their projects by providing guidance and feedback on your areas of expertise
Continually enhance your and your colleagues’ knowledge of baseball and data science through reading, research, and discussion with your teammates and the rest of the front office
Provide input in architecting the storage, ingestion, display, and analysis of baseball assessment data
Rigorously identify and vet novel data sources, collection methodologies, or technologies that could be additive to our processes, supporting the organization in implementing them in a scientifically and statistically robust manner
Required Qualifications
Deep understanding of statistics, including supervised and unsupervised learning, regularization, model assessment and selection, model inference and averaging, ensemble methods, etc.
Demonstrated experience in handling, analyzing, and interpreting high-performance data in sports
Proficiency with scripting languages such as Python, statistical software (R, S-Plus, SAS, or similar), and databases (SQL)
Demonstrated experience designing, constructing, implementing, and leading technical research projects for use by non-technical stakeholders
Proven willingness to both teach others and learn new techniques
Willingness to work as part of a team on complex projects
Proven leadership and self-direction
Preferred Qualifications
BS, MS or PhD in a related quantitative (Math, Statistics, Operations Research, etc.) or scientific field (Biology, Physics, Bioengineering, etc.), or equivalent practical experience
Familiarity with best practices in machine learning operations (Git, Docker, MLFlow or the equivalent)
Experience designing and running experiments
Experience managing or overseeing the work of other data scientists or analysts
0-5+ years of relevant work experience
Interested applicants should submit both their resume and an answer to the following question:
Our R&D Department informs us that a recently drafted pitcher would have a big league arsenal if they gained 2mph in the off-season. What models would you build or metrics would you look at to determine whether the player can achieve the proposed gain in velocity? (250 word limit)
Tip: There’s no defined right or wrong answer. Responses are used to get some insight into how you approach problem solving and baseball in general.
Ben Lindbergh and Meg Rowley banter about whether the plight of running backs in the NFL is analogous to any data-driven developments in MLB, the benefits and drawbacks of more or fewer off days during postseason series (and the problem with pluralizing “series”), the White Sox going 0-11 in games when they’ve struck out 14 or more opposing batters, Cody Bellinger and the surging Cubs, the flailing Diamondbacks and how the sequencing of slumps and hot streaks affects the perception of seasons, and more. Then (1:16:05) they meet major leaguers Slade Cecconi and Davis Schneider, followed by (1:36:45) a Future Blast from 2044.
Wednesday night in Philadelphia didn’t start off as a celebration of Michael Lorenzen. Making his first home start after joining the team at the trade deadline, he struggled to get comfortable on the mound. The first batter of the game, CJ Abrams, smashed a pitch to the warning track in the deepest part of the field. The next three batters worked full counts, with one walking. Keibert Ruiz worked another walk to lead off the second inning. Lorenzen threw 53 pitches in the first three frames. Through four, he had three strikeouts and three walks.
Luckily, he didn’t need to be the focus, because a celebration in Philly was happening one way or another. Weston Wilson smashed a home run in his first major league at-bat. Nick Castellanos popped a two-run shot in the first and followed up with a solo shot in the third. The Phillies were romping over the Nationals on a glorious summer evening. And that’s leaving the best part for last: Ryan Howard was in the booth to celebrate opening a new chicken and waffles stand in the stadium.
I won’t lie to you; those waffles looked good. John Kruk was nearly rapturous as he contemplated them. At one point, he openly begged Alex Call to finish an at-bat quickly so the booth could go to commercial and he could eat. Howard seemed happy, too, and the Phillies continued to pile up runs while he recapped the genesis of his foodie vision. After four innings, the Phillies led 6-0, and the celebration was in full swing.
Obviously, though, you aren’t here to read about Howard’s chicken and waffles, or to learn, as I did, that Kruk avoids spicy food. You’re here because a funny thing happened in the back half of this game. Lorenzen, staked to an enormous lead, started attacking the strike zone. He dared the Nationals to swing – four-seamers middle-middle and belt-high sinkers, calling out to be swung at. When he fell behind in the count, he fired one down the pipe and said “hit it.”
