November 5, 2009
Cavs Through 5: WP48 & PER

I wrote about WP48 earlier and will add to this conversation soon with a post about +/- which, arguably, tries to measure a player’s defensive impact outside of blocks, rebounds, and steals. But for now, here’s another breakdown of the Cavs through 5 games, courtesy of the birthday boy, Holland. Remember .100 is average:

WP48 (5)         WP48 (4)     WP48 (2)
LeBron             .383            .361           .365
Shaq                .175            .061           .018
Mo                   .077            .045          -.075
Delonte            .178            .308
Varejao            .171            .118          -.125
Z                     .022            .103           -.021
Parker             -.049            .053          -.068
Moon                .082           .114           .177
Boobie             .108            .040           -.068
Hickson           -.170           -.202           -.337

You can see how these cumulative numbers shift from game to game - the stats through 2 games and 4 games are also listed. And with a small sample size we shouldn’t jump to any significant conclusions (upwards of 20 games will give us a better indicator).

I’m generalizing, but much of what we see here is a result of how well a player shoots the ball, how well he holds his turnovers down, and how well he rebounds. This, I suspect, is why Boobie is above average for the first time in a long time (he was at .147 in the ‘06-‘07 playoffs) - he’s shooting 52.2% from 3P, and I still maintain that a high volume, high percentage three point shooter is the most valuable single skill in basketball. If they’re a good defender, like Boobie was on Tuesday night against Gilbert Arenas particularly, then that’s an added bonus.

Anthony Parker, on the other hand, who’s shooting 33.3% from the field and 47.4% from 3 - while averaging 2.3 AST/36 minutes and 1.7 TO/36 - has been less effective. This is, in part, due to the fact that he’s taking about 5 two point shots for every 2 threes. AP won’t continue shooting 33.3% from the field for the rest season (he’s at 45.5% for his career) but one could argue that his shot selection has been poor, as the percentages suggest he should be taking a lot more threes than twos.

This is a good example, I think, of how statistics can help coaches and players strengthen on court performance.

The big thing to monitor over the course of this season is going to be the playing time of Jamario Moon. In limited minutes against the Wizards on Tuesday, I thought he played well, staying in front of his man on defense, hitting a big three, and rebounding at a high rate (3 boards in 9 minutes). Historically, he’s a more productive player than Parker, but he seems to be losing minutes to Boobie and Delonte at the moment. I believe this can be attributed to match-ups. When the Cavs play teams like Toronto and Orlando with bigger, stretchier SFs and PFs, Moon will likely steal back the minutes Boobie has earned.

But time will tell. Coldstone has a tendency to not be very consistent with his rotations, which is one knock against him as a coach. Part of what’s good about this Cavs team is its depth, and the many different line-ups that can be thrown out on the floor, with multiple players who can play multiple positions. But this also gives Coldstone a lot of options - and that may not be a good thing. When Leon Powe comes back in February, the rotation will go into experimentation mode again, likely. I mark this down as a continuing development to monitor.

Keep in mind that with these WP48 numbers, two adjustments are being made to the pure box score statistics, for position and for team in order to set the average and predict future performance. These adjustments have been somewhat controversial in the APBRmetrics Forum, but I’ll discuss this a little bit more when I write about +/-

Just for kicks, I’ve decided to add the PER stats for each of these Cavs players, so we can compare how the two metrics measure each player’s value. For PER 15.0 is average.

WP48 (5)         PER (5)
LeBron             .383            29.0
Shaq                .175            17.9 
Mo                   .077            16.1
Delonte            .178            16.3
Varejao            .171            14.8 
Z                     .022             13.3
Parker             -.049           7.4
Moon                .082           10.5
Boobie             .108            12.5
Hickson           -.170           .9

I’m surprised that Boobie’s PER is below average and his WP48 is above. I’m not surprised by the discrepancy in Andy’s numbers or in Mo’s, given the value PER places in shot attempts over elements like rebounding.

Still, all of this adds up to an interesting look at the Cavs early on and by the numbers. We’ll try to continue peeking at the stats over the course of the year in intelligent ways that help inform what we see on TV three or four nights a week.

Bulls tonight.

blog comments powered by Disqus