Good idea, flawed execution?

Justin Bopp over at Beyond the Box Score has attempted to explicitly visualize UZR/150 through what he calls the “UZR TargetView method.” I think it’s an excellent idea, as defensive metrics are often difficult to consume numerically. Thus, a graphical depiction would appear to be ideal in today’s very visual society. However, while I think he’s on the right track, I do wonder if Justin has implemented his idea correctly. For instance, note the following visualization of the Yankees’ UZR per 150 defensive games.

The problem with this view that Justin and others are discussing in the comments section of his post, is that it appears to be a strict visualization of “ground covered” by the players, rather than of UZR/150 which incorporates arm runs, DP runs, and error runs, in addition to range runs. If Justin is trying to provide an illustrative representation of the team’s range, which seems to be the case in this graph, maybe it would be best to utilize range runs instead of UZR/150. In my opinion, this would make the visual a more accurate one (or at least it would eliminate some confusion).

For instance, according to UZR/150, Nick Swisher is worth -1.6 runs below average in right field, meaning that he is basically an average right fielder. Justin has mapped that out with his UZR depiction. However, in terms of actual range (RngR), Swisher is 5.5 runs above average. His overall UZR/150 is hurt by his below average arm, therefore, Swisher’s representation in this particular visual seems to imply that he has limited range in the outfield when, in fact, he’s one of the best outfielders—range-wise—in the American League. Again, perhaps using range runs would make the analysis better, or, perhaps adding other illustrative elements to the visual would allow for a more comprehensive look at each player (UZR/150 is a nuanced statistic—it’s not all about ground covered).

Still, I’d like to give Justin credit because this is an excellent idea. The execution needs some tweaking, but the visuals are extremely interesting. Beyond specific implementation exercised here, what do you think about the graph? For the most part, it still rings true for the players although if range numbers are used, the depictions can fluctuate (it would not change the general direction of the Yankees graph, however). Robinson Cano, Teixeira, Damon, A-Rod—they all seem to cost the team some runs, according to range and that’s represented in this graph, whereas Jeter, Swisher, and Melky have been above average, range-wise.