Confirmation Games

KatnissThe New York Times recently published an interactive illustration of confirmation bias: guess the rule obeyed by a sequence of three numbers. I won’t go into detail about the game other than to say that it’s wonderfully simple and I definitely fell into the mind trap. Some implications for politics and business are presented after players submit their answers and this can provide a launch point for class discussion.

The puzzle nicely complements the Zendo game in which players create hypotheses about an arrangement of blocks — by demonstrating the cognitive biases that affect much of our decision making.

Reflections on a prisoner’s dilemma: guest post from Katherine Wright

In this week’s seminars for our first-year module Introduction to Politics: Power and the State, on Rational Choice theory, I ran a prisoner’s dilemma game with the students. Students were told that they had been arrested for a crime along with their accomplice and could not communicate with this partner in crime.

In each of the three seminar groups we played 2 rounds of the game. For the first round I gave them a name of a celebrity (and I settled on Kim Kardashian as someone they MUST have heard of [certainly have here at ALPSBlog]) on a piece of paper. I told them that they were not allowed to show the name they had to anyone else in the class and that if they had their own name to let me know. I wanted to give them the impression that they did not all have the same name and that someone could have their name.

Prisoners dilemma

After the first round I tallied up on the board who had betrayed/stayed silent and asked each student to justify their reasoning.

The second round saw the students given a name of a classmate, someone among the 12 or so students in the class that day.

The students began to see how their perceptions of their classmates influenced their decision to betray them or stay silent. Many in the second round chose to stay silent because they had seen their partner in crime stay silent in the first round, only to find that their partner had completely shifted positions in the second round and they were left feeling betrayed. Perhaps next time I will play the two rounds without the students revealing how they have voted in the first round until after we have played the second round..

Some students spoke about how their conscience would not let them betray their partner, that they had been brought up not to “rat out” someone else or that the their partner in crime  was a friend they trusted. Others were resolutely rational and said that they would not risk 10 years in prison and 5 years was clearly the most rational decision to make.

Surprisingly, among the three seminar groups the first group – just BSc Politics students – was the most rational with the majority choosing to betray their partner in crime. I had expected that the final group – with BSc Politics & Economics students in it – would take the most rational stance, this was far from the case however and in the main they showed a noble sense of camaraderie (along with the second group of BSc Politics & Sociology students)!

For the last two groups, those who took a risk on a longer sentence and acted ‘irrationality’ were rewarded. However, for the first group this was not the case and those who played it safe and acted ‘rationally’ were rewarded. This set the groups up for an interesting discussion on the merits of Rational Choice theory and one in which I think they felt invested – especially given that many of their peers had just betrayed them!

Katharine A. M. Wright is a Doctoral Researcher and Seminar Tutor in the School of Politics, where she holds a Faculty of Arts and Human Sciences Scholarship. Her thesis examines NATO’s implementation of UN Security Council Resolution 1325.

POL1012 Group 1

POL1012 Group 2

POL1012 Group 3

When Hobbes Turned Liberal Institutionalist….

This week I returned to my roots to replay another version of Victor Asal’s Hobbes game. One of my favorite things about the Hobbes game is that it can always be slightly altered to introduce new kinds of interaction in the game.

In order to do this you really do need to play Victor’s original game first.

Then, a week or two later…. hand out the cards again. I wanted Wednesday’s interaction to be more akin to the picture of the international system represented by liberal institutional models of interaction. Specifically, I wanted to introduce variation, not simply in terms of individual power levels, but types of potential cooperation.

Tell them, today, the kind of card you get matters. (watch them peer at the front of their card anxiously….see how the people holding aces and kings begin to smile)

Project an image of the hierarchy of poker hands on the board and explain which kinds of hands are better than others. (you will get confused looks from people who have never played cards…don’t linger on this …. just smile and say…. everything will be alright)

The text and images below are from the website but you can find the image anywhere really.

Straight Flush & Royal Flush


Four of a Kind:

Four of a Kind

Full House:

Full House





Three of a kind:

Three of a Kind

Two pair

Two Pair

One pair

One Pair

High card

High Card

Now… tell them that individual cards can challenge each other. In that case, the higher value card wins, takes the other card, and then the loser sits down.

BUT… they are also free to create pairs, triples, and even complete 5 card poker hands. (I limited them to five…. continue to ignore the people who don’t know cards)

Then…. say…. go!

Observations by Students in Debrief:

  1. MASSIVE VARIATION in behavior from the original game….. some team up and produce collective security others go it alone.

