Q & A with Ken Pomeroy

Sportando's Q & A with Ken Pomeroy, the founder of kenpom.com

Sportando: Can you explain how a meteorologist without a real basketball background has managed to become one of the symbols of the analytic movement?

Ken: It was a lucky accident. In 2004, when analytics were maturing in baseball, there was no work being done on college basketball. I started doing it in my spare time and then writing about it, and eventually coaches and people in the media started noticing. The internet can be used for some bad things, but one of the best things about it is that amateurs can mingle with professionals fairly easily, and that really helped me get into the business.

Sportando: how would you 'sell' the advanced stats to a person who hasn't never heard of it?

Ken: Advanced stats do a better job of accounting for context than traditional stats. For example, points per game on a team level tells you something about a team's offensive ability but it is also influenced by a team's pace. Points per possession separates a team's offensive ability from its pace. One of the goals of analytics is to make better use of raw statistical data, and improve our knowledge of what is happening on the court.

Sportando: In which aspect do you think the numbers are more useful: scouting, analysis or future prediction?

Ken: This is difficult to answer because data is useful in all of those areas, it's just a matter of having enough of it and using it properly. In the U.S., scouting high school prospects isn't helped too much by analytics because there isn't much data available, but that's changing. If you work hard enough to collect data, it can be useful in any field.

Sportando: Who are the NCAA coaches who immediately embraced the analytic movement? and in the NBA?

Ken: Guys like John Groce and Sean Miller were the first people I heard about that were using my site, but I'm sure there were coaches before them (and me) that were using their own form of analytics. In the NBA, the pattern has been for the front office to accept analytics more quickly than coaches, and since my focus is on the college game, I'm not aware of any NBA coaches that are particularly focused on advanced stats.

Sportando: There is a NCAA player on which you would have bet (based on its statistical profile) and instead has failed in the NBA?

Ken: Mike Sweetney from Georgetown is the one that almost every mathematical model missed on. He was a model of offensive efficiency that could block shots and rebound at an elite level. But he couldn't control his weight and never came close to reaching his potential as a professional.

Sportando: Have you already identified some big sleepers for next season?

Ken: I'm just now gathering the data necessary to run the preseason rankings, but the formula for finding sleepers is pretty simple: find a team that lost a lot of close games, didn't perform well in the postseason, and returns most of its roster. The problem is that even people not into analytics have cracked that code in recent seasons. So I'd say Utah and Arkansas are good candidates this season, but I'm not the first person to suggest those teams. Among more obscure teams, Georgia State is worth watching with Ryan Harrow and R.J. Hunter back from a team that went 25-9 but lost in its conference tournament.

Sportando: What is the achievement of which you are really proud of?

Ken: Just the general idea that a few people in the game take my work seriously. I would have never thought that possible when I was in college. I mean, I wasn't even considering working in basketball then. The fact that I get to play with basketball statistics for a living is pretty amazing.