Scoring a baseball game in the Savant era is a paradox. After the final out, what remains on the page—eraser smears, hastily added totals, a coffee stain—is a limited account of what has happened. Masataka Yoshida went 0 for 4, I said to myself one night in September, closing my book on a Blue Jays win over the Red Sox. But in the second, he lined out to second base, and in the seventh, George Springer made a web gem on a ball to right field. Regardless, on my scoresheet, there were only these annotations: L4, F9.
In truth, I’d tried to give Yoshida partial credit, and had scored his double-turned-flyout with an exclamation point: F9! It was something.
Fanalytics
It’s these moments in baseball, these stat-line injustices that have defined the sport since people played with no gloves and a dead ball, that are slowly disappearing in the era of Savant and Statcast, two tools that MLB makes more available to fans than it does broadcasts of the games themselves. 20 years ago, say, here’s what we would have known about Yoshida’s night: 0-4, 2K, and removed for a pinch-hitter as the game went into extra innings. Fans at the game might have known more, so too the TV audience, and people scoring the game might have remembered later (F9!) that he’d been unlucky. But there was no emotional recourse, no concrete way for fans to look at F9 or even F9! and know that Yoshida had deserved better.
But now, when almost nobody keeps score and almost everybody knows about Statcast, fans can jump straight from the broadcast to Savant and perform what we might term “fanalytics,” the use of stats for emotional fulfillment. Yoshida fans like me were soothed by the knowledge that that even his second-inning lineout, relatively unremarkable as it looked, had an Expected Batting Average of .650. And what about the xBA of the robbery? Surely it was higher?
Surprisingly, upsettingly: .110. This didn’t match my perception of reality; the eye test told me that Springer wouldn’t make that play more than 50% of the time, and that other outfielders had even less of a shot.
So does my scorebook have it backward? Should it have been L4! and F9? Are there better peripherals for this situation? I am, admittedly, far from an analytics expert. But I am an inveterate scorer of baseball games, and I’d say there’s still a place for bringing out the old pen and paper.
The Fanalytics Mindset
One of the strangest parts of attending baseball games in the 2020s is that, amid the huge video boards playing highlights and the enormous sound systems blasting music, amid the hot-dog races and the other on-field entertainments, there are still huge swaths of downtime. Between pitches and between innings, yes, but also while the umps are watching a replay, or while there’s a mound visit, or while everyone waits for the grounds crew to repair the pitcher’s landing spot. What people did 25 years ago I have no idea, but now we all gravitate to the little oracle in our pockets that we call the smartphone. When I’m at a baseball game, and a player hits a ball especially hard, I’ll check the exit velocity on my phone during the next break. If I’m rooting for the batter, I can complain that he was robbed if someone caught it. This strategy is, it turns out, psychologically impervious: if next year, while I’m complaining that Triston Casas got robbed, Juan Soto has a homer taken away at Fenway, I can say that the Red Sox employed excellent defensive positioning and athleticism. If my player hits a bloop shot over second base at 65 miles an hour, that’s smart and scrappy. If their player does? Cheap shot, unlucky.
This is the fanalytics mindset, rife on Twitter and everyone else online: not objective, or scientific, or useful to anyone except a fan. In some ways, though, it’s become essential to my experience of a game. This summer, after celebrating a homer and seeing the new Fenway light show, it was fun to know that Adam Duvall hit that ball 415 feet, and that it would have been gone in 30/30 parks.
Unassailable. He deserved that one.
Front Offices
The recently-fired Red Sox executive Chaim Bloom, a part of whose job for the Sox was to revamp the farm system in an analytical image, might look at all this a different way. It doesn’t matter to him, as it does to me, that Duvall’s homer would have left the 29 other parks. Duvall was signed as a free agent for the Sox precisely because he pulls the ball with a lot of power, and because hitting very high fly balls to left at Fenway results in home runs, even if they wouldn’t leave any other park.
Bloom might also have looked at more proprietary analytics and noticed that Duvall sells out average for power. This can be a drawback or an advantage, depending on who bats in front or behind him. Placing Duvall in front of a high-average, low-power hitter like Luis Arraez would be a dream, since Duvall strikes out much more often than he grounds out. (In 2023, Duvall didn’t ground into a double play once.) So you could count on Arraez to get on, and Duvall to hit homers an appreciable percentage of the time. Plus, you could be reasonably sure he wouldn’t ground into a double play.
Bloom’s analysis, then, is almost certainly more dispassionate than what I do on Baseball Savant. Smarter and deeper, too.
Moneyball
And yet there the numbers are, in great detail and with pleasing formatting, on a website provided to us for free by MLB. No regional blackouts, no paywalls, nothing. What does that tell us?
First, it tells us that fans like the statistical complexity of baseball, and that they enjoy feeling like insiders. One need only watch the Friday Night Baseball broadcasts on Apple TV+ to realize this: on what is otherwise a minimalistic broadcast screen, stats rotate in and out of the bottom right corner: live probabilities for home runs, hits, strikeouts. I’ve never seen a local broadcaster do this. But Apple TV is aware that they’re showing only a couple games, only on one night a week, and making somewhat an event of it. Putting “Home Run Probability: 10%” under Shohei Ohtani will make even the most distracted viewer sit up a little straighter. Stats are a time machine in both directions; we can, or at least we think we can, look incisively into the past and future at once.
In Moneyball (2011), Brad Pitt, playing Athletics executive Billy Beane, asks Jonah Hill’s character a well-known question: “How can you not be romantic about baseball?”
Fair enough. The two are watching a poignant recreation of a real sequence: minor-leaguer Jeremy Brown, large and not particularly athletic, and not realizing he had hit a home run, trips around first and falls over before being told that the ball has left the park.
But that sentence—How can you not be romantic about baseball?—has a bigger implication than anyone in the movie knows. Billy Beane spent his tenure with the A’s learning exactly how not to romanticize baseball. How to forget the emotion of someone’s story or the thrill of their athleticism and to focus squarely on the numbers. And not the usual ones. He gets on base, they say constantly. He gets on base. Of course, the whole de-romanticization process was then quickly re-romanticized by Michael Lewis’ book and the subsequent movie. How can you not be romantic about moneyball?
Scoring the Modern Game
I said before that I think scoring ballgames has a place in the era of sleek statistics websites, and I think it’s because of this romanticism. We think we’ve forgotten it—that somehow Savant present us with objective truth. But the truth is that we’re just as likely now to romanticize someone’s barrel percentage and hard-hit rate as we were to romanticize a prospect’s plus-plus speed and bat-to-ball skills in 1997, when Billy Beane was promoted to Athletics GM and we all jumped onboard the statistics train. Scoring a game: the essential boredom of it, the tediousness, the way every pitch might set off something interesting, keeps you focused. In the end, it teaches you how stats are made. Not all at once, and never in the same way twice. Now, in some ways, is exactly the time to be doing something so analog and backward, so slow and myopic, as writing on paper.