Monday, February 22, 2010

Duke Weaknesses (Perceived and Real)

Clearly, this year's Duke team is neither entitled nor even favored to win the National Championship. The widely respected human opinion polls put, at time of writing, half a dozen teams above a Duke team that has failed to receive a single first-place vote this entire season. Clearly, something scatters the blue sunshine. There are many common memes surrounding Duke, though only a few of them actually apply to this years' team. Let's take a look at some of the common criticisms levied against this year's Duke squad, analyzed from a tempo-free perspective (statistics, as always, from Ken Pomeroy).

1) Duke has played a soft schedule. If we go based on the tempo-free stats, nothing could be further from the truth. Pomeroy ranks Duke's schedule second in the nation, behind West Virginia. Sure, their non-conference schedule is ranked 83rd, but take a look at the teams on top of the non-conference SOS rankings: other than Butler, none of the top 20 have a prayer at an at-large tournament bid. And a number of the teams up top are probably there because they played Duke... this criticism is tired and should be sent to the land of misfit memes.

Making the argument worse is the observation that not only does Duke have the #3 offense in raw efficiency (1.17 points/possession), they have accomplished such lofty efficiency against a harder slate of defenses than anyone in the nation (adjusted efficiency of defenses Duke has played: 0.95 points/possession).

Taken more subjectively, Duke will play 15 non-conference games this regular season (counting Thursday's game against Tulsa). Of these, as many as 7 teams could make the NCAA tournament; what is a better preparation for NCAA tournament play than contests against teams that will actually be there? For comparison, Duke is #2 in the Sagarin strength of schedule as well, and played the fifth-toughest slate in this version of RPI.

2) Lack of backcourt depth. It seems only yesterday that the meme was about the lack of frontcourt depth (well, last year anyway). Assuming that Kyle Singler counts as a guard, Duke's main men play a ton of minutes, while backup shooting guard Andre Dawkins has struggled in limited minutes. I have a problem with this criticism on a number of levels. Part of this argument comes from the same area as the Duke Fade, which is that too many minutes are given to the Big Three. Here's the tempo-free stats.


First of all, pretty cool that all three are used in the same percentage of possessions. Second, if you have three players capable of contributing those massive tempo-free player ratings, why would it be a bad thing to play them all the time? If a player was used so much that they tired in-game, his offensive rating would plummet, due to the extended number of possessions played. Duke's top three men (by % minutes played) have a higher combined rating (362.5) than the top three men on Kansas (358.5), Syracuse (333.4), Purdue (343.1), West Virginia (354.4) and Kentucky (334.5), to pick a few. Scheyer/Singler/Smith are doing it in far more minutes than any of those teams, save perhaps WV, who do get 80%+ minutes from Butler and Jones.

For another avenue of argument, I shall paraphrase what I overheard this morning on 99.9 The Fan in Raleigh: "While Duke has the best trio of players in the country, what happens when they are all shut down?" This, of course, has an easy answer: Duke would lose. The better question is: "what's the likelihood of all three being shut down?" The three Duke stars average a combined 53.8 points/game; the fewest they've scored, combined, was 41 in the Preseason NIT championship win over UConn. In Duke's four losses, the trio has averaged about the same amount, and scored a combined 61 points in the loss at Georgetown. Indeed, it seems that Duke's wins will come down to how many points they can drag out of their frontcourt. Which leads to...

3) Duke has no talent in the frontcourt. I should hope that the last three games would silent anyone who would make such a bold claim. There is clearly some talent in the frontcourt, and that is in the (massive) form of Brian Zoubek. The man who makes girls swoon so hard that their marriage proposal signs are upside down has averaged nearly a double-double (9.7 points, 12.7 rebounds) in those games. With his new-found ability to avoid foul trouble, Zoubek has slipped back above the 40% minutes played mark. This finds him back atop the nation in offensive rebounding. Let's nail that home for the readers too lazy to click on the link:

Category: Percentage of a team's missed shots rebounded by a player while he's on the court.
Qualification: 40% of a team's minutes played (that's about 16 min/game, ignoring OTs)

1. Brian Zoubek, Duke 23.0%
2. Demarcus Cousins, Kentucky, 22.8%
3. Anthony Johnson, Fairfield, 17.4%

So, other than the stud freshman at Kentucky, no one is even close to Brian Zoubek in offensive rebounding ability. Against Maryland, Zoubek used missed shots by the Big Three to have an amazing offensive night, scoring 16 points with 17 rebounds (8 OR). Then, against Virginia Tech, Zoubek benefited from Duke's cold shooting for another 16 rebounds (8 OR). This time he tended to kick the ball out to open men, including two assists on Nolan Smith three-pointers.
The tempo-free world is only six years old, but in that time, only three players have had an OR% above 21%: DeJuan Blair was at 23.6% last year. (The next highest season total is 20.8%).

