Commentary

Memphis, UCLA most consistent No. 1-seed candidates

Originally Published: February 7, 2008
By Andy Glockner | ESPN.com

In any knockout-style event, where one bad game can send you home, it's just as important to be consistent as it is to be very good.

In looking at this season's potential No. 1 seeds, it is interesting to look at their mid-level offensive and defensive performances to get a sense of how much room for error they have. Since overall stats are littered with games against overmatched opponents, the chart below details the figures for the seven contenders against RPI top-100 foes (through Tuesday's games):

The Contenders
  Kansas Memphis Duke Tenn. UNC UCLA G'Town
RPI 100 Games 10 9 12 14 15 11 7
Median Off Eff 106.5 108.7 110.3 112.6 113.5 118.0 104.6
Median Def Eff 92.6 90.6 94.6 101.9 86.0 101.7 107.0
Spread 13.9 18.1 15.8 10.8 27.5 16.3 -2.4
(Top 100 used instead of Top 50 because of greater sample size and at least one NCAA opponent will be outside the Top 50. Median efficiency used to find true mid-level performance. Efficiency numbers normalized for points per 100 possessions. The individual game efficiency ratings to calculate the medians and variances are from kenpom.com.)

On the surface, there are a few things that really stand out:
• North Carolina's margin for error appears to be much larger than its peers, and its much-maligned defense, on paper, appears to be much better than expected. (More on that in a minute.)
• UCLA's defense has been much less effective than expected.
• Georgetown actually has a negative spread, despite being 5-2 in its top-100 games.

The second component that helps establish potential vulnerability is variance. Higher variance equals greater inconsistency, and the less consistent you are, the more likely you are to have a performance that can get you beat. Here is each's team's variance on the offense and defense ends:

Levels Of Variance
  Kansas Memphis Duke Tenn. UNC UCLA G'Town
Variance Off Eff 16.0 11.8 10.7 8.1 10.6 11.9 6.8
Variance Def Eff 9.2 9.0 10.8 11.0 10.4 11.0 11.2
(Variance is the average deviation from a team's season-long average efficiency numbers. Efficiency numbers normalized for points per 100 possessions. The individual game efficiency ratings to calculate the medians and variances are from kenpom.com.)

Interestingly, Kansas' offense against better opponents has been wildly inconsistent, while Tennessee's run-and-gun approach has been surprisingly consistent.

The final piece of the puzzle is the level of competition. This chart details the average RPI of those top-100 opponents and the number of those teams that were top 50:

The Competition
  Kansas Memphis Duke Tenn. UNC UCLA G'Town
Average RPI of Top 100 55.5 41.9 48.3 42.6 62.5 41.5 29.7
No. of Top 50 opponents 4 6 6 9 6 5 7

Remember how good North Carolina's defense looked in the chart above? Maybe that's because nine of its 15 top-100 games are against teams in the 51-100 category. That implies the Heels likely will blow out their first-round opponent in the NCAAs. What it doesn't provide is rationalization for their mediocre overall defensive numbers. Simply put, North Carolina's D is a significant problem as a national title contender.

On the opposite end, all seven of Georgetown's top-100 games have been against top-50 foes. While it helps explain the spread that's so far below the peer group, Hoyas fans shouldn't be all that happy. These are mostly games against second-round and Sweet 16-level competition, which is basically where the Hoyas grade out at the moment.

Here's a more in-depth look at each team and what their numbers suggest:

Kansas (22-1, 7-1 Big 12)
The Jayhawks have been mauling inferior teams all season, but their performances against the better part of the schedule haven't been nearly as dominant. They beat Arizona by four in overtime at home, won by four at USC, ripped Oklahoma (in part because Blake Griffin went down injured) and lost at Kansas State. KU fans should be concerned about the offensive efficiency variance of 16.0. While that implies there are nights where KU explodes offensively, it also means the Jayhawks have the potential to have an off-night offensively, and that's when they have gotten beat in recent NCAAs. This is a really good KU team, but that's something to keep an eye on.

Memphis (22-0, 8-0 Conference USA)
The Tigers' numbers are very impressive across the board. Despite nights when their offense looks very hit-or-miss (and the free-throw shooting often does), their overall performance has been pretty consistent on both ends. But, there's no true road game in the six top-50 contests. There are three neutral-site games, though, where the Tigers won by 10 (over Oklahoma), 11 (over UConn) and four (over USC).

