New theory: players generally improve (on average) throughout their career, but maybe the adjustments to a new system (on average) cause a step back in year one?
Our incoming transfers (excluding those jumping up a level) had a lower 3pt% regardless of how many years they had played at their prior school:
>1 year: 257-707 (36.4%) -> 293-857 (34.2%). This is Bos, Guerrier, Mayer, TSJ, Plummer
1 year: 146-449 (32.5%) -> 77-277 (27.8%). This is Stoj, Z, Booth, White. Smaller sample size, but a big enough difference for it to likely be meaningful.
Our outgoing transfers (excluding those going down a level) had a lower 3pt% if they had played >1 year at Illinois, but improved if they had played only one year:
>1 year: 205-576 (35.6%) -> 175-521 (33.6%). This is Hawkins, Goode, Melendez, Grandison, and Curbelo
1 year: 127-391 (32.5%) -> 242-719 (33.7%). This is White, Epps, Clark, and Miller
Since sample sizes (number of shots and number of players) are getting small here, I also looked at the stats combining both the incoming and outgoing transfers:
>1 year: 462-1283 (36.0%) -> 468-1378 (34%)
1 year: 273-840 (32.5%) -> 319-996 (32.0%)
That's some support for the idea that transferring may cause a (bigger) step back for guys who had grown accustomed to their prior team, but someone would want to look at lots more transfers across lots more teams to really know anything.