Mar 19, 2015; Louisville, KY, USA; Kentucky Wildcats forward Karl-Anthony Towns (left) and Kentucky Wildcats forward Willie Cauley-Stein (right) react from the bench during the second half against the Hampton Pirates in the second round of the 2015 NCAA Tournament at KFC Yum! Center. Kentucky wins 79-56. Mandatory Credit: Brian Spurlock-USA TODAY Sports
This week, Larry Brown had high praise for the Kentucky men’s basketball team, and for good reason, they’re a good team! You’d think Brown would have insightful things to say given his experience in both the NBA and college basketball. Here’s the quote from Wednesday:
"I think they’d honestly make the playoffs in the Eastern Conference if they were in the NBA."
Basketball blogs, sports shows, and office water coolers have beaten this horse to death, but here we are again. It doesn’t seem worthwhile to go any more in depth into why the Wildcats would not beat the Heat, Celtics, or Pacers. A variant of this question raised on twitter seems a little more worthwhile.
Much of this comparison utilizes back-of-the-envelope style calculations, but I also use Layne Vashro’s NCAA Prospect Comparison Tool, including the stat Expected Wins Peak, I’ll let the man himself say a few words about it:
"“Expected Wins Peak” is a quasi-poisson regression model used to project how well a player will perform at the peak of his NBA career. NCAA and international prospects are projected using two similar but distinct models. The NCAA model uses age, experience, MPG, SOS, Height, Weight, and basic per-possession box-score statistics fit from all players to play in both the NCAA and NBA since 1990."
For the time being, I used the top five Kentucky players by minutes played, which gives me:
To find NBA comparisons, I left all sliders on Layne’s tool on their default settings. By selecting an NBA player as a comparison point I am building in a certain amount of bias for these players. Among the top five comparisons for any player, those that played in the NBA are a better-case scenario than those that did not. For a deeper-dive into specifically how Kentucky players in general adapt to the NBA see Layne Vashro’s piece. Here’s the top five for Kentucky by minutes played, their EWP, NBA comparison, and that NBA player’s second year PER, after they’ve had a chance to settle in.
In reference to the original question, “average” can mean a lot things so I took the median wins, which gives me the Oklahoma City Blue (23-20), six back in the Southwest region. It’s helpful to remember that these D-League teams are not full of scrubs. They’re full of All-Americans, All-Conference and former blue-chip players. Jeremy Lin, Danny Green, Gerald Green, and Chris Anderson all spent time in the D-League.
Taking the Blue’s top five by minutes played this season gives us the 2013 A-10 rookie of the year, Semaj Christon, 2012 NCAA champion Marquis Teague, All-ACC honorable mention Talib Zanna, Josh Huestis (subject of a great Zach Lowe article), and Maurice Sutton, formerly of Villanova. I don’t have EWP for all of these players so I won’t use it for direct comparison. Instead I calculate an estimated 2nd year NBA PER by taking their D-League PER from this year and multiplying it by a factor derived from previous D-Leaguers who have gone on to play significant minutes in the NBA.
Here’s the comparison between this Oklahoma City Blue “starting five” and the Kentucky five.
They’re fairly close, but these five Kentucky players edge out the Blue, 63 to 55 “PER points”. Using this method shows just how rose-tinted taking some of the original NBA comparisons are, just for starters expecting Karl Towns to play at Tim Duncan productions levels is extrapolating a lot on the 20 year-old’s ability to transition his game to the next level.
If you’ve watched any Kentucky basketball this year you may have heard about John Calipari’s “platoons”, planned hockey line-up like changes that keep nine Kentucky players consistently coming on and off the court. So there’s some depth unaccounted for by taking only five players. But there’s even more depth on D-League team benches. To get a handle on this here’s the same comparison expanded to Kentucky’s nine along with the top nine by minutes played for the Blue. Also included are the top nine for the bottom and top of the D-League, the Idaho Stampede and the Bakersfield Jam.
Kentucky still comes out on top. But, keep in mind what this is saying – a team of young NBA players as determined by their similarity score to the current Kentucky team is “better” as measured by PER, given that the D-League player’s PER is handicapped by a factor relative to previous D-Leaguers who then played in the NBA.
I think the back-of-the-envelope nature of this is pretty clear, and mostly a thought experiment. But the point I hope to make is that there is a middle-ground between the two sides where we can accept that this Kentucky team is really really good.