As I understand it, he looked at how individual players relative performance at different positions. DDJ would be a good example—maybe DDJ is a -5 centerfielder, but a +10 left-fielder, so Tango’s method would record that as a data point that says center field is worth 15 more runs than left field. I believe he ended up concluding that a league average offensive player playing league average CF is worth .5 more wins than a league average offensive player playing league average LF.

I’m not at all impressed with the method. There’s an implicit assumption that the required skills for CF, LF, and RF are all the same, and a player’s performance at each of those positions is simply a matter of the quantity of those same skills that the player has. Additionally, this method only uses data points from outfielders who play both CF and at least one corner outfield position. Can we really conclude that since there’s about a 20 run difference between CF and LF for tweener outfielders like DDJ, that this would be the case for a Manny Ramirez?

I’m sure I can come up with more objections along those lines if we want to continue, but like I said, the relative values between the OF positions are the strongest points of the adjustments. The absolute values are, in my opinion, frightening. I’m sure you can find his write-ups with a little google-ing, but basically, to get the absolute adjustments, he looked at the relative performance of players who split time between corner outfield spots and first base. Then he looked at the relative performance between players who split time at first base as compared to other infield positions. That’s two very tenuous leaps since it’s very clear that the necessary skills are not at all the same between OF and 1B, nor are they at all the same between SS and 1B. Meanwhile, the cather adjustment is quite literally pulled out of thin air.

The best part is that after he did all that, he decided that the adjustments for the OF were to generous while the adjustments for the infield were too small, so he essentially bumped the infield spots up .5 wins for not throwing left-handed. Look it up, I swear.

The end result of all this gives you position adjustments that look about like you’d expect. So use them if you like, but take them as common baseball sense guesses rather than robust statistics.