Notes on Rogers Hornsby: The limits of fielding statistics

Rajah fielding
Sure, he hit a ton, but how was Roger Hornsby in the field?

As I mentioned in the baseball rankings, I rated Rogers Hornsby higher than Bill James did in the “New Historical Baseball Abstract,” which, at the time, was the sort of thing that would send me back to the drawing board to re-check all my calculations.

In double-checking everything, I came across some interesting discrepancies in how Hornsby’s fielding statistics were interpreted, especially his stint at shortstop early in his career, in 1917 and 1918. James’ Win Shares system gave him credit for 59.63 Defensive Win Shares in 2,259 games played, which is abysmal. For reference, Joe Morgan has 90.94, Napa Lajoie (whose defense James spends several pages criticizing) is at 95.91 and Eddie Collins reaches 108.71. Compared to those, 59.63 is … not good.

Linear Weights has Hornsby even worse, with -77 fielding runs. Baseball Reference, however, has him as a positive fielder, with 54 runs above average from fielding, leading to 13.9 defensive Wins Above Replacement.

Without getting to far into the weeds on how the same set of defensive statistics can be interpreted so differently, I want to focus on those two years at shortstop. WAR, which sees Hornsby as a slightly above-average fielder overall, rates him as a genuine asset at shortstop. Of those 54 runs saved, 18 came in 1917 and 10 more in 1918. In addition, WAR grants a player runs for “positional scarcity” and Hornsby gets 10 and 7 for his full-time shortstop years. As a second baseman, it was never higher than 5.

Linear Weights also likes his shortstop years, granting him 19 fielding runs in 1917 and 9 more in 1918 (out of, remember, -77 overall). Win Shares is a little less bullish: His 7.01 defensive Win Shares in 1917 is the highest of his career, but he gets a mediocre 3.52 in 1918. Whatever they think of his defense, all three systems rate him as a better shortstop than second baseman.

The problem, of course, is that he wasn’t. Hornsby committed 52 errors in 1917 (tied for second-most in the National League) and 46 more in 1918 (fifth most). His range was pretty good, and 1917 in particular was a bad year for shortstops, so the analysis sees him as OK by comparison. But in the real world he was a liability at short, and 31 games into the 1919 season, the Cardinals had had enough. They moved him to third base, where he was even worse, and finally to second, where he would play indifferently but competently for the rest of his Hall of Fame career.

Apples & Oranges sees Rogers Hornsby as a slightly below-average shortstop, and an average second baseman, rating him a bit higher than Win Shares, a lot higher than Linear Weights and a lot lower than WAR. He finishes with 93.94 games’ worth of defense, compared to 126.66 for Collins, 117.11 for Morgan, and 111.28 for Lajoie.

One thought on “Notes on Rogers Hornsby: The limits of fielding statistics

  1. Pingback: Rogers Hornsby – Apples & Oranges

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s