Monday, March 28, 2011

Rose v. Howard: Why Do Chicago and Orlando Win?

Back we go to the MVP, I just can't resist. I have not favored Derrick Rose in the 2010-11 NBA MVP race, but what he did on Saturday night on the "road" in Milwaukee was something to behold.

Trailing 83-79 with five minutes to go, Chicago outscored the Bucks 16-4 the rest of the way, with Rose responsible for every single point, tallying 10 points and three assists at crunch time. Over the last three minutes, the Bulls scored 12 points on six possessions thanks to Rose's heroics.

It was certainly not the first time this season that Rose closed a game brilliantly in the fourth quarter, and it all made me go back to the drawing board and ask: man, am I really sure that Derrick Rose isn't the NBA MVP this season? I decided to dive back into the numbers and give D-Rose the fairest shake I possibly could. (After all, my mythical MVP vote is on the line!)

I decided to start with the Four Factors, the statistics which correlate most closely with winning, as delineated by analyst Dean Oliver (now with ESPN) in his book Basketball on Paper:
    - Shooting: Effective field-goal percentage (eFG%)
    - Rebounding: Offensive/defensive rebounding percentage
    - Turnovers: Turnover percentage (TOs per possession)
    - Free Throws: FT/FGA
In practicality, there are eight factors in which teams can be ranked - both their own performance in the Four Factors, and also their opponents' performance, as well.

How do the Chicago Bulls win games?
That's the question I was trying to get to the heart of. For many of you, this particular story is familiar: Chicago wins with defense. In terms of points per possession, the Bulls rank 12th in offense and 2nd in defense. Here's their profile in terms of league rankings for the Four Factors, for both own stats and opponents':

          OWN       OPP
- eFG%: 14 1
- TO: 15 12
- REB: 4 3
- FT/FGA: 10 (17) 12 (13)
I was unsurprised to see the Bulls rank best in the league in effective field-goal percentage on defense (note: eFG% is simply normal FG% with an extra .5 credited for every made three-pointer), but I guess I didn't fully grasp their rebounding excellence.

Not only do they rank high in both offensive and defensive rebounding, but combine those into overall rebounding percentage, and the Bulls are the best rebounding team in the league, by a fairly significant margin.

The Chicago Bulls win primarily because of defense and rebounding, it is clear. And their defensive prowess is based in forcing opponents to shoot poorly from the floor, rather than creating turnovers. Generally speaking, Chicago's strengths as a team do not correlate with the individual strengths of Derrick Rose.

[Note: Many others have noted that Chicago's bench has been an underrated reason for its success, as well - it's quietly been one of the best benches in the league, excelling especially on the defensive end, anchored by tough defensive bigs Taj Gibson, Kurt Thomas and Omer Asik.]

OK, so the evidence from the Four Factors wasn't in D-Rose's favor. But the Bulls have been on a hellacious 15-2 run to take over the top seed in the East since dropping the first game after the break to Toronto. I figured it would be fair to Rose to check on his stats during this stretch, to examine how he's driven Chicago to the top. It's here that I was fairly stunned. Here are Rose's pre- and post-All Star break splits:
       FG      FG%   3P     3P%   FT     FT%   PTS  REB  AST
Pre 9.1-20.2 .450 1.5-4.3 .355 5.2-6.2 .838 24.9 4.4 8.2
Post 8.2-20.2 .404 1.7-6.1 .284 7.0-7.9 .887 25.1 3.7 7.1
Wait, what? That's your MVP push? 40% field-goal percentage? Baron Davis-quality gunning from behind the arc?

I do credit Rose for continuing to get to the line more often, as it had previously been a key deficiency, but his true shooting percentage (TS%) has still been below league average (.529) since the break (Rose is at the league average of .540 for the whole season).

The three-point shooting is truly stunning. Another oft-cited weakness of Rose's game prior to the season, he reportedly worked on his long-distance shooting throughout the offseason, and such evidence can be seen in his pre-All-Star break numbers.

What's shocking about Rose's outside-shooting numbers is not that his percentage has regressed back near that of his first two seasons (.242 overall), but that he's still jacking them up at such a high volume (his 109 threes since the break is almost as many as he attempted in his first two *seasons* combined - 132).

Derrick Rose is a great player who has had an outstanding season. He is the best player on the best team in the Eastern Conference. He has been the driving offensive force in pulling out wins in the fourth quarter on several occasions. The Bulls offense as a whole has been excellent when he's been on the floor.

But Derrick Rose has been one of several key factors - defense (and the coaching behind it), Rose, rebounding, bench - in the Bulls' surprising success this season. He has not been the single primary factor.