This being the Nationals, they mostly didn’t hit it. Calling their offense punchless might be going too far, but they’re towards the bottom of the league in every offensive category, and that doesn’t account for the fact that they traded their best hitter at the deadline. Abrams is coming on, and Lane Thomas has been good all year, but we’re not quite talking about Murderers’ Row here.
Suddenly it was the seventh inning, and the Nationals were still hitless. Lorenzen pulled his secondary pitches back out; he buried Jake Alu under a pile of changeups for his fourth strikeout and then mixed four-seamers high with changeups low to coax a groundout (smashed, great play by Rodolfo Castro) out of Ildemaro Vargas. Seven innings, 100 pitches, no hits – was this happening?
That last out of the seventh inning awoke the Philadelphia crowd from its post-homer lethargy. They’d been enjoying a casual demolishing of the little brothers of the NL East. Now, they might be witnessing history. A roar broke out, and the crowd rose to its feet to collectively cheer Lorenzen as he strode off the field. Six outs, six measly outs – surely he could do it.
Lorenzen came out sharp in the eighth – by which I mean, he threw some good pitches when the count made that possible and otherwise made the Nationals beat him by putting the ball in play. It was a brilliant plan all night; the Phillies recorded 15 outs in the air, few of them threatening to be anything more than cans of corn. Most importantly for Lorenzen, that eighth inning took only 11 pitches, which gave him enough runway to come back out for the ninth.
I’ve spent a lot of this writeup talking about Lorenzen’s ability to adapt his pitching to the situation, and that was on display in his last inning of work. The strike zone widens when no-hitters near the finish line. Hitters’ pulses rise – you don’t want to be on that highlight reel, you know? Lorenzen aimed for the corners to get ahead, then snapped off ridiculous breaking balls whenever he had the chance, hoping for a miserable flail from a desperate opponent.
That plan dealt with Thomas and Joey Meneses, the latter a victim of a called strike three that was both clearly outside and clearly a pitch you have to swing at in the ninth inning of a no-hitter. That left only Dominic Smith, but he wasn’t going down easily. After falling behind 1-2, he took and fouled his way back to a 3-2 count. Lorenzen looked gassed. “One more pitch,” Kruk breathed on the broadcast, almost a mantra. And Lorenzen left it up to the gods of contact one more time. He threw a slider right down main street at 85 and dared Smith to do his worst:
After the momentous end to the seventh inning, Citizen’s Bank Park had turned raucous. That energy carried right through to the end of the game. The place positively shook when Meneses struck out, and erupted even more when Johan Rojas squeezed Smith’s fly ball for the final out. Sorry Weston, and sorry Ryan; it was Lorenzen’s night now, and the crowd bathed him in applause as he exulted in his achievement.
If you didn’t know he hadn’t allowed any hits, Lorenzen’s line wouldn’t turn any heads. Five strikeouts, four walks; it’s not exactly the stuff of aces. But Lorenzen has never been an ace, and he wouldn’t tell you otherwise. He’s never been a high octane strikeout pitcher, and now that he’s transitioned from the bullpen to the rotation, he’s leaning more than ever on his savvy. Tonight was the crowning achievement of that style.
As the stadium roared and Lorenzen’s mom beamed from the crowd, the team mobbed him. What a glorious feeling it must be to combine the pinnacle of individual achievement with your first real taste at team success. Lorenzen has played for a winning team exactly twice in his major league career – the 2020 Reds went 31-29 and the 2021 edition finished 83-79.
This year’s Phillies are a cut above that – the defending National League champions, near-locks to make the playoffs and another run at the title. And he’s one of them now, indelibly linked with this team, this city. You won’t be able to tell the story of the 2023 Phillies without mentioning this night, which means you won’t be able to tell it without mentioning Lorenzen. How wonderful that must feel after nearly a decade in the wilderness, hoping to start, then getting your wish only to toil in obscurity.
