The Baseball Analyst Issue 2: a review

Bill James published 40 issues of The Baseball Analyst in the 1980s. All are available on SABR’s website.

The second, August 1982, issue of the Baseball Analyst features a fine Dallas Adams contribution, interpreting careers as expressed in major league games played by age (completely without examining any other statistics). While there’s really only some perfunctory analysis, there’s a ton of information about the aging patterns of ballplayers, displayed in three tables and a half-dozen graphs. Adams concludes that Hall of Fame players are about two years “better” than the average everyday player. He also breaks out catcher careers, and concludes that they reach the majors about a year older but are otherwise pretty much indistinguishable from everyone else. Interesting stuff.

Craig Wright offers a critique of Paul Schwarzenbart’s fielding analysis in the first edition of the Analyst, pointing out some weaknesses and commenting on some findings. This works all right as a comment, but there’s no analysis and Wright’s alternative numbers are only superficially explained. Of course, this journal’s audience knew Wright’s credentials, so this isn’t really problematical.

Dick O’Brien’s two brief contributions–about batting order and clutch hitting–are too brief and too dense to be readily accessible. What does show is that he’s done a lot of research and analysis.

Jim Morrow tries to improve Slugging Average. He offers some statistical analysis attempting to break scoring into its components, using regression analysis. The effort’s a failure in the sense that his “linear weights”-like calculations are less predictive of actual performance than SLG. Why the technique failed merits further investigation; that does not occur in this issue of the journal.

The last article is another Dallas Adams study, this one apparently showing that heavily-used rookie pitchers tend to have short careers. The general notion is now accepted wisdom, but I’m frankly skeptical about his underlying data; it looks skewed in ways that really needed discussion. I’m aware of commentary on this study in subsequent issues of the Analyst; we’ll see how that develops.

Once again, these offerings demonstrate an enormous amount of data collection and number crunching, long before Retrosheet and Baseball Reference. (Bill James comments on that in the introduction.) All in all, a useful outing, but flawed.