One other interesting factor I discovered while exploring the Bulls is the team's bizarre injury profile. A significant factor behind the Rose-for-MVP narrative is that he has kept the Bulls afloat despite substantial injuries to the team's second- and third-best players (Carlos Boozer, 23 missed games; Joakim Noah, 31 missed games).

Yet, what I found remarkable is that the Bulls have otherwise had almost pristine health. Out of the other nine players who've seen significant time this season, there have been three games lost to injury, total (Rose, 1; Gibson, 2). That's a shockingly low number. I think the perception is that Chicago has been hit hard by injuries. In reality, I'd guess the loss of two stars for significant time balanced by almost no other missed games leaves the Bulls with a fairly average injury profile overall.

How do the Orlando Magic win games?
My mythical vote is not necessarily a Rose vs. Howard proposition at the moment, but I wanted to check out how the Four Factors argument worked out for another top candidate, to see if my reasoning above was unduly or unfairly harsh towards Rose. So, I took a look at the Orlando Magic and Dwight Howard, who rank 14th in offense and 3rd in defense overall:

          OWN       OPP
- eFG%: 5 4
- TO: 25 15
- REB: 17 1
- FT/FGA: 5 (15) 15 (8)
The Magic are strong in effective field-goal percentage at both ends of the floor. Dwight Howard's .601 eFG% is second in the league. He's also the undisputed force behind the team's field-goal defense, as a multiple Defensive Player of the Year winner without much, if any, imposing defensive talent surrounding him.

Orlando's also exceptional at defensive rebounding. Howard is second in the league in defensive rebounds per game (10.2) and third in defensive rebounding percentage (30.3).

The FT/FGA factor is an interesting one, as there is some question as to whether that factor is better calculated as FTM/FGA or FTA/FGA. There is secondary value in simply drawing a large number of fouls, mainly to get the opposition into foul trouble. For both teams, FTA/FGA and FTM/FGA (in parens) are listed.

The different measurements change Orlando's ranking significantly, mainly because Dwight Howard draws a lot of fouls, but also misses a lot of foul shots. Howard shoots 11.8 free throws per game, leading the league by a huge margin of more than 3 FTA per game. I believe that drawing fouls is a major positive factor, and thus favor FTA/FGA as a calculation, though I can accept that you may prefer FTM/FGA, which changes the reading of how vital Howard is in this category.

In any event, I strongly believe that FTA correlates more closely with winning than FT%. I'd much rather have Howard shooting 12 FTA per game, and making only 59%, than having Stephen Curry shooting 3 FTA per game, but making 93%.

It's also worth noting that the Magic are particularly bad in committing turnovers. Dwight Howard ranks slightly above average in turnover rate for a center. The real culprits for Orlando in this category are point guards Jameer Nelson, Gilbert Arenas and Chris Duhon.

All of this is to say that the success of the Orlando Magic correlates almost directly with the strengths of Dwight Howard. The Magic win games primarily because of Dwight Howard, period. Secondarily, they make a lot of three-pointers on average, and have an excellent coach. But Dwight is the driving force.

Of course, things get fairly reductive at some point, and people view both Chicago and Orlando as teams with one superstar, and thus point to W-L records - Chicago's 53-19 is 6.5 games better than Orlando's 47-26 - as evidence that Rose is better than Howard.

As described above, I believe that this kind of reasoning severely underrates that Dwight Howard is much more responsible for Orlando's wins than Derrick Rose is for Chicago's. I'd also say that, even with Boozer and Noah missing a combined 54 games, I'd still much rather have the talent surrounding Rose than the collection of pretty good and/or aging players around Howard.

I was fully ready to swing to the Rose MVP camp following his performance late in the game on Saturday. However, after digging into the evidence, I feel more strongly than ever that Derrick Rose does not deserve to be the 2010-11 NBA MVP, though I have little doubt at this point that he will win the award. I'm going to withhold judgment on my mythical ballot as a whole for another couple weeks.

Sunday, March 06, 2011

Top Five Teams of The Future, on SI's The Point Forward

Zach Lowe, having a Blake Griffin-like rookie season with the new NBA blog The Point Forward on, was kind enough to invite us to contribute a post while he was attending the Sloan Sports Analytics Conference in Boston on Friday.

We wrote a piece titled Top Five Teams of The Future, taking a long view in the aftermath of the wild 2011 trade deadline to try to determine which teams are best-positioned to be elite for the next five years or so, following the moves and non-moves this season.

Hope you enjoy, thanks to all our readers, and welcome to any SI readers making their way to this space for the first time.