Baseball is about a lot of things. It’s about the crack of the bat and the roar of the crowd, the beauty of close plays and the shocking speed and strength of grown men wearing ridiculous pajamas. Increasingly, it’s about numbers too – teams are getting smarter and smarter about separating what seems important from what is important. But regardless of the numbers, tonight was important. Baseball isn’t just about who wins the trophy at the end of the year. It’s about nights like these, and players like these. What a glorious night for Lorenzen, and what a wonderful celebration of baseball.
No team has come out of the August 1 trade deadline hotter than the Texas Rangers. Not only did they make substantialadditions to theirroster — most notably by adding Max Scherzer and Jordan Montgomery to their rotation — but they’ve reeled off a season-high eight-game winning streak to start this month, enabling them to expand their AL West lead over the Astros to three games. Alas, they suffered a substantial blow along the way, as rookie third baseman Josh Jung fractured his left thumb in Sunday’s game. He’ll undergo surgery to stabilize the injury, which could sideline him for the next six weeks, but fortunately the Rangers have the depth to withstand his absence.
The 25-year-old Jung injured himself while knocking down a 109-mph line drive off the bat of the Marlins’ Jorge Soler in the sixth inning of Sunday’s game. The ball hit the base of the thumb of his glove hand and squirted out, but he had the presence of mind to recover it and step on third base for a force out, then throw to second to complete a double play.
One batter later, Jung left the game, to the surprise of manager Bruce Bochy. “I went out to the mound, I had no idea he got hurt on that play,” said Bochy. “The calmness that he showed picking up the ball, stepping on third … he didn’t grimace, he didn’t do anything to make us think that he was hurt.” Read the rest of this entry »
One of the strange things about projecting baseball players is that even results themselves are small samples. Full seasons result in specific numbers that have minimal predictive value, such as BABIP for pitchers. The predictive value isn’t literally zero — individual seasons form much of the basis of projections, whether math-y ones like ZiPS or simply our personal opinions on how good a player is — but we have to develop tools that improve our ability to explain some of these stats. It’s not enough to know that the number of home runs allowed by a pitcher is volatile; we need to know how and why pitchers allow homers beyond a general sense of pitching poorly or being Jordan Lyles.
Data like that which StatCast provides gives us the ability to get at what’s more elemental, such as exit velocities and launch angles and the like — things that are in themselves more predictive than their end products (the number of homers). StatCast has its own implementation of this kind of exercise in its various “x” stats. ZiPS uses slightly different models with a similar purpose, which I’ve dubbed zStats. (I’m going to make you guess what the z stands for!) The differences in the models can be significant. For example, when talking about grounders, balls hit directly toward the second base bag became singles 48.7% of the time from 2012 to ’19, with 51.0% outs and 0.2% doubles. But grounders hit 16 degrees to the “left” of the bag only became hits 10.6% of the time over the same stretch, and toward the second base side, it was 9.8%. ZiPS uses data like sprint speed when calculating hitter BABIP, because how fast a player is has an effect on BABIP and extra-base hits.
ZiPS doesn’t discard actual stats; the models all improve from knowing the actual numbers in addition to the zStats. You can read more on how zStats relate to actual stats here. For those curious about the r-squared values between zStats and real stats for the offensive components, it’s 0.59 for zBABIP, 0.86 for strikeouts, 0.83 for walks, and 0.78 for homers. Those relationships are what make these stats useful for predicting the future. If you can explain 78% of the variance in home run rate between hitters with no information about how many homers they actually hit, you’ve answered a lot of the riddle. All of these numbers correlate better than the actual numbers with future numbers, though a model that uses both zStats and actual ones, as the full model of ZiPS does, is superior to either by themselves. Read the rest of this entry »
Dan Szymborski
Sponsor Us on Patreon
Facebook Group
Twitter Account
EW Subreddit
Effectively Wild Wiki
iTunes Feed
Get Our Merch!
