Earlier this year, College Hoops Update ranked the eligible transfers. Here were some of their top players:

Several transfers have seen a bump in their ORtgs with their new team. But because the early season provides a number of cupcake games, the vast majority of BCS players look more efficient at this point in the season.  I’ll check back later in the year to see which players maintain a high level of play.

What should we expect as the year progresses? Looking at the 186 D1 transfers into BCS programs in the last 8 years (with over 25 possessions used at both schools), the player’s efficiency at the previous school is usually not a great predictor of performance at the new school:

There is clearly a huge amount of variance in the table. Thus we should be cautious to draw strong conclusions based on how a player performed elsewhere. Wesley Johnson was not a very efficient scorer at Iowa St., but he became a superstar at Syracuse, and those types of transformations are quite common among transfers. But there is information in the performance at the previous D1 school. Notice that the lower right corner of the figure is mostly empty. If the player had an ORtg above 100 at the previous school (and didn’t transfer into DePaul), he is unlikely to struggle with his new team.

Comparing transfers to returning players, the ORtg with the previous team tends to have about half the explanatory power of an ORtg for a player that returns to the same team. The unexplained variance emphasizes how important circumstances are to a transfer’s success. If you transfer into Duke and get to take wide open threes, odds are pretty good that you will have a strong ORtg. If you transfer into a team with no offensive structure and weak complimentary players, odds are you will struggle.

But keep in mind that wild fluctuations in ORtgs from season to season are not uncommon, even for players that return to the same team. In 2010-2011, Tim Jarmusz of Wisconsin, Biko Paris of Boston College, Darius Morris of Michigan, Scott Wood of NC State, Christian Watford of Indiana, Jon Diebler of Ohio St., Jordan Taylor of Wisconsin, Jeff Brooks of Penn St., Alex Oriakhi of Connecticut, Brad Tinsley of Vanderbilt, and Bill Cole of Illinois, all had shocking improvements in their ORtg.

The next table compares changes in ORtg at various percentiles for D1 transfers and returning players. The 90^th percentile means that 10% of D1 transfers experience ORtg improvements of over 21.7. But notice that among returning players, 10% experience ORtg improvements over 18.0.

Players with few possessions on the season (small sample size) explain a lot of these big changes. But even if we restrict the data to players that receive major playing time in both seasons, 10 point swings in ORtg are not rare. 

Notice that on average D1 transfers improve more than returning players. There are probably multiple causes. First, transfers may find a coach that can use their skills better. Second, most transfers sit out a year, and so it is not surprising that their improvement nearly doubles the improvement of returning players.

Another trend is worth pointing out in this table. Notice how the 90^th percentile is different when we restrict the data to players with over 50% of the minutes in both seasons. This is what I call the “coaches are generally not stupid” factor and is explained in the next table: 

Players that receive more playing time tend to be the players that have improved the most. If you played over 50% of the minutes last year, you were probably already good. But if you make a big leap forward, you may move from the under 50% to the over 50% playing time group.