UK’s curiously linear relationship between returning experience and performance


Reid Travis is older and more experienced than your usual Calipari-era Kentucky player. That may turn out to matter, even if he’s not, strictly speaking, “returning.”

Duke may very well beat Kentucky at the Champions Classic this week, and, if you didn’t think so before, consider the fact that no fewer than 28 out of 30 writers surveyed (including yours truly) think the Blue Devils are going to lose. That kind of consensus fairly begs to be wrong. To paraphrase Groucho Marx, I would never want to be a member of a club voicing 93 percent agreement on what’s probably more of a 53 percent chance kind of thing.

This is where I additionally point out that a Duke “upset” win (ha) would indeed hold a measurable if still modest amount of predictive heft with regards to the rest of the season. Say “it’s only November” all you want.

The truth is they’ve been playing this four-team Champions thing now for seven years. The 14 winners have, on average, gone on to be seeded almost two full lines higher in the ensuing NCAA tournament than have the 14 losers. In fact, prepare to be shocked. One Champions loser actually missed the tournament entirely.

That team was, of course, Kentucky in 2013. The February injury suffered by Nerlens Noel did those Wildcats no favors, surely. Then again, that team (which, to be clear, was a very good but hardly legendary 17-6 even with a healthy Noel) stood out for another reason. It was the youngest roster John Calipari’s ever had in Lexington.

Returning experience and performance

Calipari-era UK rosters ranked on returning experience
%RPMs: % of returning possession-minutes from previous year
           %RPMs    W-L    NCAA seed    NCAA wins
1.  2015     60    38-1        1            4
2.  2010     55    35-3        1            3 
3.  2012     52    38-2        1            6
4.  2019     32
5.  2014     29   29-11        8            5
6.  2017     26    32-6        2            3
7.  2016     16    27-9        4            1
8.  2011     12    29-9        4            4
9.  2018      7   26-11        5            2
10. 2013      6   21-12       N/A          N/A        

So, sure, even making allowances for the unpredictability of a single-elimination tournament, there appears to be a trend at work. When Kentucky’s old, it gets a No. 1 seed. When it’s not, it doesn’t.

On paper this year’s roster looks functionally identical, in terms of returning experience, to the 2014 team that was labeled a huge disappointment (there were “40-0” t-shirts in the preseason) and that limped into the tournament as a No. 8 seed before mounting a thrilling run to the national title game.

Yes, but that 2014 team didn’t have a major-conference transfer with 82 career starts already under his belt. Reid Travis will turn 23 in 20 days. He’s older than D’Angelo Russell, Jahlil Okafor, or Devin Booker. No, Travis doesn’t leap tall buildings, Stanford opponents last season did not necessarily tremble much less surrender when he took the floor, and, sure, the Twitter hyperventilating when he announced he was transferring to Kentucky was a bit excessive.

Still, the material point is merely that a Wildcat roster will, for a change, have access to a top-30 recruit who’s already recorded four (albeit somewhat injury-blighted) seasons of development and growth. Four is more than the zero that UK usually makes do with before its players move on to the NBA.

It would be double-counting to simply take the minutes and possessions Travis logged in Palo Alto last season, add that to what we saw in 2017-18 from PJ Washington, Quade Green, and Nick Richards, and list the resulting number (44 percent) as this Kentucky roster’s quote-unquote true figure for RPMs. Nevertheless, we should remain open to the possibility that these guys will play older than a team returning 32 percent of its possession-minutes. Throw in talented freshmen like Keldon Johnson, EJ Montgomery, and Tyler Herro, and Calipari would seem to have some talent at his disposal, even by Kentucky standards.

A while back I came up with the idea of a “category 5 roster,” one that returns at least 40 percent of its possession-minutes while adding a freshman class that clocks in at 25 recruiting points or higher according to Drew Cannon’s canonical front-loaded recruiting curve. That idea worked well enough until Duke came along in 2017 and rather callously flung my cherished analytic bauble to the floor and sent debris flying in every direction.

A hobbled Harry Giles did not have the impact that a player variously listed as No. 2 or even No. 1 in the nation customarily exerts. That and a number of other factors made the Blue Devils the first category 5 roster of the one-and-done era not to earn a No. 1 seed.

Category 5 rosters, 2007-19

                      Recruiting  NCAA   NCAA
          Year  %RPMs   points    seed   wins
UNC       2007    79     35.0       1      3
Kentucky  2010    55     30.1       1      3
Kentucky  2012    52     33.1       1      6
Kentucky  2015    60     26.8       1      4
Duke      2015    42     29.7       1      6
Duke      2017    56     36.6       2      1    

For the record, UK and Duke look like this for the upcoming season….

          Year  %RPMs   points 
Kentucky  2019    32     29.7
Duke      2019    14     39.0   

No category 5s here, but plenty to watch with interest. Duke is indeed just as extreme in both youth and in freshman talent as everyone is saying.

As for the Wildcats, pencil them in, effectively if not literally, for something higher than 32 on %RPMs, based on Reid. Kentucky might not deserve to be classed as such an overwhelming favorite against the Blue Devils this week, but a safe if revocable working assumption would be the Wildcats could very well be pretty good this season.