For the past 24 to 36 months, high school recruiting has been declared all but dead in the water by a number of outlets and even by some head coaches in the college game. The reasoning is simple: with the transfer portal now basically serving as the college version of free agency, why go for high school players who take longer to develop and whose outcomes have higher variance when you can just go for players who are already proven in the college game?
Last year, I pushed back against that sentiment in Quantifying the Role of Experience in College Basketball Roster Building, where, after analyzing 1,500+ college basketball rosters from the past five years, the numbers suggested that transfer recruiting is far from the golden ticket it has been portrayed to be, as there wasn’t a correlation between stacking your team with transfers and actually winning.
The numbers also showed that the strongest correlation with winning came from teams that gave returning players a bigger share of their minutes.
So, of course, Michigan goes on to win the national championship this year with a roster that is seen as confirmation that not only can you win big with transfers, but that you might actually need them to do so.
So, has the right way of building a basketball roster fundamentally changed over the last year? Were the numbers in my article from last year wrong, or did they suggest something I didn’t see? Or is the framing of Michigan as a “team of transfers” not exactly right?
In order to answer that, we’ll have to go, once again, to the numbers.
