Posted by Allanfan20:
tomnerve, you're post is well researched and well thought of and you deserve 2 thumbs up, but there's some flaw to it, as well as JoeC. The way I see it (And I think there's a few posters here who would agree) that using purely stats does not tell the story.
You're right that stats don't tell the whole story, of course, but on the other hand they do tell us something. Both Frye and Lee had PERs 4-5 points higher than other rookie in their MPG ranges. That's not an insignificant difference, but rather it's a huge one-- about the same difference between the overall average NBA player (PER set at 15) and the average borderline All-Star (PER usually ~20), or the average borderline All-Star and the average All-Star lock / borderline MVP candidate (PER usually ~25).
There is certainly more to the story than is told by PER (which in this case is just my quick-n-dirty stat to get a feel for performance). For instance, Frye's help D certainly did look awful at times, and that doesn't show up in the stats. (Although, of course, it may very well be that the better parts of Frye's D don't "show up" in the memory of the typical Knicks fan either-- subjective impressions come with their own systematic flaws as well.) Nonetheless, the PER disparity between Frye and other rookies with similar MPG is undeniably quite large. In order for that disparity to be "cancelled out" by all the things missed by PER, all the "intangible" sorts of things about Frye would have to be far, far worse than they are for those other rookies with similar MPG, and I very much doubt that that's the case. Frye may have been lacking on the D relative to other rookies in his MPG range, for instance, but if so then probably not to a massive degree. And his other "intangibles" (effort, etc) seemed to be solid last season-- probably on a par with, or better than, fellow rookies with comparable minutes.
So in a nutshell, no matter how many caveats one might like to introduce about statistical analysis of basketball (and there are good ones), the sheer size of the disparities seen here are going to carry through. Frye and Lee both played much better than other rookies who got similar PT, at least in terms of everything measured in a box score. Of course, one can take that information and make of it what one will (perhaps you might make an argument that the rookies who got similar PT to Lee actually should have gotten much less than they did, rather than Lee getting more, for example). But it's hard to deny the underlying differences in performance there.
In terms of that PER stuff, I don't really understand it, but I hear it's just an indicator of your general on court performance.
PER is basically a way to take every stat you'll find in a boxscore, give each stat an appropriate significance, adjust each stat for confounding variables like minutes played and team pace, and combine it all into one number. You put it well when you said it's an "indicator" of team performance. It's not the end-all, but for many purposes it's a useful tool for getting an objective sense of overall player performance (at least in terms of everything measured in a box score).
help treat disease with your spare computing power : http://www.worldcommunitygrid.org/
