In "Hey, Bill" we have drifted into a discussion about Doc Cramer, 1930s/1940s American League outfielder.    How good a player was he or, more accurately, how bad?

                Doc Cramer was kind of a poor man’s Richie Ashburn.  Although he was tall and thin, he was a singles hitter who didn’t walk and didn’t steal bases.  He was a left-handed hitting center fielder who rarely struck out. His batting averages were pretty good. He was a career .296 hitter, and he had 2,705 career hits.   But his walk rates were below average, and he would sometimes go thousands of at bats between home runs.  He didn’t steal many bases, and his stolen base percentage was horrible.  While he did score 100 runs in a season three times and scored as many as 116 in a season, he also made upwards of 450 outs almost every season, with a career high of 501.   Per 162 games played, Cramer would bat 661 times and get almost 200 hits, score almost 100 runs, and strike out only 25 times—all very good numbers.   But he would hit only 3 home runs, per season, with less than 30 doubles, 41 walks, 4 stolen bases, and almost 500 batting and baserunning outs.

                The other rating systems rate Cramer as a pretty awful player.  Retrosheet, using (I think) Pete Palmer’s methods, rates Cramer as 29.6 games below average over the course of his career.   Baseball Reference rates him at 228 runs below average, or 22.6 games below average.  In spite of this, Cramer was named to five All-Star teams, and was mentioned in the MVP voting five times.  Altogether, he either was named to the All Star team or was mentioned in the MVP voting in eight different seasons.

                So he was perceived, while active, by those who saw him play, as a pretty good player, but he is perceived by modern statistical ranking systems as dead weight.   So what is the truth here?

                Well. ..I don’t mean to be arrogant.  I respect the opinions of those who saw him play. However, my methods—while very different from the Baseball Reference methods—reach essentially the same conclusion.  I am also unable to document the good qualities that 1930s sportswriters saw in Cramer.

                Baseball Reference lists Cramer at +12.6 Offensive WAR, but negative 6.2 Defensive War; in other words, they see him as a bad hitter, but a worse fielder.    I see him as a not-very-good fielder, but a much worse hitter.    But adding it together, I have Cramer with a career won-lost contribution (Win Shares and Loss Shares) of 226-288, which is essentially identical to the Baseball Reference rating.  

                Why the difference?   Well… I don’t exactly understand how Baseball Reference does their rankings, but I’ll explain it as best I understand it, and if I’ve got it wrong, I’m sure some of you will be happy to correct me.    Baseball Reference, I think, compares each player to an average hitter at his position, and to an average fielder at his position.   Their conclusion is that while Cramer is below average in both areas, he is further below average as a fielder.

                This method implicitly assumes that an average center fielder is as good a player as an average first baseman or an average third baseman or an average numismatist—in other words, that the average shortstop is as good a hitter as the average first baseman, position-adjusted, and that the average first baseman is as good a fielder as the average shortstop, position-adjusted.    Of course this is not true; the average shortstop is a much better fielder than the average first baseman, but a much weaker hitter.

                My method, on the other hand, assumes that

                1)  All hitters compete with all other hitters, but

                2)  The defensive players at some positions are much better than the defensive players at other positions.

                Given those assumptions, Cramer is further below average as a hitter—because he is being compared to all hitters, rather than being compared to other center fielders—but not as far below average as a fielder, since center fielders are assumed to be carrying a larger defensive load than players at most other positions.

                I like my assumptions better and my method better, but it doesn’t really matter; we wind up, in this case, at exactly the same conclusion.   62 Win Shares below average is 22.7 Wins below average. 

                This chart summarizes Cramer’s Win Shares and Loss Shares: