How predictive are KenPom’s preseason rankings?

It’s that time of year again. The long awaited start to the college basketball season is only a few weeks away, kicking off Wednesday, November 25th. To further fuel the anticipation, Ken Pomeroy continued what has become an annual tradition, releasing his site’s preseason rankings on November 4th. Here are his top ten teams.

  1. Baylor
  2. Gonzaga
  3. Duke
  4. Villanova
  5. Kansas
  6. Texas Tech
  7. Wisconsin
  8. West Virginia
  9. Texas
  10. Ohio State

While fans will devour and debate these rankings over the next few weeks, they’ll likely be quickly forgotten once the games begin. Attention will instead understandably turn to how KenPom’s predictive rankings shift throughout the season. When it comes to evaluating teams and forecasting their performance moving forward, it won’t matter what their rankings were a month ago or at the beginning of the season; all that’ll matter is what their rankings are at that point in time, right?

While this is true, it’s worth exploring how similar KenPom’s preseason rankings have been to his final rankings. Plenty of attention has been give to the predictive value of the preseason AP Poll by myself and others (three examples: 1, 2, 3), but what about KenPom’s preseason rankings? A team’s rise or fall in the rankings from the beginning of the season gets mentioned from time to time, but there isn’t nearly the same level of analysis. One likely reason for this is that KenPom does not make his preseason rankings accessible through the site once the season begins. I had to rely on the Wayback Machine to gather the preseason rankings going back to 2011-12 (couldn’t find data any earlier than that).

So how similar have his preseason and final rankings been? Over the last nine seasons, 48/90 (53.3%) teams in KenPom’s preseason top ten finished in the top ten. Interestingly, exactly five of ten teams have done so each of the past five seasons. The number topped out at seven back in 2012-13 and has never been below five over this stretch.

Have preseason top five KenPom teams been significantly more likely to finish in the top ten? It depends how one defines “significantly,” but they have been more likely to do so. Since 2011-12, 28/45 (62.2%) top five KenPom teams have finished in the top ten of the rankings.

It’s worth noting that KenPom’s rankings are somewhat “sticky.” For example, a preseason No. 1 team that struggles at the beginning of the season might remain towards the top of the rankings longer than some believe is warranted. This actually happened as recently as last season. KenPom’s top preseason team in Michigan State got off to a 5-3 start, yet remained stubbornly high in the rankings. Still, while KenPom’s final rankings are inherently a bit biased towards their preseason counterparts, they aren’t nearly enough so to make this sort of analysis worthless. Plenty of teams significantly rise and fall in the rankings every season.

Are the KenPom or AP Poll preseason rankings more predictive of postseason success? I haven’t done the analysis yet to make this determination. What does seems reasonable to say, however, is that roughly five of KenPom’s preseason top ten teams will finish in the top ten once the season concludes. The five preseason top five teams in Baylor, Gonzaga, Duke, Villanova and Kansas should be in an even stronger position to do so.

How useful was this post?

Click on a star to rate it!

Average rating 0 / 5. Vote count: 0

No votes so far! Be the first to rate this post.

As you found this post useful...

Follow us on social media!