  2. Students who had initially powerful cards (Aces) felt more assured than those who had low number cards. But everyone had an opportunity to collaborate to build a good poker hand in order to feel safer.

  3. Students reported feeling less concerned about relative gains and more interested in absolute gains as they searched for other potential allies.

  4. All the students agreed that the structure of the game more closely aligned with what they imagined the international system to be like: more opportunities for cooperation in many different ways, but still anarchic and fraught with distrust and fear for survival.

  5. Students who had no idea about poker hands were sometimes preyed upon, but most often, were assisted by colleagues with similar cards.

Prior to this exercise the students were unconvinced by the Kupchans‘ work on collective security and the institutionalist perspective. They preferred the ‘pragmatism’ of Mearsheimer and Waltz. After the exercise, the students reported understanding, more clearly, what the Kupchans’ were getting at. I’ll definitely do this again in tandem with the original form of the Hobbes game.

Identity Politics Activity

Due to some late summer travel and other obligations, I’m working this week to put together my fall syllabus for Intro to Comparative Politics. One of my favorite class activities comes about mid-way through the semester when we talk about identity politics. One of the main ideas I hope the students take away is that identities can be manipulated for political reasons. More specifically, the following activity has these learning objectives:

  • Explain the difference between symmetric and asymmetric cleavages.
  • Identify the conditions under which particular identity categories will be politically salient.
  • Predict the consequences of a permanently excluded minority.
  • Compare and contrast the political implications of fluid and fixed identities; symmetric and asymmetric cleavages.

Students are given colored index cards to represent one identity category (Green or Pink) with a language written on it (I use Esperanto or Ido). The “Round 1” table is projected with the number of students in each group (see Table at end of post) and students are asked to form a governing coalition that represents at least 51% of the population.

Once a coalition is formed, we discuss (1) what coalition was formed, (2) why, and (3) what happens to those that are excluded? I then project the map of the hypothetical country, showing a significant natural resource in the area controlled by the permanent majority. The students predict the likely consequences of the permanent majority’s control over a natural resource (I use a bag of leftover Halloween candy to illustrate the “natural resource” that the permanent majority can choose to distribute as it wishes). Next, they discuss the likely responses of the permanently excluded minority (e.g. civil war, terrorism).

I collect and redistribute cards and project the “Round 2” table. Students again form coalitions and we continue the discussion. In “Round 2” there are different possible coalitions and identity categories are fluid. The students then compare the political implications (likely democratic stability, probability of conflict between ethnic groups) in the different rounds.

The tables below give a rough approximation of how I allocate the identities, but the table I project in class has the number of students rather than percentages. What’s important is that the “green” coalition is “obvious” in round 1, while the second round has multiple possible coalitions.

Round 1

Esperanto Ido
Green 40% 30%
Pink 10% 20%

Round 2

Esperanto Ido
Green 40% 20%
Pink 20% 20%


Let me know if you have any questions or how it works if you give it a try.

Midterm Review Game: Trivial Pursuit

As midterms approach, its a good time to think about how we help our students prepare for their exams.  Approaches vary widely: some professors design study guides and hold extra review sessions; others are more of the ‘cut the apron strings’ mode and leave students to study completely on their own.  I’m agnostic toward the method, as long as the intention behind it is sound; as with most things in teaching, the tool (say lecture, or simulations, or using a study guide) should be chosen because its the best fit for a specific task and a specific group of students, not because we are in love with the tool itself.

In my introductory courses in international relations and American politics (the only times I employ a midterm), I am a fan of the review game. My exams rarely focus on definition of terms, so I find a traditional study guide completely counterproductive: students focus on learning a list of terms, rather than thinking about application, analysis, or synthesis.  A review game like Trivial Pursuit can put the onus of preparation on the students themselves (where it belongs!), and forces them to think about questions, rather than just answers.  It also makes the process of review way more fun than a standard review session. This works great for a final exam review, too–and if you do it twice, you may find that the students produce even better questions the second time around.

