With college football season in the past and the pros winding down, my inbox is starting to fill. Let's address a couple of the more pressing issues over the past week...
I was wondering if you could tell me what North Carolina's defensive ratings looked like going through the '05 season. Roy Williams commented in an interview today that the '05 team did not start playing defense until midway through the ACC season and I was wondering how this team compares to that team at similar points.
Interesting question, Brian. Being that 2005 was my first season tracking possession data in real time throughout the season, I remember it well. At this time three years ago, we knew that UNC was dominant, yet there were a handful of critics of their defense. In fact, it's my recollection that the concerns about their defense were bigger then than they are today. The only reason this is so clear to me because I recall that it was the best case of possession stats contradicting public perception that season. UNC's ultra-fast pace (there were just two teams that averaged more possessions per game) led many mainstream analysts down the wrong path.
At this time in 2005, UNC had the eighth-best adjusted defensive efficiency in the country. They would finish the season ranked sixth. The Heels defense was solid to this point; maybe it improved slightly as the season continued, but certainly not in a statistically significant way. That's not to say that this season's defense can't improve as much as it needs to in order to make Carolina's title hopes legitimate. In fact, they're just two good defensive games or so from putting their defense back in the acceptable range. It should be noted that "acceptable" used here is relative to a team trying to win a title.
That's why UNC's ranking hovering the 30s and 40s over the last two weeks is noteworthy. Since I've been tracking adjusted defensive efficiency over the last four seasons, the worst defense to make it to the Final Four was George Mason's in 2006. The Patriots' D entering that season's tournament ranked 23rd in the country. That's why I don't yet buy that UNC is part of a Fantastic Four that has separated itself from the rest of the country. They may get there, but until the defense improves, UNC is part of the rest.
I mentioned earlier that I thought there was more criticism of the Carolina D in 2005 than this season. I can never prove that this is true, but one reason it may be is because of the quality of the offenses UNC has faced.
Opponents' Adj. OE Rank
2005 104.7 37
2008 102.5 104
The Tar Heels have played a schedule that had featured some impressive defenses, but a bunch of ordinary or worse offenses (with the exception of Clemson). This will continue through early March because the ACC as a group is much better on the defensive end. There are currently just four teams in the conference with an adjusted offensive efficiency in the top 50 nationally. One is the Tar Heels themselves and one is Georgia Tech, who they don't play again.
I read your article with interest and am curious as to why you said that the Memphis non-conference schedule was not terribly impressive. As of this date it ranks 31st, using your statistics. Just three teams in your top 30 have better schedules, one of which (Arizona) the Tigers have already beaten. The other two (Gonzaga and Tennessee) are also on their schedule.
Thanks in advance for your time,
Ah, the old strength of schedule debate. It's almost like a dig on a team's schedule is like a dig on the team itself, but the two can be totally independent.
This is one of those age-old issues that has always been difficult to handle. There's a problem with describing schedule strength using one number in that some information gets left out. Take, for instance, the situation where one team plays the best and worst team in Division I, and another plays games against two teams whose ability is slightly above the Division I average. If you're a top-ten team, you're more likely to go 2-0 against the latter schedule. In an average SOS calculation, the latter schedule would also look more difficult.
This gets back to the issue of context. I'll take as much data as I can get my hands on to evaluate a team, but in the end, that data needs context. In the case of Memphis schedule, the decision to play Arizona as opposed to Wisconsin-Green Bay is worth about the same in terms of SOS impact as the decision to play Pepperdine instead of New Jersey Tech. Yet the latter decision (which is purely hypothetical, by the way) has no impact on how difficult Memphis' schedule really is. Then there's the additional information that Arizona's Jerryd Bayless was a game-time scratch against Memphis. This, of course, is not included in any SOS calculations. Thus, blindly using one SOS number--especially if it's the RPI which doesn't account for venue--can be misleading.
That's not to say Memphis' schedule has been weak. Not by any means. But it's lacked the marquee matchups against potential Final Four teams (with the possible exception of Georgetown) that I'd like to see to get a better idea of how good they are. That's not unique to Memphis. None of the top teams have had a game like that, which strikes me as unusual. Just think, last season we had UNC/Ohio State, Florida/Kansas, Ohio State/Florida and UCLA/Texas A&M. That's four matchups among teams that were thought to be Final Four teams. Memphis/Georgetown is the closest thing to a game like that this season.
Ken Pomeroy is an author of Basketball Prospectus.
You can contact Ken by clicking here or click here to see Ken's other articles.