Since Ken Pomeroy began tracking possessions per game (aka tempo, aka pace) on his website, fans have used his measure of tempo to argue about the entertainment value of their team. But the entertainment value of tempo has long been exaggerated. Even if many fans prefer up-and-down basketball, fans typically do not like poorly played basketball. And the fastest way to speed up the game is to allow wide open shots. As fans of Providence learned during the Keno Davis era, fast-paced basketball with no defensive resistance can be painful to watch.

That’s why Ken Pomeroy’s recent use of the play-by-play data is so fascinating. This summer he used the play-by-play data to develop a measure of average possession length (APL) on offense and defense. If teams are slow-paced because they play frustrating long possessions on defense, we probably should not criticize them for not running in transition. And now with APL, we do not have to.

The following table shows the coaches that have probably been receiving too much criticism for playing slow basketball. These are coaches whose rank nationally in terms of average possession length (APL) on offense differs substantially from their rank in terms of possessions per game. (In the table I calculate the mean APL and mean possessions per game for the last 4 years for each coach. I then rank all active coaches. There are 312 D1 head coaches with previous historical data, thus the rankings vary from 1 fastest to 312 slowest.)

There are a number of championship level coaches from Jim Boeheim to Tom Izzo to Billy Donovan to Rick Pitino whose teams are really only slow because of their defensive tenacity.

Meanwhile there are also a handful of coaches who probably deserve more criticism for being boring. Their teams have only been fast-paced because of giving up too many quick baskets defensively. Call this the group the Keno Davis all-stars. (Keno Davis himself falls from 11th to 47th between the two metrics.)

Many of these teams have struggled, so it is not a surprise that their defenses gave up easy baskets while their offenses struggled to score quickly.

But Frank Martin really stands out as an outlier on this list. These tempo calculations include his time at Kansas St. where his team won a lot of games. I think what is happening here is that Martin’s teams are extremely likely to crash the boards offensively. That sometimes means his teams don’t get back defensively which leads to more transition opportunities than a typical team allows. Frank Martin’s offenses aren’t all that fast-paced, but those additional transitional lay-ups do lead to more possessions per game.

True Fastest and Slowest over the Last Four Years

By the new metric, the fastest offensive coaches in the nation are:

Amusingly, two Nolan Richardson proteges (Mike Anderson and Corliss Williamson) show up in the Top 10.

And the slowest offensive coaches in the nation are:

Of course we still need to be careful in interpreting these new numbers. Part of the problem is that we associate faster tempo with fastbreak basketball. But at the college level with 35 second shot clocks, fastbreaks are only part of the equation. I think this post by Andy Glockner really sums things up nicely. In that column Glockner described the bi-modal nature of Wisconsin’s offense. Despite having unambiguously one of the slowest offenses in the country, Wisconsin isn’t afraid to run and take transition lay-ups when those opportunities present themselves. I would describe it this way. Transition baskets are usually the highest percentage shot any offense will get. And no smart basketball coach will pass up the rare chance for an easy bucket.

What really sets offensive tempos apart at the college level is what happens when transition baskets are not available. Do teams run some quick sets and attack before the defense is fully set (North Carolina’s secondary break)? Do they impatiently take the first semi-open shot that is available, even if that shot is a low-value two point jumper? Do they run traditional sets (like the clear out pick-and-roll) that take some time to execute, but not the full shot-clock? Or do they probe endlessly, running different cuts and sets until a lay-up or wide-open three presents itself? Those differences in styles explain why at the college level each offense’s average possession length can vary so dramatically.