Re: [APBR_analysis] Re: On defense and methodology
- On Sat, 1 Nov 2003, Mike G wrote:
> Too much coffee, John ?Heh. Too little, I guess...
>Malone did play center a little bit, and the Jazz did kind of suck when they did so. That being said, I think you can trace Ostertag's fake high score to two things:
> My hangup with 82games is, first, that we see ratings like the
> Ostertag phenomenon; and second, while he/they claim to have an
> alternate set of +/- numbers that are adjusted to the competition --
> lineup vs. lineup scores (I guess) -- they don't reveal them.
> Ostertag is probably sitting on top of a statistical anomaly heap of
> mitigating factors. Here's my suspicion:
> --He is never asked to carry the bulk of the offense or the defense;
> i.e., he's never one of the primary scorers in the game, and always
> has defensive help. Maybe Sloan doesn't want him to get frustrated
> and foul out.
> --His backup is terrible or nonexistant. If Malone had to play
> center, probably the offense and defense suffered.
- He was usually on the court when both Stockton and Malone were, and
- There's one lineup that was used around 1.5% of the time (50 minutes or so the entire year) that had Ostertag in it and neither S nor M that for some reason really outscored their opponents, I'm guessing because it was so strange coming from that team that people had no idea how to match up against it. The lineup, Jackson-Cheaney-Kirilenko-Padgett-Ostertag, strikes me two ways: quick (compared to what the Jazz normally run, anyway) and garbage time-ish.
>Another problem I saw was with Vlad Radmanovic of the Sonics. Now, I am a big Vlad fan, but he was NOT the best player on the Sonics last year, and the team was probably right to draft another 4 to push him this year (as it turns out, said 4 will miss the season, but you get the idea). Radmanovic played a lot when the Sonics started out 8-2, didn't play much when they went 15-28, and then started to play a bunch again when they finished the season 18-12. Some of that improvement is due to Rad playing better in the 2nd half of the season, but most of it has nothing to do with him and he just had the good luck to lose his job right when things went really rough.
> So, I'm still hoping Roland et al have the actual numbers that will
> sort these features out from the general background of "in"
> vs. "out" -- which doesn't tell us a whole lot after all.