I use TBL in two undergraduate criminal justice courses of about 30 students each. I used forced ranking peer evaluations in the past and was happy with the results. It got students to think carefully about individual contributions to the team, although they were uncomfortable doing so. I also allowed them to work as teams and to game the system because only the high functioning teams could figure out how to do so. However, this semester I have many students who have had me before, and therefore know how to game the system already, so I changed the format of the evaluations from last semester and also required them to work individually on the evaluations. Several teams cheated or tried to cheat to distribute points evenly across all members. I dealt with the cheating issue and eventually got them to do the assignment individually, but this is going to be a continual problem since I often have students for more than one semester. I've found this thread helpful as I'm trying to think of ways around this problem that will not increase my grading time. I like the idea of evaluating the evaluations.