Entire Rotation Returning = NCAA Berth?

Now that Noah Dickerson has announced that he’s returning to the UW basketball team we can officially say that the Huskies are returning the top-8 players from last season’s rotation. That means every player who played more than 4.5 minutes per game is coming back. That doesn’t happen very often.

It stands to reason that bringing back everyone who played a major role off a team that won 20+ games would lead to good things right? Probably. I wouldn’t be doing my civil duty though if I didn’t fact check such a claim. But how to figure that out?

Kenpom.com uses a metric called Minutes Continuity (hereafter MinCon) to determine how similar a team is to the following season’s team. The Huskies return 95.1% of their minutes played from last season but that might not necessarily be the best way to go about it if those players end up playing different roles. MinCon takes the minimum of the % of minutes played for each player between year 1 and year 2 and uses that total. Jaylen Nowell played 80% of total minutes last year but if he only plays 50% next year then he would only contribute 50% towards their MinCon.

This means that the value changes as the year goes on. The Huskies get credit for everyone returning as of now but the less those returners play relative to last year than the lower their MinCon. USC returned their entire rotation going into last year except then DeAnthony Melton was ruled ineligible right before the season started and Bennie Boatwright got hurt so their actual total dropped down to 78.9%.

To figure out who our case studies should be we’ll look at 3 scenarios. #1 is that none of the incoming freshmen are able to make a dent in the minutes. Elijah Hardy takes the backup minutes from MCIII but doesn’t crack the rotation, Jamal Bey occasionally gets a look in Carlos Johnson’s role, and neither of the project big men are ready to contribute. However, let’s say that Hameir Wright steals more playing time away from Sam Timmins and Jaylen Nowell gets more time at point guard which allows Nahziah Carter to effectively steal minutes from David Crisp. That would mean the Huskies have a MinCon of 94.3%. The highest total since he began measuring it in 2008 was 93.2% (2010 Butler) so this seems unlikely.

In scenario 2 we also see Eljiah Hardy become a real competitor to David Crisp and takes on an expansive role as a freshman. Crisp sees his playing time drop and maybe has a nagging injury that keeps him out for a week or two and so only plays 50% of the minutes (a very vocal part of the fanbase cheers). In this scenario they have a MinCon of 88.3%.

Scenario 3 is the same as 2 but instead, Sam Timmins gets hurt and/or gets completely passed on the depth chart by the freshmen bigs. That drops his minutes played to 0%. Also, Jamal Bey steals 5% of Dominic Green’s minutes and 5% of Matisse Thybulle’s minutes. That gives Washington a MinCon of 79.3%.

While we may not be able to directly predict the future, the odds are that some kind of injury or unforeseen circumstance will arise that makes the distribution closer to scenario #3. So we’ll focus in on teams with a MinCon of 80% or higher starting broad and eventually trying to find some similar case studies.

Across all of Division-1 there have been 88 teams with a MinCon of 80% or higher. That already puts 2019 Washington in rarefied air. With 350+ D-1 schools that means Washington will likely be in at least the 97th percentile in that category.

The graph above is cluttered but the big take away is that any team above (to the left of) the line improved their KenPom ranking from year 1 to year 2 while below (to the right of) the line means they got worse. The teams with labels are ones that made the NCAA tournament in year 2. Schools in red are traditional power conference schools, in green are occasionally multi-bid leagues, and light yellow are single bid leagues.

The average team across all of D-1 improved by about 3 wins and climbed over 30 spots in KenPom rankings the year that they had such great continuity. If you applied that to Washington from last season it would result in 24 wins and a KenPom rating of 67. That would put the Huskies right on the verge of an NCAA at-large berth. Last season the lowest KenPom ranking for a team that got an at-large spot was St. Bonaventure at #68.

But looking at the entirety of D-1 draws in some potential huge outliers. Furman in 2016 improved from 309th to 175th in KenPom and from 11 to 19 wins. That’s an improvement of 134 spots and 8 wins. Washington literally can’t move up that many spots. Let’s look at some more similar teams.

When you only look at multi-bid leagues the average win increase actually goes up a little while the KenPom improvement goes down. The latter isn’t a shock since the teams were much better in the first place and had less room for improvement. Only 18 squads met the criteria over the last decade so not even two per season. And only one of those has happened since 2014. It’s pretty close to unprecedented in the transfer-happy era of basketball in which we now live to have a team be that intact from year to year.

Unfortunately, there aren’t many great case studies for us to glean information from. Only 3 of the 18 teams started out worse than 70th in KenPom. Those were: 2007 Baylor, 2008 South Carolina, and 2013 Mississippi State. Mississippi State won just 10 games and was ranked 256th in the first year of that stretch which is not a good comparison. South Carolina and Baylor were much closer with KenPom rankings of 106 and 107 in year 1. The difference is that they won just 14 and 15 games respectively. For the record, Baylor improved by 6 wins and 66 spots to secure an 11 seed. South Carolina went up 7 wins and 35 spots for an NIT berth.

There’s not an example in the last decade of a team that over-achieved their KenPom ranking in terms of wins like UW did last year and then brought back almost their entire team. There are two principles duking it out here. One that says UW is likely to slide back after winning a lot of close games and losing a lot of blowouts last season. And another that says they should take a big leap forward with such a veteran core.

If you want to close your eyes and just look at the raw numbers then you could say that 14 of the 18 teams (77.8%) made the NCAA tournament returning so much of their roster. But then again, more than half of those teams were a top-50 team to start with and so just needed to hold steady for an NCAA berth.

Conclusion

The one thing that seems certain is that UW’s KenPom ranking will improve in 2018/19. The only team that got worse in that respect in this sample was 2009 Notre Dame which dropped from 24th to 41st and missed the NCAA tournament. Otherwise, power conference teams improved anywhere from 2 to 66 spots.

Although it pains me to say it, there’s only so much that the numbers can tell us. If you throw out that 98th ranked KenPom number from last year the optimism is bursting. The Huskies will be bringing back the Pac-12 defensive player of the year, the Pac-12’s leading returning rebounder, and the Pac-12’s 3rd and 5th leading returning scorers (Dickerson is the rebounder and 5th scorer). No one else in the conference comes close to returning that kind of production. Of course, last year showed that being near the top of the Pac is no guarantee of tournament success.

Yes, Washington should be better next year. But while it’s certainly reasonable to expect a tournament berth next season, it’s far from guaranteed.