clock menu more-arrow no yes mobile

Filed under:

Predicting a Victor of the Big Dance Via KenPom

Over the last 14 years, the KenPom rating system has shown a fairly obvious pattern of who will win the national championship in men's college basketball. With the Sweet Sixteen upon us, who is in the best position to bring home the hardware this year according to Mr. Kenneth Pomeroy?

Bob Donnan-USA TODAY Sports

The madness of March has already wrecked it's fair share of brackets after just the first weekend, just like it always seems to do. You had Michigan State winning it all? Well, you're an idiot! Why would you pick MIchigan State to win it all you rube! (It's ok, I thought Izzo + favorable draw + one of the best players in the country would also = national championship... oh how wrong we were) But there is still two-thirds of the tournament from a round stand point and plenty of dramatic basketball left to be played. Heck, March has become so mad that many a website now do a "round-by-round" picking system to hold people's interest in playing various contests throughout the whole tournament. Which leads me to the point of this particular piece:

Who the heck is going to win it all this year?

With 16 teams remaining, ranging from 11 seeds to 1 seeds, blueblood basketball programs to programs that are traditionally more concerned about the gridiron, and every team in between, we have a wide variety of candidates to choose from, even though our field has been whittled down threefold since the beginning of the tournament. Since picking a successful bracket appears to be beyond my areas of expertise (please see This One's for Tubby!!!!! entry in the Yahoo TDG bracket pool for the evidence), let us look to a system far more intelligent than any one person could be: the KenPom efficiency rating system.

The statistics involved in the KenPom rating are often cited and quoted at this site by several of our basketball writers. Sometimes, they are met with skepticism. But looking all the way back to 2002, when Pomeroy starting tracking the statistics, and up until the present, some pretty obvious patterns reveal themselves. Below is a chart to display these particular phenomena.

National Champions

There are a few obvious conclusions to be made about national champions of the last 14 years. Here are a few I think are most relevant:

  • Teams that win it all are rated within the top 10 overall according to KenPom. Strangely, Connecticut has twice been a bit of an "outlier," winning it all while being rated 9th (2011) and 8th (2014). Remember they were seeded way below KenPom's rating of them in 2011 and 2014, 3 seed and 7 seed, respectively.
  • A champion has never been worse than 39th in offensive rating and even that is an outlier. Take out the Miracle Huskies and that number drops to 18th in offensive rating.
  • A team has never been worse than 21st in defensive rating.
  • The highest rated team (according to KenPom) has won the championship 7 of the 14 years of it's existence. That is to say, if one solely used KenPom to determine the national champion, they would have been right 50% of the time. I should probably try this...
  • Stating the fairly obvious here but... teams that are elite in a per-possession basis in both offense and defense find themselves bringing home hardware. You don't necessarily have to be the best at either one but your combined rating must be top tier if you expect to hang a banner.
So where do these findings leave us? In order to pick a probable champion, let's examine the teams remaining who fit the following criteria derived from the chart above:
  1. The team must rank in the top 18 in offensive rating (we're ignoring the UConn outliers for sake of simplification).
  2. The team must rank in the top 17 in defensive rating.
  3. The team must rank in the top 9 in overall rating.
Following these criteria, we are left with the following teams, as illustrated in the chart below:

Possible NCAA Champions

Now, some of you will look at this, roll your eyes, and say, "Thanks Captain Obvious!" due to the fact that the chart is composed of three #1 seeds and one #2 seed. But what may be more useful is to look at some of the teams this process didn't identify as possible champions. One example is the missing #1 Oregon whose defense is rated too poorly at 43rd to gain inclusion on the chart above. The one team that may have the biggest argument for inclusion is Oklahoma, who ranks 7th overall, 14th in offensive rating, and 18th in defensive rating. They just barely fall out of our "non-UConn-outlier" threshold on defense but are probably good enough to win it all, considering they have Buddy Hield.

Sweet Sixteen Notes

Getting to the Sweet Sixteen is quite an accomplishment for just about any basketball program in any given year. Heck, Minnesota has only done it four times in its entire history. So calling out a team who makes the Sweet Sixteen might be a little harsh but I just want to point out two of the squads that may have been fortuitous in their journey to the second weekend of March Madness, along with a few other interesting notes:
  • Duke and Notre Dame, while contributing to the positive vibes the ACC has been feeling all week due to having 6 participants in the Sweet Sixteen, are defensively atrocious, ranking 109th and 172nd in defensive rating, respectively. Notre Dame is "lucky" (sorry I had to) to be in the third round of the tournament, considering they got a tip in against a superior Stephen F. Austin team (according to KenPom, 26th overall, 50 off., 20 def)  in the waning moments of their second round matchup. Look for both of these teams to lose their games (Yes, I understand that means Wisconsin advances but I'm just stating emotionless facts here).
  • Of the "non-contender" teams remaining, the most balanced is the Bulldogs of Gonzaga. They sport an offensive rating ranked 23rd and a defensive rating ranked 22nd. With this kind of balance and Damontas Sabonis playing his best ball of the year, look for them to beat Syracuse and clash with top-rated Virginia in the Elite Eight.
  • A way one could have sniffed out the Michigan State upset would have been to look at their defensive rating, which was 51st in the country. They had the #1 rated offense, and they had no problem scoring points against Middle Tennessee State, but they couldn't stop the Blue Raiders at any point during that first round matchup. KenPom may have hinted towards that result.
In summary, it would appear that 3 of the 4 top seeds are favorites by popular opinion, the NCAA Committee, and KenPom's rating system. Personally, I would be surprised if a team besides Virginia, Kansas, North Carolina, and Villanova won it all but then again, I picked Michigan State at the beginning of the tournament so what the heck do I know?

Enjoy your second weekend of March Madness everyone! Go Irish! Go Ducks! Go Bulldogs!