Of course, there's more to the frontcourt than Zoubek, and the tempo-free stats have mixed feelings about Duke's other big men. Miles Plumlee is a very strong defensive rebounder, and at 22.6% is currently vying for Duke's best defensive rebounding mark since Pomeroy started tracking individual stats in 2005 (Shelden Williams' best was 22.0%). Mason Plumlee currently has the lowest offensive rating on the team (94.5), thanks to a triple-whammy of low effective field goal percentage, small rebounding numbers, and a very high turnover rate. While analysts have been high on Mason's potential, the tempo-free stats haven't shown that this season.

Lance Thomas is generally seen as the "glue" of the 2009-2010 Duke team, and the captain is certainly one of the emotional leaders. As with many in this position, his contributions are said to be "intangible." Indeed, tempo-free statistics don't have much to say about LT. He has an average offensive rating, and the only stat which really stands out is his turnover rate, which is the highest on the team: Duke turns the ball over on 26.1% of his possessions. This could be partially due to Lance, who turns the ball over 1.6 times/game; it could also be unlucky that he's on the court when turnovers occur. Don't view this as a total knock on Thomas, however. These stats of course only include offense, and while Thomas' contribution may not be exactly "intangible" it is certainly "unmeasured" by this system.

Overall, this criticism is still hanging on by a few threads. Clearly, Duke's frontcourt is responsible for a massive number of rebounds, so much so that it can prop up a poor shooting night against many teams. Whether one of the Plumlees or Zoubek can be a viable offensive threat remains to be seen. For now, only Zoubek is impressive from the tempo-free perspective.

For the heck of it, let's throw in:

4) Duke gets all the calls. While pretty hilarious in most seasons, this conspiracy theory has a pretty steep piece of evidence to overcome this year: this season, Duke has a free throw rate (FTA/FGA) of 37.8, which is 171st in the nation, good for almost exactly median. In addition, Duke gets only 21.9% of their points from the charity stripe, good for 119th in the nation. This season, the refs seem to have been paid off by the fans of Kansas State (Free Throw Rate: 53.3) and Wyoming (27% of points from FTs). What Duke does have going for them, however, is free throw percentage: 8th in the nation at 75.9%.

So, what are Duke's real weaknesses, as shown by tempo-free stats?

1) Fouls. As alluded above, Duke struggles more than usual to get to the free throw line, and so cannot benefit as much from their high percentages from the line. While in most years, Duke has multiple players drawing more than 5 fouls per 40 minutes, only Scheyer is above that mark, and no one else is very close.

On the other side of the basketball, committing fouls has been a huge problem, particularly for Brian Zoubek (8.2 Fouls Committed/40 minutes) and Mason Plumlee (7.2 FC/40). In all three of Duke's January losses, they allowed their opponents to get to the free throw line at a rate of about 50 FTA/FGA.

2) Slow Pace On each team's "Game Plan" page, Pomeroy has tables of correlation coefficients. Each of these is a measure of how correlated a particular statistic is with offensive/defensive efficiency. At the moment for Duke, one of these, on offense, is the "pace" of the game, measured by the number of possessions. As Duke's number of possessions goes up, so does their offensive efficiency. For example, Duke's fastest game this season was against Pennsylvania, in which Duke scored 115 points in 75 possessions (Offensive Efficiency: 151.8!) . In Duke's slowest game, at Clemson, the Blue Devils scored just 60 points in 61 possessions, for an efficiency of 98.8. This will definitely be a trend to keep an eye on, should Duke face a team notorious for slow play in March (looking at you, Big Ten).

3) Allowing 2-pointers Duke's defense is ranked 18th in efficiency, but it has it's definite weaknesses. While the Blue Devils are second in the nation in 3-point shooting percentage against (27.4%), they allow opponents to shoot over 45% inside the arc (81st). Put another way, Duke's opponents score 60.8% of their points from 2-point range, which is the 8th most in the nation (to be fair, the 18.4% of opponent's points from beyond the arc is fourth lowest). Duke's strategy clearly makes some sense: it doesn't take a math degree to deduce that 3 > 2. However, some teams are just not built for chucking up treys. This season, Georgetown, NC State, and Wisconsin all used Duke's spread defense to their advantage, scoring backdoor layups and short-range jumpshots.

