February 25, 2004
Baseball Prospectus Basics
How to Read a Box ScoreIn our introduction to the Baseball Prospectus Basics series we wrote, "We always want to improve our understanding of the game--each player, each play, each pitch, each throw, each hit--what does it really mean?" We have a storehouse of data to help us refine our understanding of how baseball really works and how it can be improved. We have a team of performance analysts who help us see things we might have never perceived on our own. But the unrefined essentials of what we use are harvested from the box scores you and I read every morning from April through October.
The title of this essay is misleading: there is no correct way to read a box score. Roto gamers approach a box score like it's a greatest hits record. Retrosheet's patrons dust each stroke of agate as if it was an artifact. You pay for the morning paper, you get to use your box scores however you wish, even as fishwrap.
Box scores now tell us nearly everything that occurs in a game. They tell us hot warm it was, the direction and speed of the wind, and how many people came out to the park. We can find out who the umpiring crew was. Baserunning blunders, substitution patterns, clutch hits, high-leverage relief appearances--it's all in a good box score, along with groundballs, flyballs, balls, strikes, and pitch counts.
In this ubiquitously electronic age we can get lazy. We become passive fans. We scan box scores for the highlights, and let "Baseball Tonight" tell us who hit the screaming taters. But with enough scrutiny we can figure out for ourselves most of what happened in a game simply by poring over its box score. Last summer I wrote an article that looked at the eighth inning of a superficially nondescript game between the Twins and Tigers, played in front of 9,000 fans. From the notation of a stolen base at the bottom of the box score--that game's most obvious point of fantasy interest--I was able to extrapolate everything that happened in the Twins' half of the inning: who made the outs, what kind of outs they were, who reached base and how, when a wild pitch occurred, how the frame's run was scored, and a lot more. And from that, the ninth was ascertainable too. The next week, someone wrote to ask why someone would bother to recreate a game from an online box score when with one click he could go to the game summary. It's not for everyone, and it's a lot easier and quicker to read summaries, but some people still enjoy taking a watch apart to see how its gears work.
Deconstruct box scores, and you begin to see trends. Last year you'd have seen Lance Carter resurfacing as the Devil Rays' closer in April. You'd have seen that Tony La Russa had burdened himself with a bench full of players who could play a few positions adequately but couldn't produce runs, and how it would cost him precious ground by the end of the year. And while a lot of people were talking up the Reds as a playoff contender in May, you'd have seen that their 21-19 start was misleading, and how their pitchers were going to force the team back towards the Pythagorean pace they were on. After 40 games the Reds had a Pythagorean record of 16-24, on a course for 65 wins. They ended up with 69 wins and a 19-game deficit in the standings.
Last year we showed you these pitching lines from an A's game:
IP H R R BB K HR Rincon 0 0 0 0 0 0 0 Bradford 0.2 0 0 0 0 0 0 Foulke 1.2 0 0 0 1 0 0Ricardo Rincon's line triggered some questions, and we found out that:
A pitcher who is announced but who does not throw a pitch due to an injury gets credited for one game in the official batting stats but no game pitched--analogous to a pinch hitter who is announced but does not bat.
Whether the pitcher is included in the box score if he doesn't throw a pitch depends on the organization doing the scoring and their level of expertise. The official scoring is as above, but the official scorers don't prepare a newspaper-style box score.
The box score's Batters Faced summary says Rincon faced a hitter. His blank line could result from a caught stealing, hit batter, or an error. He didn't catch anyone stealing or hit anyone. So what happened? The A's committed three errors, so one of them has to account for Rincon's batter.
And then we had the Tigers. Two box scores early in the season gave a hint that 119 losses wouldn't be out of the question. Three weeks into the season, and Alan Trammell's wankstain was all over Detroit's box scores. On April 16, Trammell tried Bobby Higginson in center field. It was the first time Higginson played the position in the majors, and he didn't play there again. And then a few days later, with his team mincing along at 1-14, and losing its 16th game 9-2, Trammell had Craig Paquette pinch hit for Higginson. It was the eighth inning, and Royals reliever Albie Lopez was working on his fifth consecutive scoreless inning, so maybe it was despair, but the move had no strategic justification. Removing your best hitter is the worst way to start a rally, isn't it? Paquette was Trammell's de facto designated pinch hitter, so maybe he was trying to keep Paquette fresh. (For the year, Paquette managed five singles and no walks in 33 plate appearances.) Pinch hitters come in cold and they can't really be kept fresh, but if even you could crisp them up with two or three swings every other day, what would be gained by wasting one of your star's turns on a bag of sand like Paquette? It wasn't a platoon decision. Higginson is a lefty and both Paquette and Lopez are righties. If Higginson was injured you couldn't tell from the box score; he had played the previous inning and was back in the lineup the next day, and there was nothing in the next day's news about any injury. Either Trammell was conceding the game by giving Higginson a breather or--this has to happen sometimes--a leisurely bathroom break. OK, so box scores don't contain everything.
A box score does tell us about personnel, context, and tactics, not just performance. They lead us to learn some things about the rules too. Analysts like Bill James, Pete Palmer, and Clay Davenport have given the world some fascinating metrics, but there's no secret knowledge at work here. Just about everything we need to know to understand the game is contained in the box score.
If there's a correct way to read a box score, it's carefully.