Here’s a how to:

  • Equipment needed: any version of trivial pursuit–just the board, die, and game pieces. A typical board is pictured above
  • Two weeks before the exam, divide the material on the exam into 5 or 6 categories.  For my intro IR class, these are: conceptual foundations (levels of analysis, states v. nations, sovereignty, etc); historical context of IR; Realism and Power Politics; Idealism and Neo-Liberalism; Constructivism and other ISMS (Marxism, Feminism, etc); and Current Issues (which covers class discussions and assignments).  For American politics, these are usually Popular Culture and Citizenship; The founding and the Constitution; federalism; civil liberties; civil rights; and current issues.  Your categories will certainly differ, and can always include a ‘wild card’ option.  The game would work just as well for comparative politics or any other course.
  • Pass out or post a one page description of the review game that also serves as a ‘midterm reminder’ guide.  Mine usually just notes what the exam will cover, the format of the exam, some brief tips on how to study, a few sample questions, and info about the review game.  Contact me if you’d like to get a copy of mine as a guide.
  • Assign students to each category (randomly is best), and tell them they are each responsible for producing a certain of questions (with answers) on that category.  Depending on the number of students, I’ll ask for between 5-10 questions.  You can decide whether a certain number have to be multiple choice (mine tend to go for short answer).  They also have to provide the answer.  Note down who is assigned to what category.
  • The questions are their ticket to the review.  Students who do not turn in questions cannot participate in the review. Added incentives can include extra credit for the winning team, and I will tell them that particularly good questions may show up on the exam itself (i always try to use at least one, sometimes with some minor edits).
  • Questions can be submitted in a variety of ways.  The easiest for the instructor is to have students write them on index cards–category at the top, question on one side, answer on the other.  They can bring them in the day of the review.  I usually have them submit them online a day or two early so I can organize the questions, and if I’m feeling charitable, post them for everyone to see after the review.  It is VERY important that you know who submitted which questions.
  • On the day of the review or sometime before, divide the students into two (or more if you have a large class; 10 people per team is a good maximum) teams making sure that there is someone from each category on each team.  Then make sure that you organize the questions (whether on note cards or electronic) so that teams do not get asked the questions they wrote.  This is really important, otherwise people just answer the questions they wrote without anyone having to think about the question.  I like to assign teams in advance, and make adjustments last minute if someone doesn’t show up.
  • Bring additional questions. Many of the questions the students write will overlap, and so its great to have some of your own to use in a crunch.  This is where those textbook test banks come in really handy.
  • Ask each team to choose a die roller and a spokesperson.  This minimizes down time during the review, and makes clear who is communicating the ‘final answer’ to the instructor, who asks all the questions.
  • Arrange the seats in the classroom around a central table so everyone can see the game board.
  • Assign each category a color. Put it on the board so everyone can see.
  • The basic rules of trivial pursuit apply.  Students roll the die and move their pieces, and answer questions corresponding to the color they land on.  Correct answers allow them to go again; incorrect answers give the next team a turn. The large spaces at the end of the spokes of the wheel earn correct answers a little colored wedge; the goal of the game is to collect one wedge of each color and then answer a question correctly after moving to the center of the board.  The only additional rule I use is that after either three correct answers or earning a wedge, the next team gets to go.  Otherwise you can have a team go on a spree, which can lead to students on other teams checking out of the game.

That’s it!  It involves very little preparation on the part of the professor, and instead puts the burden on the students, which is great for us but also for them, as the process of writing questions and then working together to answer them can really help them retain information.  I’ve seen the average on the midterm improve since using this review game, although the usual limits on a such a claim apply since I do not have pretests or true control groups.

Try it out, and as always, report back here on how it goes.

Michelle Allendoerfer: Fearon’s Rational Theory of War

Today we have the first of two posts by Dr. Michelle Allendoerfer, Assistant Professor of Political Science, Women’s Leadership Program (, George Washington University:

Rationality of WarOf the concepts typically taught in an Introduction to International Relations class, Fearon’s rationalist theories of war is one of the most challenging for students.  Getting students to actively engage with these concepts makes them more concrete and less abstract.  There are bargaining activities out there to illustrate the bargaining theory of war, but many treat wars as “all or nothing” — the winner gets the spoils while the loser gets nothing (minus the costs of war to both parties).  I wanted an activity that reflected a wider range of war outcomes in which the loser may end up with some division of the spoils.

I divided my class into dyads made up of two groups.  I made eight groups of 3-4 students (four dyads).  Each dyad gets a modified deck of cards (I used only 2-9 with more cards in the middle of the range), tokens (10 per round), a coin, and a handout.

To play, each group takes a card from the deck. In all rounds but the first, this information is kept private.  The challenger, selected by a coin toss, makes an offer to divide 10 tokens. I used Hershey kisses because my class is small and motivated by chocolate; anything that incentivizes the students will work, such as poker chips tied to extra credit.  If the defender accepts, the tokens are divided according to the offer and the round ends.

