Week 13: The Prisoner's Dilemma
Win-Win vs. Zero-Sum
Last week we saw how choices create ripples. This week we explore the most famous puzzle in game theory: what happens when two people must decide WITHOUT knowing what the other will do?
We play one of the most famous thought experiments in all of decision science: The Prisoner's Dilemma. It reveals a deep truth about human cooperation: what's best for each individual, if everyone does it, is worst for everyone. We also learn the crucial distinction between zero-sum situations (one person's gain is another's loss) and positive-sum situations (everyone can gain if they cooperate). Knowing which kind of situation you're in changes everything about how you should decide.
- The Prisoner's Dilemma is best played with 2+ people. If you're one-on-one, you play against the student.
- Let the game unfold naturally — resist the urge to coach cooperation. The learning comes from experiencing the tension.
- The zero-sum / positive-sum distinction is one of the most useful frameworks in the whole course. It applies to sibling conflicts, friendships, classroom dynamics, and eventually to business, politics, and international relations.
Week at a Glance
| Component | Details |
|---|---|
| Key Vocabulary | prisoner's dilemma, cooperate, defect, zero-sum, positive-sum |
| Difficulty | Advanced |
| Prep Time | ~15 minutes |
| Guided Session 1 | The Prisoner's Dilemma |
| Guided Session 2 | Fixed Pie vs. Growing Pie |
| Independent Practice | Conflict Analyzer + Mini Prisoner's Dilemma |
Facilitator Preparation
- Prepare score cards and tokens for the Prisoner's Dilemma (details below)
- Make "Cooperate" and "Defect" cards for each player (index cards work great)
- Prepare the payoff matrix on a visible poster or whiteboard
- Have tokens/counters for scorekeeping (20+ per player)
- Think of real examples of zero-sum and positive-sum situations from your family
The Prisoner's Dilemma feels unfair — that's the point. It demonstrates a real tension in human interaction. Don't resolve the tension too quickly. Let the student sit with the frustrating insight that rational individual behavior can lead to collectively terrible outcomes.
For Younger Learners (Ages 8–9)
Simplest version of the concept: "When your choice depends on what someone else chooses, you need to think about what they might do — not just what you want to do."
What to shorten or skip:
- Focus on the Sharing Game (Prisoner's Dilemma) — it's concrete and playable.
- Skip the formal payoff matrix notation. Use "What happens" storytelling instead.
- Keep Nash Equilibrium intuitive: "The spot where nobody wants to change their mind."
- Don't use the term "Nash Equilibrium" with younger learners — call it "the stuck spot" or "the landing spot."
- Keep sessions to 20 minutes.
Adapting the activities:
- Play the Sharing Game with actual small items (stickers, crackers, tokens). Physical rewards make the payoffs real.
- Use only 3–4 rounds, not the full 10. The insight clicks fast.
- For the real-world discussion, stick to playground and sibling examples — "Your sister can share the iPad or hide it. You can share or hide yours."
Journal alternative: "In the Sharing Game, I chose to ___ because ___. Next time I might ___ because ___." Spoken is fine.
What success looks like: The learner can explain "I had to think about what the other person would do, not just what I wanted" in their own words.
- Full payoff matrix with numbers. Challenge students to find Nash Equilibrium by elimination.
- Discuss real-world prisoner's dilemmas: arms races, group projects, environmental cooperation.
- Explore why the "rational" choice (defect) leads to a worse outcome for everyone.
Guided Session 1
The Prisoner's Dilemma
Learning Goal
By the end of this session, the student can:
- play the Prisoner's Dilemma game and explain the payoff structure
- explain why it's tempting to defect even though mutual cooperation is better
- describe what happens when both players think only about themselves
Activities
1. The Story
Tell this version:
"Two friends are caught doing something they shouldn't (let's say they accidentally broke a window with a ball). They're put in separate rooms and each is told:"
- If you BOTH stay quiet (cooperate), you each do 1 hour of community service.
- If you BOTH blame the other (defect), you each do 3 hours.
- If you stay quiet and the other blames you, YOU do 5 hours and THEY go free.
- If you blame them and they stay quiet, you go free and THEY do 5 hours.
Ask: "What would you do? Remember, you CAN'T talk to your friend."