On offense, Memphis' 3-point shooting concerns are overblown. Despite the team's overall 3-point rate (213th in D-I), Chris Douglas-Roberts, Derrick Rose, Doneal Mack and Willie Kemp are all decent or better from the arc. The worries about free throws, though, aren't overstated. Memphis doesn't have a player with 20 attempts or more who's shooting better than 67.6 percent. That's a good way to lose to a comparable opponent in the Elite Eight or deeper.

Duke (20-1, 8-0 ACC)
The Blue Devils are second in the country behind Kansas in adjusted scoring margin (per ESPN's InsideRPI) and have played a much stronger overall schedule than the Jayhawks. They have been equally consistent on both ends, even though the offense is clearly the better part of this team. While they're a decent rebounding team, the one gigantic red flag is that the Blue Devils are only 113th in Division I in 2-point field goal defense. In four of their last six ACC games, the Devils have been suspect defensively, but their offense has been so good that the winning margins have been very convincing. It's not enough to pound Duke on the glass. Clemson dominated the boards and still lost by 13. North Carolina was plus-11 Wednesday night and lost by double digits. The key to beating Duke is also forcing the Devils into a bad shooting night. Easier said than done, but Pitt managed it and Temple came decently close, too, with that formula.

Tennessee (20-2, 7-1 SEC)
No real surprises in the numbers in the chart aside from the Vols' offense being much more consistent than you would imagine a high-octane, 3-point shooting team would be. Even with that, though, they are working with considerably less margin for error than their peers because of their defensive shortcomings. One bit of hope is that Chris Lofton's recent explosion continues. If the Vols' offensive efficiency can get even better, a la North Carolina's, they won't need to improve the defense to improve their chances of surviving an off-night. Like Duke and North Carolina, the imbalance is a concern. It's a primary reason why Kansas, Memphis and UCLA stand out as the national title favorites; they do it well on both ends, which is a much safer plan.

North Carolina (21-2, 6-2 ACC)
If the Tar Heels win the national title this season, they will have done it in an extremely unlikely way. UNC is currently fourth in adjusted scoring margin despite having a -0.7 defensive component, which means the Heels have been slightly worse defensively than the average opponent performs against its schedule. For all the talk about Tennessee's lack of defense, even the Vols are +2.7 in that category. Most nights, Carolina's devastating offense (fourth in the country in overall efficiency and sixth in tempo) is good enough to win, but in four of their past six ACC games, the Heels have been well below average defensively. This season, they're allowing 47.6 percent shooting from inside the arc. Statistically, this team is much closer to (but worse than) last season's team than the 2005 national champs, which was an elite defensive team, too. Getting the Raleigh/Charlotte path, and the huge home-court advantage that will come with it, may be a must for this team to get to San Antonio. And right now, that's Duke's path to lose.

UCLA (20-2, 8-1 Pac-10)
It was very surprising that the Bruins' median defensive performance against top-100 opponents was so suspect. Yes, there are reasons for a couple of them (the injury woes against Oregon and USC; Wazzu's late barrage of 3s that juiced up its numbers), but it still makes you take a second look at the them to see if they are legit on that end. Like Duke, the 2-point FG defense (79th in Division I) raises some eyebrows. But unlike the Devils, UCLA is a dominant team on the defensive glass and doesn't send opponents to the free-throw line. If you can eliminate cheap points on second chances and foul shots, giving up an occasional extra basket isn't that damaging. The bigger reason why UCLA is a leading contender, though, is the offense, which has gone from quite good to statistically dominant with Kevin Love in the lineup. Because the Bruins' tempo is so slow, you wouldn't realize it but they have utterly shredded five of their last six Pac-10 foes.

Georgetown (19-2, 9-1 Big East)
The Hoyas' numbers might have been the most shocking thing in this exercise, but in thinking about them, they make sense. Georgetown beat UConn on a last-second 3, beat West Virginia on a last-second block and beat Syracuse by two in OT. Toss in the Memphis loss, and you have a team that's 5-2 against the top 100 (actually the top 50) with a median performance on offense that's actually worse than their defense. The Hoyas will have more chances down the stretch to improve, but right now these numbers suggest that they are indeed a vulnerable team against comparable opposition. That shouldn't be totally surprising, since they were all but eliminated in the Sweet 16 last season by Vanderbilt. Right now, their numbers suggest that kind of ceiling this season, even though their overall field goal percentage differential is outstanding.

Based on all this, which teams look most likely to make title runs? Probably Memphis and UCLA. If those two do meet in the final, let's just hope it's not a repeat of their 50-45 grind in the 2006 Elite Eight.

Andy Glockner is a regular contributor to ESPN.com's college basketball coverage and is the host of the ESPNU College Basketball Insider podcast. He can be reached at bubblewatch@gmail.com.