Once again Henry Abbott has asked me to participate in the TrueHoop Stat Geek Smackdown. The contest works as follows: A number of statistically inclined persons choose the winner of each playoff series (you get five points for getting this right in each series). In addition to choosing the winner, we also pick how long each series will last (this can earn you an additional two points). Whoever earns the most points across the 15 playoff series of the NBA post-season wins the contest and gets a cool prize.
When it comes to stat analysis and basketball, disagreements seem to be everywhere. These disagreements, though, are mostly centered on the evaluation of individual players. The evaluation of teams is generally agreed upon. For example, the model I employ argues that the outcome of each series is a function of each team’s efficiency differential (offensive efficiency minus defensive efficiency) and home court advantage. Since I suspect everyone employs a similar model, how is it that we come up with different picks? Well, sometimes we don’t (as one can see if you review the picks from this year). But when there are differences it probably comes down to how we think various injuries and roster changes will impact a team’s chances.
What follows is the story I think these injuries and roster changes tell. Of course, others can – and did – tell a somewhat different story.
Let’s start in the Eastern Conference
Chicago (1) vs. Indiana (8)
Chicago has homecourt advantage and an efficiency differential of 7.8. Meanwhile, the Pacers are the worst team in the playoffs with a differential of -1.1. So this pick is easy.
Chicago 4-0.
Miami (2) vs. Philadelphia (7)
Miami has homecourt advantage and the league’s best differential (8.0). The Sixers had a nice season, but with a differential of 1.6, Philadelphia is not likely to defeat the Super -Friends.
Miami 4-1
Boston vs. New York
Boston’s efficiency differential for the entire season is 5.8. And although many people thought Carmelo Anthony would substantially improve the Knicks; that really didn’t happen. The team’s efficiency differential at the end of the season – a 0.8 mark – is not much different from what it was before Melo came to town. What has changed in the last few months of the season, though, is the play of the Celtics. Boston sent Kendrick Perkins to the Oklahoma City Thunder for Jeff Green and Nedad Krstic. Both of these players were below average with the Thunder. And both are below average with Boston. Furthermore – and not surprising (given the infusion of unproductive talent) — the Celtics efficiency differential since the All-Star break has only been 2.5 (thanks to Arturo Galletti for sending me the post All-Star differentials for each team). That suggests this series is going to be closer than the season differentials suggest. But although the gap is closer, I still am going to pick the Celtics to advance.
Let me note – as Henry observed – that I am the only one in the contest with this much faith in the Knicks. Perhaps that means I am the only one with this little faith in Green or Krstick (or maybe I am just confused).
Boston (4-3)
Orlando vs. Atlanta
Atlanta has gone 3-1 against the Magic this year. And the trade the Magic made last December really didn’t transform Orlando into a team that could contend with Chicago and Miami. That being said, the Magic’s differential for the entire season is 5.8. And Atlanta’s mark is -0.9. In other words, if it wasn’t for the Pacers, the Hawks would be the worst team in the NBA playoffs. These numbers suggest that Magic will be able to sweep the Hawks (by the way, I am the only one who thinks the Magic sweep the Hawks).
Orlando (4-0)
Now here is the Western Conference.
San Antonio vs. Memphis
Injuries have clearly hurt the Spurs (pun intended). Since the All-Star break the Spurs differential is 2.8. And the Grizzlies mark is 4.3 since the break (a mark achieved primarily without Rudy Gay). So this suggests the Grizzlies – a team that has never won an NBA playoff game – might defeat the number one seed in the Western Conference. Although I am tempted to make such a pick, I am going to guess that Manu Ginobili will eventually be healthy enough to make a contribution. So my pick is…
San Antonio (4-2)
LA Lakers vs. New Orleans
The Hornets were actually the last team to lose during the 2010-11. But New Orleans – with a differential of 1.0 – is simply not on par with the defending champions. So assuming Andrew Bynum can play (and I don’t know for certain this will happen), the Lakers should win take this series.
LA Lakers (4-1)
Dallas vs. Portland
Portland’s acquisition of Gerald Wallace has certainly helped the Blazers. And of all the teams that won 55 or more games this year, the Mavericks are the weakest. But I still think the Mavericks – with a differential of 4.5 and homecourt advantage in this series – are favored against Portland.
This is one of only two series where the participants in this contest disagree. And I certainly can see Portland prevailing. Of course, now that I have picked the Mavericks, I guess I will be rooting for Dallas to prevail.
Dallas (4-3)
Oklahoma City vs. Denver
This is the hardest series to call. Each team made a major trade this season. Each team got much better after the trade. The Thunder’s post All-Star break differential is 6.7. And that means the Thunder are comparable to the best teams across the entire regular season. The Nuggets’ mark after the All-Star break, though, is 9.9. This mark ranks the Nuggets among the all-time great teams in NBA history.