If the defender rejects, the groups go to war by showing their cards. I linked the division of tokens to the margin of victory. For example, if side A plays a 6 and side B plays a 3, the margin of victory is 3 in favor of side A and this corresponds to a 8/2 split of the tokens – this information is on the handout provided to the students. If a war occurs both sides pay the cost of 2 tokens.

A practice round helps get the students going.  An initial practice round where cards are not hidden also illustrates to students the crucial point of Fearon that war is ex post inefficient and should not happen in conditions of complete information.

In my next post, I’ll share some of my reflections on the activity and some notes on debriefing.  Feel free to email me (mallendo-at-gwu-dot-edu) with questions or for a copy of the handout I use to facilitate the game.  I’m happy to share those materials.

The Attack of Dr. X: Collective Security Role Play

Here’s a quick role play exercise that takes about 5 minutes and zero preparation.  Announce to the class that Dr. X down the hall (pick any faculty member, preferably someone they might know or who seems completely non-threatening) has just run into the classroom and randomly tried to tackle a member of the class.  Then ask them what they plan to do about it.  If they hesitate, you can start giving a play by play of what’s going on, and then prompt them again to see if someone is going to intervene.  Usually at this point if not sooner, a student or two announce that they will tackle Dr. X; others might offer to hold him/her down or kick them.  You can continue storytelling–maybe the first tackler gets roughed up a bit. Once five or six students have announced their actions, however, offers dry up as the threat seems to be handled.  Then start quizzing the students who were not quick to act on why–you will get a range of reasons, but at least one student will note that their efforts weren’t necessary.

That is, of course, your in to start a discussion about collective security and the problem of free-riding.  Their willingness to allow other students to handle the threat is exactly the kind of behavior we predict amongst sovereign states in a collective security arrangement, where as long as the threat is handled it is in no single state’s interest to expend the resources for that purpose.

I use this exercise whenever I discuss collective security; usually this is in the context of discussing the League of Nations and why it failed.  Its useful because it is very quick, requires no preparation for faculty or students, and is very effective in making the point.  As an ‘organic’ simulation (see Kollars and Rosen 2013), it allows students to show the truth and practice of the theory through their own actions before studying it formally–thus making them more likely to understand and accept the theory.  Plus, Dr. X can become a villain for the rest of the course, and you can constantly reference him/her whenever you need to discuss ‘threats’ to state systems.


Whodunnit? Part 2: Murder Mystery in the Classroom

Back in July I mentioned that I was toying with the idea of using a murder mystery party as a classroom tool.  Today I did just that in my methods class as a nice use of the half-week before Thanksgiving.

Essentially I bought and then adapted a commercially available murder mystery and ran it during a 1h20 session of class.  I assigned each student a role a month in advance, giving them a basic character sheet, some information about the world we were in (it was fairy tale themed), and instructions on what would happen on the day of the event.  The only thing required was willing participation, but I awarded extra credit for those who showed up in costume, threw themselves into the performance, or correctly guessed the murderer.  At the event itself (which I held in our usual classroom, but played music through the computer and brought in some baked goods) I gave them each a nametag, a list of initial objectives (questions to ask other characters and information to impart), some play money to use for bribing information out of others, and another copy of their character sheet.  They had 30 minutes to achieve their initial objectives before the actual murder occurred; then they received a second set of objectives and information and had another 30 minutes to try to figure out who committed the crime, why, and how.  The last 20 minutes I reserved for accusations, explanations, and debrief.

In terms of the event, everything went very smoothly and the students all appeared to enjoy themselves.  I played the victim so they could all focus on solving the crime, but I think that many of them got a bit too caught up in playing their roles and paid less attention to the evidence that I had gathered for them.  None of the students managed to guess the murderer, although several correctly identified the motive and/or the means.  Quite a few of them based their accusations solely on ‘shady’ behavior by a given character; some students admitted to never looking at the physical evidence despite my posting it prominently on the board and suggesting they do so.  I almost considered posting copies of the evidence online and tasking them with thinking everything through a bit more and not revealing the solution until next week, but decided against it mainly because I could not trust that the student playing the murdered would be able to keep quiet for a week, and had no opportunity to ask without raising suspicion. In retrospect, I should have done so.

The debrief phase did involve me tying the exercise back to methods–the whole point was that this was an exercise in analyzing evidence and drawing conclusions, and as students explained their reasoning I pushed them to discuss the evidence and how it influenced (or in several cases played no role in) their deliberations.  But as with most simulations, there simply was not enough time to really dig into this enough to satisfy me.

