What does Kenpom tell us about Past Champions?

With just a few weeks remaining in the regular season, it is time for the championship talk to start in college basketball. One of the best data analyses in college basketball comes from Kenpom, developed by Ken Pomeroy. Many analysts have questioned whether there truly is a dominant team in the nation this year, and Kenpom has some interesting insights.

First, let’s take a look at the history of champions and where they ended up ranked in terms of adjusted-offensive efficiency (an estimate of the offensive efficiency - points scored per 100 possessions - a team would have against the average D-I defense) and adjusted-defensive efficiency (an estimate of the defensive efficiency - points allowed per 100 possessions - a team would have against the average D-I offense).

When we pop those numbers into a graph (excluding UConn in 2014 to keep the data consistent) we get this.

Aside from the UConn anomaly in 2014, every champion finished the season ranked in the top 20 in adjusted-offensive efficiency and adjusted-defensive efficiency. Taking a look at this year’s rankings, there are currently only two teams that fall into this category: Purdue and Michigan St (Tennessee was in it yesterday but their 61-59 victory of Kentucky actually dropped them to 28th on the offensive side).

Of course these rankings are changing after every game (the MSU-Purdue match up on Saturday will definitely change some things), but seeing MSU there was certainly surprising. They suffered a 16 point loss away against Ohio State, went to OT against Rutgers (#153 on Kenpom), and had a 10 point loss home against Michigan. They recovered and are currently on a 7 game win streak, but have an average margin of victory of just 5 points in their last 4 games while the average Kenpom ranking of those 4 opponents was 75.

Meanwhile, Purdue has jumped on a 17 game win streak after suffering both of their losses this season in the Battle for Atlantis (which ‘Nova won). Their loss against Ohio State last night shifted their numbers a little, but they have to travel to Michigan St. on Saturday which could certainly change the rankings even more.

Now, where does Villanova stand in all this? They have one of the best scoring offenses in college basketball, averaging 87.9 points per game and owning the number one ranked adjusted-offensive efficiencies. Defense has been the biggest problem this season, and they are currently ranked #53 in adjusted-defensive efficiency. The team has suffered a rash of injuries which likely impacted these rankings, especially with a great defensive guard in Phil Booth and as of yesterday, Eric Paschall out with a concussion for at least a week. In the last Kenpom rankings before Booth’s injury, the ‘Cats were ranked #24 in AdjD. While their offensive ranking has remained at #1, the defense has certainly slipped a little.

Now, end of the season rankings certainly don’t reflect where the teams were at this point in the season in years past. Using the beauty of the WayBackMachine with the Internet Archive, I found Kenpom rankings for teams as close to February 7th as the Internet archives would allow (rankings started in 2007, and data was not included for 2008 because the page wasn’t archived at all in the first few months). Throwing those rankings into a graph, we get this.

As you can see, the data is certainly more varied. Four out of the ten teams didn’t have both a top-20 offense and defense at this time of the season. Looking at the differences between this time and the end of the season, the offense ranking moved an average of 4.5 spots while the defensive ranking moved an average of 10.2 spots.

Throwing some of the top teams of this season onto the trend and we get something like this.

As you can see, Purdue, Kansas, UNC, MSU, and WVU are closest to match this trendline. Of course there are only 10 teams on the original trend, so there is a great amount of variability.

By no means am I saying that you need a top-20 offense and defense to win a championship. UConn won it all in 2014 with the 39th ranked AdjO and 10th ranked AdjD. However, based on the past 16 years, no team has won the championship and finished the year with a AdjD rating outside the top-20. Looking at the data from early February in seasons past, the lowest ranked defensive-efficiency was the 2015 Duke Blue Devils with their 48th ranked AdjD (they went on the finish ranked #11 in that category).

So what to make of all this data? Good question. Villanova certainly isn’t statistically/historically out of the picture if they want to win the championship. With the return of Samuels in the coming days and Booth hopefully returning around the Big East Tournament, keep an eye on the Wildcats’ AdjD ranking and how it changes in the final weeks of the season.