If we look at the performance of the players on each team before the trade, the mark we see for the Thunder conforms to expectations (as I noted a few days ago). The Nuggets were not expected to get much worse after Melo left town. But the performance we have seen in Denver since Anthony made people in New York so happy has not been expected.
So this puts me in an odd position. Do I take the Nuggets performance without Melo – across a relatively small sample of games – as the true measure of this team’s quality? Or do I argue that past performance of the players Denver employs is the best measure of performance? Given a choice, I am going with the larger sample. And that means, my pick in this series is…
Oklahoma City (4-3)
Okay, those are my picks.
The participants in this contest only disagreed on the last two series listed above. So those series will essentially determine who comes out of the first round in front.
- DJ
dberri,
Why don’t you factor in WP to your predictions? Unproductive players tend to have (some of) their minutes replaced by productive players in the playoffs. Top-heavy teams are obviously favored.
The Nuggets will still be the same Nuggets even if their starters’ minutes go up, but the Heat will be drastically better with Wade and James playing more. For the Magic, Jameer’s minutes going up (or perhaps more importantly, Arenas’s minutes going down) is huge.
I doubt it’ll matter in first-round series, which are still essentially coin-flips or quite lopsided with no in between. But it might in the second round.
Something else worth considering is that the Memphis looks VERY dangerous. They’re producing decent WP48, and Manu may not be able to play this serious.
Philip,
IMO you are making a really good point about depth being less of an advantage in the playoffs.
Without any back to backs, 3 nights out of 4, or 4 nights out of 5 on the road etc… coaches are way more willing to up the minutes of their starters, shorten the rotations, and not worry about who is going to wear down over the course of the season. It’s do or die now.
That could also impact Dallas negatively.
Setting aside the issue with Manu (which is critical), I think we can be sure Duncan is going to playing a lot more now.
There are a lot of cases like that.
This is somewhat off topic, but does anyone know if (and by how much) the average TS% drops during the playoffs?
Oh dear, Prof Berri. The non-Howard Magic (helped by Dwight’s 8 TOs) have ruined your prediction on the first night.
rad e cool
he helped a lot more with 46 and 19. not his fault his gm sucks.
Very interesting but this analysis fails to factor in the clutch players that step up their performance during the playoffs so outperform their regular season stats.
I think the issue of “recent” efficiency differential vs. seasonal efficiency differential is a very interesting one.
I’ve been casually tracking that for every team in the NBA this year. For most teams ED seemed to reach a normalized level within 20 or so games and then it fluctuated up and down around that. There were instances of sharp deviations when key players were injured or traded, returned from injury, or had mid season coaching changes and totally changed their style of play, but only a few sharp sustained reversals with no explanation.
The 76er were one strange case.
They started out terribly, then had a sustained period where they were playing almost as well an anyone in the league, then they tailed off. There were no clear cut triggers or changes I could identify.
The Denver case is especially interesting to me.
You can argue they were injured and beat up early in the season. So they weren’t playing up to their potential. Then they got healthy right around the time of the Melo trade. People are associating the improved play with the Melo trade when in fact several key players (Anderson and Martin) finally started playing well.
It’s really difficult to tell how good they are, but it’s probably somewhere in between their annual ED and their recent ED.
Also note that in addition to the Magic losing, the Bulls almost lost game 1. They’ll still win the series, but 4-0 looks like a longshot.
IS, can you elaborate on the negative impact for the Mavs? They strike me as the model “half-baked” team. If Dirk doesn’t take 20 shots a game.
Prof,
Why 4-3 on the Celts? I started this comment before the game but didn’t get to come back home until now. It seems to me that even with your data, 4-2 Celts over the Knicks looks much more likely. You really have that bad a feeling about their OKC-acquired players, even when their minutes are likely to go down in the playoffs?
Or perhaps you went 4-3 for strategic reasons bc others are more likely to pick 4-2 or 4-1 Celts so you stand out and win the thing more often by picking 4-3?
todd2,
IMO, Dallas is one of the deeper teams in the NBA. Despite having quite a few older players, they were actually pretty good on back to backs. They can rest key players when they know the schedule is about to get tough and get away with it better than most. That’s less of an issue/advantage in the playoffs.
It was nice seeing CP3 play yesterday. Aside from his stellar performance, he was the first one by Gray’s side when he went down. Classy.
@todd2 While it was a “classy” gesture to support a fallen teammate, I’m not sure Chris Paul is any more or less “classy” than the average player, especially considering he intentionally punched Julius Hodges with a closed fist directly in the groin/genitals during a game. He is a awesome point guard and shooter, though!
Point taken. I missed that one. I like the guy’s leadership.