scouting and scoring

Reviewed by Tim Morris, University of Texas at Arlington

8 august 2019 archive

A book on the history of record-keeping in baseball might not seem in the wheelhouse of Princeton University Press, but Princeton has been interested lately in qualitative studies of the data that increasingly guide all decision-making. Scouting and Scoring is thus well in line with a recent Princeton title like Jerry Muller's Tyranny of Metrics (2018). Author Christopher Phillips argues that "we have to think more carefully about how [people] create data" (5).

Books on the new data sciences characteristically spend little or no time discussing the human labor by which data are made. … The belief … is that with enough sophistication in processing and analyzing, any faults or improprieties in collection can be managed. (5)

In mathematical terms, that may be so. But Phillips, via baseball, observes how the stuff we think we've reduced to pure numbers is often a mess of judgment calls.

Scouting and Scoring is a definitive study of one aspect of baseball history, or perhaps parahistory (not in an occult sense; but scouts and scorers are not the game proper). Phillips does not really make baseball records a platform for larger critiques of how we generate and analyze data. He sticks to careful documentation of how baseball has arrived at its exemplary statistical archives.

Baseball is often invoked as an ideal case of the reduction of human activity to measurable quanta. Baseball's discrete units of play lend themselves to tabulation, and those tabulations can provide real insights that feed back into better strategies and tactics. Baseball statistics seem so transparently "given" (the root sense of "data") that we rarely think about them. But take a double, for instance, a baseball event that Phillips frequently invokes. The batter hits the ball and ends up on second base: what could be clearer? But recording a double depends on a decision made by the official scorer. Some significant part of the time, the scorer must decide whether it was a double or a two-base error; whether it was a double or a single where the batter-runner advanced to second on a throw. How do we know a given event occurred, till someone actively decides it did?

And how do major-leaguers get to the majors to hit those maybe-doubles? At some point, before a major-league prospect accumulates a reliable statistical record, he must be judged directly by a scout. Thanks to the film Moneyball, even more than Michael Lewis' book, people often see stats and scouts as using diametrically opposite worldviews. But Phillips shows how, within the scouting fraternity, there has been a constant drive through the years to objectify and standardize qualitative observations.

The audience for Scouting and Scoring is likely to be limited to people who already know a good deal of the history it conveys – those most interested would even know some of the people that Phillips mentions as involved in the rise of sabermetric data-gathering. The sections on both scoring and scouting are a little abstruse even for me, and I would seem within the target audience. But via some extrapolation, I think many readers could draw useful insights from Phillips' book. Statistics on the economy, crime, trade, immigration, and healthcare – any of the key issues of our time – are generated by methods analogous to those that Phillips painstakingly documents for baseball. Numbers are made, not found: which is not to say they are fabricated, only that they must be the result of judgment calls, like those made by scorers and scouts. In an era when we turn to statistics to validate ideologies or even morality, we should be more aware of how those statistics come to be.

to the top of this page