Grading, game theory, and group evaluations..

When I was an undergrad econ minor, second semester senior year I took a business and government class, primarily because it was one of the only econ classes that was offered on Tuesday/Thursday. The prof I had (who wrote this textbook) had a great grading policy. He said that he didn't want grading to be a "zero-sum game". His thought was that in a traditional grade-on-a-curve, students are angry at those who suceed because they throw off the curve. So instead he built about a 15-point curve into the grading, which was great, since by the end of the semester it worked out that I needed like a 68 on the final to get an A in the class.

One of the MBA classes I'm taking this semester is just the opposite. We had a group project that we had to work on, in self-selected groups of 2-5 people. At the end of the semester, our prof passed out the most evil group evaluation I had ever seen. We had to give our group members, including ourselves, points, which had to total 11. Points assigned could not be fractional. In other words, if your group has more than 1 but fewer than 11 members (which all did), someone had to get fewer points.

Our group of 4 was pretty happy with each other - none of us were perfect, but we all did our best and worked together pretty well. So we decided on some OPEC-style cartel collusion. We each gave ourselves 2 points and all the other group members 3 points. We figured this would average out as equal plus show that we felt our fellow members pitched in We all watched each other fill out and hand in our surveys, to ensure no cheating.

I don't know how effective this was so far - final grades haven't been turned in yet. While it is certainly creative, it also seems somewhat unfair to create a grading system where someone has to lose.