A simple game theory classroom game with Clickers

[#1]

This is a very simply classroom game I do with my students to cover ideas of Nash equilibrium and Prisoner's Dilemma.  It can also show ideas about public goods and externalities.  It is an adaption from Hemenway et al 1987 (see below).  But, it has also been used in different ways by many different instructors. This is the basic set up.

1- Put a slide on the projector with the following information (my ppt slides are included below)

Extra Point Question

     If everyone chooses collude, all students get 10 bonus points in final exam

     If everyone chooses to collude, but one person defect, that person defecting gets 50 bonus points and no other student get any points.

     If more than 1 person chooses to defect, no student get any points.

2- Allow the students to vote (I use iClicker in my class; but any system would work well).

3- Show students the results.

Students are usually very upset when I show the results for the first time.  So I always allow them play more than one rounds.  Sometimes in between rounds I ask them for ideas on how to "solve" the dilemma.  Notice on the video below how the last round students are allowed to switch iClickers, yet the outcome does not change!!!

Anyway, it is a great activity because it takes very little effort to prepare, students like it a lot, and it has many points for applications and discussion.

VIDEO:

The video below shows an instance of me playing the game with my students.  In this occasion, I had about 500+ students in the classroom playing the game.  Notice the distribution of the answers after I close the voting (remember, A) is collude and B) is defect): most students choose to collude, but a few (around 30 or so) choose not to collude. 

A Simply Game Theory Classroom Example - YouTube

PPT Slides:

Dropbox - Slide-for-Game-Theory-Activity.pptx

REFERENCES

Hemenway, David, Robert Moore, and James Whitney. "The Oligopoly Game." Economic Inquiry, 25, Oct. 1987, pp. 727-730; contains copy for class instructions; moore@oxy.edu ; whitney@oxy.edu 

https://www.marietta.edu/~delemeeg/games/games1-10.htm

[chiang]

Do you ever get negative feedback from this experiment, in the sense that you're offering this potentially valuable prize, but due to bad incentives, it is never awarded? I can see a student on the borderline of passing the class claim that these points could have allowed him or her to pass Has that ever happened?

[vazquj1234]

Great question, Eric!  Students indeed react very strongly to this game, which is, of course, what I want as an instructor.  Most are frustrated by the result, since they loose a lot of points due those few who choose to defect.  After the first round I make sure to highlight that I am not trying to "trick" them with the rules, and that the results are simply a function of the prisoner’s dilemma.  But, what really helps with the reactions is to allow students to play several rounds of the game. After each round, I stop and ask students to discuss strategies to "solve" the dilemma.  Some suggests to change the rewards (rules of the game), some suggests exchanging clickers, etc.  I made sure to hear their suggestions, and to try as many as I can.  In the end, they realize first hand how pervasive is the prisoner’s dilemma problem, which is what I want.

I should say the strategy that seems to work the best in getting people to collude is when I allow students to penalize the defectors.  For instance, I change the rules to the following:

- If everyone colludes all students get 10 points

- If everyone colludes and one person defects, the defector gets 50 points and no one else get anything.

- If more than 1 person defect, the defectors loose 5 points, and no one else get anything.

[chiang]

Thanks, Jose. Punishing defectors is an interesting twist to the game. I've never thought of that. But interestingly, that would contradict the original intention of the classic prisoner's dilemma case to incentivize as many suspects to confess. Perhaps it would be useful to discuss after this extension of the experiment that it requires significant "rules changes" in order to overcome the prisoner's dilemma, which is why it's such a powerful concept.

[arobi888]

-- Professor Peter Froehlich, who teaches computer science classes at the highly competitive Johns Hopkins University, contractually grades "on a curve," automatically marking the highest grade an A, with other grades trailing based on their proximity to the class's best. One clever student tried to organize the entire class for December's final exam, to persuade everyone to do no work at all -- thus rendering the "highest" grade a zero, meaning an A for everyone. (Of course, if a single student broke ranks, everyone except that student would receive an absolute zero.) Fortunately for the students, according to InsideHigherEd.com, the class held together, and a shocked professor Froehlich nonetheless honored his contract, giving everyone an A (but subsequently closing the loophole). [Inside Higher Education, 2-12-2013]