As the rankings sit today, Duke is the best team in the nation, measured from a tempo-free perspective. They have the best offense, a good defense, and have played a very strong schedule. We've seen that many of the common criticisms of Duke fall apart when seen from the perspective of tempo-free stats. Still, this is a college basketball world without perfection, and while Pomeroy's statistics can help us see that Duke is a little underrated by the human polls, there is still room for improvement as March looms.

Wednesday, February 17, 2010

The Duke Fade: A Response

Earlier today, Alex Fanaroff of Duke's daily newspaper The Chronicle posted a wonderful article, "Duke Does Decline, Objectively Speaking." It's a refreshing look at basketball in a way that we love here at the Immaculate Inning: using tempo-free statistics. Fanaroff describes the general issue:

1) Duke fades down the stretch.
2) This is due to the starters playing way too many minutes, causing them to get tired.

Fanaroff went into the article expecting to at least debunk #1 as a myth, but ended up finding a strong correlation between Duke's efficiency margin (Offensive Efficiency - Defensive Efficiency), and the number of ACC games played. This suggested that over the last seven years, the data indeed does suggest a Duke Fade. Fanaroff was unable, however, to show that "minutes by starters" had any effect on efficiency margin. Naturally, I was intrigued, and wanted to dial further into the data. Here are a number of issues:

1) Year effects do matter. In his accompanying blog post, Faranoff says "Predictably, most seasons failed to produce robust trends due to the limited sample size, though all seasons from 2004-2009 demonstrated a downward trend." This statement is misleading on two accounts. First of all, here is each year plotted individually, with the linear regression line on each:

True, there is a downward trend most years, and Faranoff does mention that Duke's trend is technically upwards this season. Most seasons, that line is not significant: there is not enough evidence to say that the regression line differs from no slope.

However, the null hypothesis can be rejected in 2004 (r-squared 0.2177, p = 0.04406) and 2008 (r-squared 0.2993, p = 0.01879). Here's my question: if a declining efficiency margin during ACC play is a bad thing, then why did the 2004 team make the Final Four, despite having one of of the few significant in-year declines?

2) Opponents matter. As many folks over at the fine institution of DBR have pointed out, Duke always plays Carolina as its last ACC regular season game. Then, while the early rounds of the ACC tourney may provide a brief dip in competition, Duke almost always advances to the end of the tournament, and they naturally find better teams there. We can look at the effect of Opponent Rank (determined by Pomeroy) on Efficiency:

The scatterplot doesn't look like much, but the trend line is there, and it is significant: Duke's opponents get tougher as the season goes on. And Duke does play much better against inferior competition:

Since Duke's opponents are harder later in the season, and Duke has lower efficiency margins against better opponents, how can we take this into effect? One way is with an Analysis of Variance (ANOVA: more info here), in which we can specify variance components. Basically, we have two main effects on efficiency margin: ACC Game # and Opponent Rank. If we account for the variance accounted for by the strength of the opponent, is the correlation still significant? For the stat-geeks out there, here's the ANOVA table:

Anova Table (Type III tests)

Response: Delta.100
......Sum Sq....... Df........ F value........ Pr(>F)
(Intercept) 0.12776 ....1 ........6.0593 ..........0.015267 *
Opp.Rank 0.48675 ..1...... 23.0858 .........4.563e-06 ***
Game 0.19808 .....1 ........9.3947 ...........0.002693 **
Residuals 2.50903 ....119

To summarize, there is still an overall significant downward trend in efficiency margin as the season progresses, although the significance is reduced.

3) Home Court Matters? When Ken Pomeroy adjusts for home court, he adds 1.4% to the home team's offensive efficiency and visiting team's defensive efficiency, and subtracts the same from the home team's DE and visiting team's OE. So the difference between Faranoff's raw data and what Pomeroy would consider "adjusted" is -2.8 for Duke's home games and +2.8 for Duke's away games. What happens to the correlation if we make the "Pomeroy Adjustment"?

The correlation coefficient decreases, as does the significance, but not to the point where the slope becomes non-significant. There is still a Duke Fade when we account for home and away games.

I must say I come away unimpressed with other explanations for the Duke Fade. Opponents do get tougher but not enough to overcome the efficiency drop. Adjusting for home and away games also doesn't have much of an effect. It is very important to note that nowhere have I suggested a causal agent for the Duke Fade. This is to avoid the common fallacy that correlation implies causation. Clearly, Duke can still have an historic season (2004) despite having one of the few significant in-year Duke Fades.

Instead, I'll take an "I Report, You Decide" kind of approach here. These are the statistical facts, and I'll be happy to attempt more rigorous investigations if they are suggested in the comments.