2017 NCAA Tournament: Should Kansas Speed up Purdue?

I’ve heard more than one sports reporter suggest that given Purdue’s size advantage against Kansas, it would behoove the Jayhawks to get out and run, and try to wear down Isaac Haas and Caleb Swanigan. This makes sense on the surface, since KU’s advantage would seem to lie in the athleticism department. However, Purdue averages 69 possessions per game, which is in the top third of Division I, indicating they aren’t exactly scared of a more up-tempo game.

I could delve into my personal observations on pace and what I think Kansas should do, but as most of you know, that’s not exactly my style. My style is Excel charts. So, get ready for some scatter plots!

My initial thought was to compare margin of victory to possessions, and see if there is a correlation. I looked only at conference and postseason games, since those should provide a better picture of what Purdue is doing now, rather than including blowouts against mid majors in November. Since Purdue has had a few overtime contests, I measured possessions in possessions/minute, since the extra five minutes of overtime would skew the results if I simply went by number of possessions in each game. Here’s what I found:

Blah. Noise. Nothing to see here. With a correlation coefficient of just -.05, this suggests that maybe playing slower results in sliiiiightly better outcomes for the Boilermakers, but with a sample of 21 games, reading into a number that small would be pointless. How disappointing. The end.

Or is it?

Maybe pure margin of victory isn’t the number by which we should measure success. What if the distribution of tempo used by teams at the top and bottom of the Big 10 is playing with the numbers? Just to make sure, I decided to weigh the margin of victory against Ken Pomeroy’s expected margin of victory. Now, we’re measuring how Purdue plays vs expectation and comparing that to pace, instead of just how much they won or lost by. Does that change things?

That’s a little clearer. Honestly, the line would still probably be pretty straight, except we can see that in all five games where Purdue played games with at least 1.8 possessions per minute (72 possessions in a regulation length game), they’ve performed below KenPom expectations. 5 games worth of evidence is far from definitive, but since KenPom gives KU a 3 point advantage in this game, that would suggest it may be in KU’s interests to push the tempo a bit. Kansas averages 70.3 possessions per game, so that would be a slight increase in pace for them as well.

Obviously you have to take into account KU’s lack of depth when considering whether speeding up is a good option. How much 7’2 center Isaac Haas ends up playing is probably another factor, since word is Purdue doesn’t intend to have Swanigan defend Josh Jackson, which would suggest that they (strangely) wouldn’t intend to play their two biggest players together much. This by no means guarantees a better outcome if KU speeds things up, but there is some reason to believe it’s in their best interest.