The Pros:

–the students had a blast, and anytime you can associate methods with ‘fun’ is a win

–the exercise really does have excellent ties to the overall course lessons and learning outcomes.

The cons:

–the exercise definitely needs more time.  Everyone managed to achieve their objectives, but there was little time available for students to really think through the information and evidence, and the debriefing was more abbreviated than I would have liked.

–this was also somewhat time consuming for me.  Although the mystery I bought contained everything necessary to run the event, I need to make a number of adaptations.  This may be particular to this mystery and/or company, but there were several mistakes in the mystery (conflicting information about where one character was living, for example) and the murder was pretty much impossible to solve as written.  I know this because I did a dry run with some friends (mostly faculty from various departments at my university) and got a lot of useful feedback on it.  I therefore had to re-write all of the objectives for each of the 15 characters, both to correct for the inaccurate information and to add information so that every player had enough to go on to be able to have a possibility of solving the crime.  That, as you can imagine, took some time, as did putting together the pre-game envelopes for each student.  All told, it took probably 3-4 hours of prep time for me in the last two days–not including the original character assignment or any time spent doing the prep for the dry run.

I still think this exercise has a lot of potential, so I plan to do it again but with some adjustments.  I would still do the roleplay in a single class session, but give them the evidence and the ability to continue questioning each other online over several days.  Then each of them will have to turn in a 1 page report stating their case, due at the next class (say Thursday for role play, Tuesday for paper + solution and debrief).

I also may experiment with smaller versions of this–ditch the roleplaying and just lay out the facts of the case and evidence for them and set them to solving the crime, rather than participating in it.  I think it will still be fun, but will keep the session more focused on the lessons at hand.

The Beer Game

Poverty Games, Part 5: Inequality Monopoly

Some of my sociologist friends turned me onto a common game they use in their field to teach about social class and inequality: Inequality Monopoly (also called Modified Monopoly, Development Monopoly, or Stratified Monopoly).

The basic idea is to have students play a classic game of Monopoly (8 students to a board, multiple boards as needed) but with a twist: starting resources, salary, jail, and other effects of the game change based on which ‘social class’ you are randomly assigned at the beginning. So the upper class player may start with $2500 and a couple of hotels they can place on the first property they land on, while the middle class starts with $1500 and a house, the working class $1000 and lower class $500.  Salaries change too; only the upper class gets the full $200 for passing Go.  You can also put limits on who can purchase utilities or railroads, make it easier for the wealthy to get of jail quickly, and institute a lottery system that costs 5 or 10% of your income to play.

You can also speed up the game by simply distributing property in advance; some of the rule sets below have ideas on ways to do this.

My colleagues who use this game find it a very effective way to dig into problems of poverty, social class, income inequality, and the challenges of ‘equality of opportunity’. I haven’t tried it myself, but its worth importing to our field and seeing what happens.

Resources to Play the Game:

  • The best resource available is this cite: Fisher, E.M. (July 2008). USA Stratified Monopoly: A Simulation Game about Social Class Stratification. Teaching Sociology 36(3): 272-282.

That article contains everything you need–learning objectives, instructions for the instructor, a great rules chart for each player, timeline (broken down for classes of 50, 70, and 100 minutes), discussion questions, related written assignments, and assessment.

  • This document is a great resource that’s easy accessible.  It has a set of rules, instructors for the instructor, worksheets for the students, debriefing questions and essay questions.
  • Here’s another document with very similar rules.
  • This is a wonderful set of powerpoint slides that detail several different rule sets, have notes on important concepts, and discussion questions.
  • Richard Harvey at Saint Louis University comments here on the rule set he uses and his assessment of the game.
  • Here is an interesting piece on the use of Monopoly to study institutionalized racism.
  • This video made the social media rounds a month or two ago, and is about a series of studies on the unethical behavior of the rich.  One study involved a game of monopoly with ‘rich’ and ‘poor’ players–poor players getting half the starting money, half the Go salary, and only getting to use one die (thus not being able to roll doubles and take an extra turn)–and showed that the randomly determined rich players at the end of the game believed that they deserved to win the game due to good gameplay–not luck of the draw.

See earlier entries in the Poverty Games series:

Poverty Games Pt 1: Ayiti the Cost of Life

Poverty Games Pt 2: Third World Farmer

Poverty Games Pt 3: Free Rice

Poverty Games Pt 4: Spent