2. Play the Game
Turn it into a repeating game with points (positive framing — you're earning rewards, not punishments):
The Payoff Matrix:
| Partner Cooperates | Partner Defects | |
|---|---|---|
| You Cooperate | You: 3 pts / Partner: 3 pts | You: 0 pts / Partner: 5 pts |
| You Defect | You: 5 pts / Partner: 0 pts | You: 1 pt / Partner: 1 pt |
Rules:
- Each round, both players secretly choose "Cooperate" or "Defect" (hold up a card face-down, then flip simultaneously)
- Record scores for each round
- Play 10 rounds
Let it play out without coaching. Track scores:
| Round | Your Choice | Partner's Choice | Your Points | Partner's Points |
|---|---|---|---|---|
| 1 | ||||
| 2 | ||||
| ... | ||||
| 10 | ||||
| Total |
If only one learner + facilitator: Play the Sharing Game as learner vs. facilitator. The facilitator should vary strategy — cooperate some rounds, defect others — so the learner experiences the full range of outcomes.
If truly solo: Use a decision tree instead. "Imagine your friend can share or hide. For each of their choices, what would you do? Draw or write what happens in each case." The learner maps all 4 outcomes and circles the "stuck spot."
3. Debrief
Discuss what happened:
- "What strategy did you use? Did it change over the 10 rounds?"
- "When did you cooperate? When did you defect? Why?"
- "What's the BEST possible outcome for both players together?" (Both cooperate every round = 3+3 = 6 total points per round = 60 total)
- "What's the WORST?" (Both defect every round = 1+1 = 2 total per round = 20 total)
- "What usually happened?" (Probably somewhere in between)
Key insights:
"The dilemma is that defecting is ALWAYS better for you individually, no matter what your partner does. But if BOTH players follow that logic, you both end up with 1 point instead of 3. Individual rationality leads to collective disaster."
4. The Winning Strategy
Share what decades of research have found:
One of the most historically influential strategies in repeated Prisoner's Dilemma tournaments is called Tit for Tat. In Robert Axelrod's famous computer tournaments in the 1980s, this simple strategy beat far more complex ones:
- Start by cooperating.
- After that, do whatever your partner did last round.
- If they cooperate, you cooperate. If they defect, you defect next round.
- Be willing to forgive — if they go back to cooperating, you do too.
Why it works:
- It's nice (starts cooperative)
- It's retaliatory (doesn't let people exploit you)
- It's forgiving (doesn't hold grudges)
- It's clear (the other person can figure out your pattern and choose cooperation)
This maps to real life: "Be kind by default. Stand up for yourself if someone takes advantage. But always be willing to rebuild trust."
Later research found that slight variations — like occasionally forgiving an extra defection — can do even better. But the core principles remain powerful: be kind by default, stand up for yourself if someone takes advantage, and always be willing to rebuild trust.
Guided Session 2
Fixed Pie vs. Growing Pie
Learning Goal
By the end of this session, the student can:
- explain the difference between zero-sum and positive-sum situations
- identify real-life examples of each
- describe how to look for win-win solutions in situations that seem zero-sum
Activities
1. The Pizza Problem
"There's one pizza with 8 slices. Two people want pizza. Every slice one person takes is a slice the other doesn't get."
This is zero-sum (also called "fixed pie"). Your gain is my loss. The total is fixed.
Now consider:
"Two kids have different chores they hate. Alex hates washing dishes but doesn't mind vacuuming. Jordan hates vacuuming but doesn't mind dishes. If they trade chores, BOTH are happier."
This is positive-sum (also called "growing pie"). Both people gained. The total happiness INCREASED.
"The key question in any conflict is: Am I in a fixed-pie situation, or is there a way to grow the pie?"
2. Sort the Situations
For each scenario, discuss: Is it zero-sum, positive-sum, or could it be either?
| Situation | Type | Why |
|---|---|---|
| Two kids want the last cookie | Zero-sum | One cookie, two people |
| Two friends want to play different games | Could be positive-sum! | Take turns, or find a game both enjoy |
| Competing for first place in a race | Zero-sum | Only one first place |
| Studying together for a test | Positive-sum | Both learn more |
| Sharing space on the couch | Seems zero-sum | But could negotiate a compromise |
| Trading Pokemon/baseball cards | Positive-sum | Each values different cards |
| Sibling fight over the TV remote | Seems zero-sum | But could agree on a schedule |
| A group science project | Positive-sum | Different skills combine |
Key discussion:
"Most real-life situations that SEEM zero-sum actually have positive-sum solutions if you look hard enough. Most conflicts aren't really about one pizza. They're about different preferences that can be creatively combined."
3. The Win-Win Finder
Take a recent conflict or disagreement the student experienced (or a hypothetical one). Walk through this process:
-
What does each person actually want? (Not their position, but their underlying interest)
- "I want the TV" → Actually: "I want to relax and be entertained"
- "I want the TV" → Actually: "I want to watch my favorite show before it's spoiled"
-
Are these interests truly opposed? (Can both people get what they actually want?)
