When Statheads Age

This is Joe Sheehan’s first piece as part of his April residency at FanGraphs. A founding member of Baseball Prospectus, Joe currently publishes an eponymous Baseball Newsletter. You can find him on Twitter, as well.

Like a lot of fans, I watched Sunday’s and Monday’s games fascinated by the number of pitchers who seemed to be throwing harder than they did last season. So it was a relief to see Dave Cameron’s note here Tuesday about why readings were higher. It will take some mental gymnastics to compare velo figures from 2017 to previous years, but I’m sure it will be second nature aft…

… wait, what?

This is what it’s like being a baseball fan in 2017. The issues you face are ones of reconciling changes in how the velocity of every single pitch thrown in MLB is tracked. It’s not about getting that data, but rather, about sussing out the difference between measurement points of the pitch on the way from the pitcher’s hand to home plate.

We had a different set of problems when we were putting together the first Baseball Prospectus annual, back in the winter of 1995-96. The challenges we faced weren’t discerning which measure of velocity to use, but rather, when we would get access to minor-league statistics, and how soon lefty/righty splits would be in our hands, and would anyone at all talk to us about prospects we only knew by their stat lines.

I fell into Prospectus by chance, before it even had a name. Clay Davenport had been publishing his Translations — a system that improved upon minor-league equivalencies — online for a few years, and he and Gary Huckabay had an idea for a book, an annual that would include the Translations and Gary’s Vladimir projections, and commentary on players. and team essays. I knew both from the Usenet newsgroup rec.sport.baseball (r.s.b), which was the primordial soup of internet baseball writers. I can still remember the conversation, in November of 1995, in which Gary invited me to work with them. I was a journalism major and my first real job involved layout, so I was able to edit and assemble the early annuals. We wanted to write about the importance of OBP to offense, and the necessity of limiting workloads to the health of young pitchers, and the silliness of using fielding percentage to determine who was good at playing defense. We wanted to write the book we wanted to read.

It’s been almost 22 years since baseball went from a game I love to a job I love. It’s been almost 30 since a kid in my local Strat-O-Matic league (ITBL!) showed me the 1988 edition of the Bill James Baseball Abstract. Strat and Bill James were how I learned there was more to baseball stats than what showed up on the baseball cards or the television. R.s.b built upon that knowledge, teaching me about replacement level, and the importance of age in evaluating prospects, and why throwing 130 pitches in a game is bad for 21-year-olds.

This all sounds quaint in 2017, but it was on these arguments and others like them — some we later learned were wrong — that Prospectus was built. We worked with the data we had, and we were incredibly happy to have it. The Stats, Inc. books were something of a bible back then, especially the minor-league versions. There was no Baseball-Reference [pause to genuflect], no Play Index, no sortable stats on FanGraphs. I can remember using Total Baseball and the Baseball Encyclopedia in those early days. The idea that we might some day know the velocity of every pitch thrown was visible from that moment; that we would know the spin, and break, and from where it was released, would have seemed far fetched.

For someone who remembers assembling those early books, who can see them from where he’s writing this, where we are now is astounding. Sunday’s Opening Day coverage on ESPN featured references to defensive WAR, “barrels,” and win probability. At Great American Ball Park in Cincinnati — the city where major-league baseball began — the scoreboard listed WAR, OPS+, and ISO, with text explanations throughout the game. We spent the better part of a decade at BP on a war footing, trying to get fans, media, and industry professionals to look at these ideas. There’s no war any longer.

In fact, much of the innovation now is being done by the league itself. With first PITCHF/x, and now Statcast, it is MLB, through its MLB Advanced Media arm, that has accelerated the process by which we’re learning about the game. Teams, all 30 of them, use data to make decisions as banal as where to play the second baseman in the fourth inning, and as complex as how to manage the health of their pitchers. As the industry has embraced these tools and, more importantly, the mindset that data drives decisions, the media that covers the game has come along.

The place I helped build, the site you’re on now… this is the mainstream media. That’s not a pejorative. I write for Sports Illustrated. Christina Kahrl is an editor at ESPN.com. Rob Neyer was on FS1’s baseball coverage. Jay Jaffe goes on MLB Network, as do Dave Cameron and Jeff Sullivan. We’re not that long removed from MLB refusing to let me into the Rule 5 draft in Dallas. Today, dozens of internet-only writers are credentialed. There’s an entirely new category of baseball writing that blends analysis with reportage, best exemplified by the work of Eno Sarris, but available too in places like the Boston Globe (Alex Speier) and the New York Times (James Wagner). Coverage of baseball has never been stronger, never been smarter, never been seen by more people.

Change has come so rapidly that I find myself at loose ends a bit. I’m a stathead, but I was never a sabermetrician. My greatest work in that vein is probably a research piece on Jack Morris’s career. Words are my skill, not numbers, and with each passing year, the numbers become a little more complex, a bit harder to manage, a bit more daunting. The Play Index remains my primary tool, but the real work is being done far afield from seasonal stat lines and game logs and OPS+. The real work is being done a terabyte at a time with radar that collects every movement of every solid object — bats, balls, people — on a baseball field.

I can actually understand the confusion and frustration of the ballwriters of the 1990s, who had been able to talk about batting average and RBIs and wins for years, and who now found themselves being called dinosaurs for doing their job. I look at BABIP or HR/FB, numbers that were advanced as recently as five or six years ago, and I feel like I’m not providing enough of the story. If our revolution was from context-sensitive stats to context-neutral ones, the new one is from outcome stats to skill ones. If Matt Adams hits a ball that, based on launch angle and exit velocity is a home run 92% of the time, but Albert Almora makes a great play on it, has he succeeded or failed? Jharel Cotton allowed five runs in 4.1 innings Wednesday night; as Eno Sarris tweeted, however, the four run-scoring hits off Cotton averaged 75 mph off the bat. So was Cotton bad?

This is where the conversation is now, and like those writers a generation ago, I’m faced with the choice between sticking to what I’ve done, or learning. It’s not easy; I had to corner a friend recently to get a primer on exactly what I’m looking for when it comes to spin rates on different pitches. I’m trying to figure out whether “barrels” tells us anything we didn’t know about who the best hitters are. On most days, I’m just trying to figure out how to get access to this information. The further we get from box scores, the more specialized skills, the more proprietary tools, are needed just to make sense of those terabytes.

No matter the year or the lessons, we’re all still learning about baseball. In 1996, it was translating performance based on competition and run environment; in 2000, it was DIPS theory, and what pitchers could and could not control; in 2009, it was measuring the differences among catchers in how balls and strikes were called (“pitch framing” to most, “systemic violation of Rule 2.00” to me). Now, it’s spin and launch angles and route efficiency. In 2026, I assure you, it will be something else. The lesson that runs through all these moments is this: if you cling to what you once knew, you’ll be left behind. Keep learning.