Monday, October 27, 2008

The Big NBA Season Preview, Round II - Statistical Odds and Ends

This is kind of the trivia section of my NBA Season Preview. Think of it as little bits and pieces that may shed some light on the shape of the league going into opening night.

I. Pythagorean All-Stars
Anyone who has followed sports for a while, particularly baseball, is probably familiar with the notion of Pythagorean wins. The idea is that a team's scoring differential or efficiency differential (efficiency is just score adjusted for pace) has more to say about how good a team is than its actual record. This is as valid in basketball as anywhere, where the outcome of a close game that involves 90-100 possessions for each team might hinge on one lucky bounce during one of those 180-200 possessions, meaning the winning team didn't exactly prove that it is definitively better than the losing team. Meanwhile, a team winning by 10 or more points generally played better than the team it beat, and wasn't relying on a lucky bounce. Now, we expect lucky bounces in close games to even out over time, so here I want to take a look at teams that were significantly better or worse than their records indicated. I used the season summary from Basketball Reference for the data.

A. Toronto Raptors
Last year, the Toronto Raptors won 41 games, which makes them sound like an about average team. However, their efficiency differential was that of a 49-win team, making them one of the top 4 teams in the East last year. Unfortunately, they went up against one of the other three (Orlando) in the first round of the playoffs. Still, if the Jermaine O'Neal trade works out, we might see this team win 50+ games next year, and make it at least to the second round.

B. New Jersey Nets
The Nets won only 34 games last year, but they were even worse than that record indicates -- putting up the efficiency differential of a 27-win team. We were already expecting this team to be worse this year due to the loss of Richard Jefferson (Yi Jianlian isn't, at this point at anyways, an equivalent replacement), but even if they were exactly the same team we'd have to expect them to win 7 fewer games than they did last year. Combined, these factors make it look like we might see the bottom fall out a little bit, with 20-25 wins.

C. Cleveland Cavaliers
The Cavs went 45-37 last year, but during the season they were actually outscored by their opponents, meaning they might just as well have been a sub-.500 team. I expect them to be improved this year, but now I'm wondering if an effectively 40-win team has made enough changes to win 50 games this year, as I predicted in the previews. As it is, they might end up being a much-improved team and still winning 45-47 games.

D. Utah Jazz
Utah was 54-28, but they underperformed their efficiency differential by 5 games last year. At 59 wins, they could have ended up with homecourt advantage against the Lakers in the second round of the playoffs last year. I'm not sure that it would have made a difference, but it's interesting to consider. With few roster changes, it shouldn't be a huge surprise if a healthy Jazz wins 60 games this year. Unfortunately, Deron Williams has already suffered an ankle injury, but reports are that he won't miss too much time (Williams played all 82 games last year, and has never played fewer than 80 in a season).

II. Small Sample Size All-Stars
Every year there are a handful of young players who look like potential breakout candidates in extremely limited minutes. Here are some such players from last year who might be primed to break out if given some minutes.

A. Ramon Sessions
Ramon Sessions averaged 20.9 points, 7.6 assists, and 6.6 rebounds per game in the D-League last year. He also got to the free throw line an average of 9.2 times per game. As an NBA starter to finish out the year last year, he averaged 13.1 points, 13.1 assists, and 5.6 rebounds. Here is a list of players who were able to average 13, 13, and 5 over the season last year:


Nobody. In fact, forget about the rebounds, there was no one in the league to average 13 points and 13 assists. The only player to average over 10 points, 8 assists, and 5 rebounds was Jason Kidd. Similarly, the only players to average at least 13 points and 10 assists were Chris Paul, Deron Williams, and Steve Nash.

Overall, including the games where he came off the bench, Sessions played in a total of 17 NBA games last year (in addition to the 24 D-League games referenced above). In those 17 games, Sessions averaged 11 points, 10.2 assists, and 4.6 rebounds per 36 minutes. That's a more efficient version of Jason Kidd, or a lower-scoring version of Chris Paul.

With Mo Williams now in Cleveland, the only person standing between Sessions and a full-time starting gig is Luke Ridnour.

B. Leon Powe
Powe stood out in a handful of playoff games for the Celtics last year, but he was already looking good during the regular season. In limited time, he averaged 20 points and 10 rebounds per 36 minutes. He's not about to become a starter with Kevin Garnett and Kendrick Perkins in front of him, but I can see him having another solid season with about 20 minutes per game.

C. Ian Mahinmi
Like Sessions, Mahinmi was a D-League star, finishing his stint with the third highest PER in the D-League. He only ended up playing a total of 24 minutes in the NBA last season, but in that time he racked up 21 points, 5 rebounds, and 4 blocks, and shot 50% from the field. There's no way of knowing whether he'll be able to continue to produce consistently at the NBA level, but we might see him get the chance to crack the frontcourt rotation for the Spurs during the first half of the season to keep 36 year-old Kurt Thomas as fresh as possible for the playoffs.

D. Amir Johnson
Amir Johnson averaged 10 points, 11 rebounds, and 4 blocks (4 blocks!!!) per 36 minutes last year. He's already been penciled in as a starter in Detroit. As in San Antonio, Johnson's early-season minutes here could keep the wear and tear away from Antonio McDyess so he can be ready for the postseason.

E. Brandan Wright
Per 36 minutes, Wright averaged 14.5 points, 9.5 rebounds, and 2.1 blocks last season. I have no idea if Don Nelson will give him more minutes, but fingers crossed! While he has generally been reluctant to play rookies, Nelson has shown a willingness to play some second-year players, with Monta Ellis the most recent example. Then again, it's hard to tell what Nelson might do with frontcourt players who don't shoot the three. He couldn't be bothered to play Andris Biedrins more than 27 minutes a game last year despite excellent production at both ends of the court last year.

No comments:

Post a Comment