- One person could watch their show now, the other relaxes with a book and watches later
- They could watch together and find a compromise show
-
What creative solutions exist?
- Take turns with a schedule
- Use a different device for one person
- Watch together and alternate who picks
"The magic move is going from 'We both want the same thing and can't both have it' to 'What do we each ACTUALLY need here? Is there a way we both get enough?'"
Independent Practice
Goal
Apply the Prisoner's Dilemma and zero-sum/positive-sum thinking to real relationships.
Activities
The Mini Prisoner's Dilemma requires a partner — play with a family member, friend, or classmate.
1. Conflict Analyzer
Think of a recent conflict with a friend, sibling, or classmate. Answer these questions:
- What did each person want?
- Was it truly zero-sum, or was there a way to grow the pie?
- Did anyone "defect" (act selfishly when cooperation would have been better)?
- What would the Tit-for-Tat strategy suggest?
- What would have been the win-win outcome?
2. Mini Prisoner's Dilemma
Play a simplified version with a family member at least 5 times this week (can use rock-paper-scissors timing with thumbs up = cooperate, thumbs down = defect). Track the results. Did you find a cooperative rhythm?
Decision Journal
Write about a situation where you chose to cooperate even though defecting would have benefited you in the short term. What happened? Looking back, are you glad you cooperated? How does this connect to Tit-for-Tat?
Minimum viable version (younger learners): After dinner or a family activity, answer one question: "Did I cooperate or compete today? What happened because of my choice?" Say it out loud or draw a two-outcome picture.
Reflection Questions
- Is it ever right to defect? When?
- Think about your closest friendship. Is it mostly zero-sum or positive-sum? What makes it that way?
- What would the world look like if everyone played Tit-for-Tat?
Quick Mastery Check
After this week, check whether the learner can:
- Explain interdependence: "How is choosing what to eat for lunch different from choosing whether to share your toys with a friend?" (Looking for: "What I eat doesn't depend on someone else's choice, but sharing does — it depends on whether they share back.")
- Predict a response: "If you always cooperate in the Sharing Game, what will most people eventually do?" (Looking for: "Take advantage" or "Defect, because they get the best deal.")
- Find the stuck spot: "In the Sharing Game, why do both players often end up both hiding — even though both sharing would be better?" (Looking for: "Because each person is scared the other will hide, so they hide too.")
If the learner can explain why "both hide" happens even though "both share" is better, they've grasped the core paradox.
Pause and Notice
After the Sharing Game, ask:
"When your partner chose to hide while you shared, how did that feel? Did it make you want to hide next time — even though you believe sharing is the right thing to do?"
"Game theory shows us that doing the 'smart' thing and doing the 'kind' thing sometimes pull in different directions. That tension is real — and there's no shame in feeling it."
Trust, fairness, and loyalty all live inside these games. When someone cooperates despite the risk, they're choosing to trust — and that takes courage. When someone defects, they might be protecting themselves. Understanding the game helps you choose deliberately instead of just reacting.
This week's takeaway: The best decision-makers don't just calculate outcomes — they also decide what kind of person they want to be in the game.
Spiral Review
- From Week 12: "Every round of the Sharing Game creates ripples. Your choice to cooperate or defect doesn't just affect this round — it shapes whether the other person trusts you next round."
- From Week 9: "Expected value still applies. If you cooperate and your partner cooperates 60% of the time, calculate: is cooperation or defection the better bet? Now factor in the relationship cost."
- From Week 6: "Loss aversion shows up here. The sting of being betrayed (you shared, they hid) feels worse than the satisfaction of mutual sharing — which pushes people toward hiding."
- From Week 11: "Is each round of the Sharing Game a two-way door or a one-way door? In a one-shot game it's closer to one-way. In a repeated game, each round is more two-way — you can rebuild trust."
- Use simplified scoring (cooperate = 2 pts each, defect vs. cooperate = 3 pts / 0 pts, both defect = 1 pt each).
- Focus on the FEELING: "Did it feel good to cooperate? Did it feel bad when someone defected?"
- For zero-sum vs. positive-sum, stick to concrete examples: pizza slices (zero-sum) vs. trading chores (positive-sum).
- Calculate total scores under every possible strategy combination for 10 rounds.
- Discuss real-world applications: international trade agreements, environmental treaties, arms races.
- Research Axelrod's tournaments and why simple strategies outperformed complex ones.