Mosi Platt – of “Miami Heat Index” – has offered a fascinating look at Team USA at Bleacher Report. It is fascinating because
- it uses Wins Produced (always great)
- it has a slide show (also, always great)
In all seriousness, everyone should take a look at Kevin Durant & Team USA: A Less Dominant Road To Redemption. Mosi – or Reservoir God – has done some excellent work.
I enjoyed that thorough analysis. It helped explain why Rondo might have perceived himself as on the bubble, or why Coach K was not pleading with him to stay. It also explains why this “B” team might win it all: team defense, along with opponents who also have lost people since the Olympics.
On the other hand, the warm-ups for the tournament provide such a small statistical sample that it seems very hard to predict the outcome. Furthermore, there’s a very real possibility that at least some warm-up teams deliberately tank against the U.S. so as not to give away any of their strategies — Greece, in particular, might have adopted that strategy.
Like the NCAA tournament, the World Championships are unpredictable because so few games are played. Better teams can lose to worse teams, hot players can defeat cold players. Surprise strategies or bad match-ups can throw players off guard and create an upset. Coaching, teamwork, and “hot” players may matter more than they would over the course of an entire NBA season, or even over a seven-game series in the playoffs.
I’m also interested in the performance of players who are coming to the NBA after this tournament — especially Omer Asik, who is coming to my Bulls. I would love to see him perform well, although I’m more interested in seeing Rose improve. Rose is playing at a star level, but not a superstar level. Still, if he keeps up this level of play throughout this season it will be an improvement over last year.
What’s interesting is that Rose is being treated like the second “star” on the team after Durant. The stats don’t show him being that much different from the rest of the guards — in fact, he’s been outplayed by Westbrook and Gordon — but for some reason he seems to get star treatment. I’m happy about that even as I am skeptical. The stat geeks all think he is overrated, but maybe he’s just young and, like Durant last year, Rose will blossom (no pun intended) next year.
DJ – Thanks for the compliment and posting the link. BTW – You can call me Mosi. I was just joking about the secret identity.
Chicago Tim – Good point about the similarity between the World Championships and the NCAA tournament. It’s not single-elimination, but one game can waste a larger body of work. One point I should’ve made is something that’s often brought up here – has our bias towards offense driven much of the rhetoric that this team is nowhere near as good as the Redeem Team? Would we appreciate this team more if we were more appreciative of defense?
reservoirgod — Wins Produced does take into account rebounds, blocks, and steals, all hallmarks of good defense. Arguably even offensive stats are hallmarks of good defense, since defensive stops often lead to easy points at the other end. I am more inclined to believe that we just have a small sample of games so far, and that the 2010 team really is significantly worse than the 2008 team — but still good enough to win the tournament.
But let’s revisit this subject after the tournament is over, please!
Chicago Tim:
First, I agree w/ you about sample size but I’ve always found it interesting when star players get together on one team. As for an appreciation of defense, I know that I’m guilty of an offensive bias. I posted a comment on Andres Alvarez’s blog that said I would prefer to watch Carmelo over Iguodala if I knew my team wasn’t a title contender, even if it meant a few less wins. WoW Journal readers may appreciate the outputs of good defense more than the average fan (who typically only cares about double-doubles, triple-doubles, breakaway dunks and shots swatted into the mezzanine). But I’m not sure an appreciation of Wins Produced = an appreciation of the process for producing a stellar defensive efficiency. With that said, I also tend to think the 2010 team won’t be as dominant as the Redeem Team but definitely good enough to win the World Championships. My offensive bias question was really geared towards those fans and writers that think this team isn’t good enough to win the gold medal.
Mosi,
Awesome post and a good read after my vacation. Loved the bit about Westbrook. He and Durant on OKC next year will be scary. Gordon and Westbrook got minutes in the 1st game and played well. Love got 10 rebs in 13 minutes.
Before I ask this, I want you to know that I do agree with the wins produced metric and have read many posts on this blog and the other affiliate blogs, but I’m still confused about something. In the team adjustment, how is pace adjusted for. I know it is, but I don’t see how it’s adjusted for. I read how it’s calculated, but I see defensive factors. I don’t get how pace is adjusted for. I have read that wins produced is calculated per single possession as opposed to 100 possessions. I just want to understand specifically how pace is adjusted for.
Also, hoopdata.com tracks charges as well as And1 shots. Couldn’t you use those two metrics to make the wins produced formula even more accurate. It does an excellent job right now, but by adding charges and getting a specific number of how And1 freethrows a player takes can make it even better.
Thanks, Arturo. I plan to post recaps throughout the World Championships but this was a bad weekend for me to start so I’ll have to get caught up on Monday.
EntityAbyss:
Prof. Berri posted on charges some time ago. If you search the blog for charges it should come up.
Yea, I read his post and comments on charges. I was just wondering why he didn’t use it. Also, I’m confused as to how the defensive adjustment adjusts for pace. For a fast paced team, wouldn’t their wins produced add up if they played above average defense and offense?
EA,
I believe the prof has tried every variable to see what can add to the consistency and explanatory power of the formula. What works is in there, what doesn’t work isn’t in there.
EA,
The metric measures marginal win value produced by a team per 48 minutes so pace is actually built in. So if you typically play a 220 possession game your marginal value will account for that. Since marginal value is calculated vs average league wide production a defensive adjustment is required to account for how more/less productive a teams typical opponent is.
One thing that’s extremely interesting is that the model points to the fact that slowing the game down makes your cumulative marginal advantage smaller. So bad teams need to slow the game down to improve their chances against better teams and good teams need to speed it up.
I’ll point you here:
http://arturogalletti.wordpress.com/2010/07/19/measuring-the-quality-of-basketball-in-the-nba-part2-adjusting-for-pace/
and here:
http://arturogalletti.wordpress.com/2010/07/05/defense-adjust-wins-produced1st-pass/
for some of the work I’ve done around this.
Arturo,
The difference in pace between even the fastest and slowest NBA teams isn’t much more than 10 possessions/game, about 12%. While an increased sample size will always benefit the better team, whether it be over the course of an entire series or a game, a 12% difference will never be huge.
I would expect it might be a bigger factor in the college game, where difference in pace is far more pronounced; grinding down pace might be a worthwhile strategy for an underdog that benefits from fewer possessions (and thus a smaller sample size). But I expect that the minor returns for an NBA team would be outweighed by the losses of artificially changing their own pace, going against their playing style.
I would guess that some teams play better fast, and some play better slow. For instance, the early 2000s Lakers were dominant, but needed to play slowly to accomodate Shaq. The Spurs have been dominant forever, and have always played slowly. I realize this is anecdotal, but it does pass the smell test.
(Which I realize isn’t a test at all).