A woman accepts her partner's apology every time he cancels plans at the last minute. She tells herself she's being understanding. What she's actually doing is teaching him that cancelling carries no cost. Six months later, she's furious — and he's genuinely confused, because from his side, nothing changed. The pattern was set months ago. She just didn't see herself setting it.

A man runs a small team. One person consistently delivers late. The first time, he has a quiet word. The second time, he covers the gap himself. The third time, he vents to a colleague instead. By the fourth time, the whole team knows: deadlines are optional for that person, and the manager won't do anything real about it. He didn't lose one battle. He lost the rules of the game.

A guy keeps lending money to his brother. Each loan is small — fifty here, a hundred there. Each time he tells himself it's family. Each time his brother says he'll pay it back. After a year, the running total is significant and nothing has returned. He's resentful. His brother is oblivious. The game was decided in the first round, when the first loan went unrepaid and nothing happened.

Three different situations. Same underlying structure: each person made a reasonable choice in the moment that set a terrible pattern for every moment after. Not because they were uninformed. Because they were solving the wrong problem — managing the discomfort of this interaction instead of the shape of all the ones to come.

This is the first post in a ten-part Game Theory series. The primer covers what repeated games are and why they matter. This post is about the specific ways people wreck them — and a practical process for changing the pattern once it's already set.

The Gap Between Knowing and Doing

You can understand repeated games perfectly and still lose them. Most people do.

The problem isn't the concept. It's the two-second gap between stimulus and response — the moment where your emotional system makes a move before your strategic thinking catches up.

Under pressure, your brain doesn't consult your game theory knowledge. It does what it always does: reduce the threat as fast as possible. The sharp comeback. The silent treatment. The "fine, whatever." These aren't strategic failures. They're survival reflexes running in contexts where survival isn't actually at stake. Your nervous system doesn't distinguish between genuine danger and your partner's disappointed tone. It responds the same way.

This means the failure point in repeated games is almost never "I didn't know the right move." It's "I knew the right move and did something else because the pressure was on." The gap between understanding and performance is where most of the damage happens.

And that gap is driven by five predictable patterns.

Five Ways You Wreck Repeated Games Without Realising

1. Emotional accounting

You keep a mental ledger of who owes what in your relationships. The problem is that your ledger is rigged.

You remember your sacrifices in high definition — the time you stayed late, the favour you did, the thing you let go. These are stored vividly, with emotional weight attached. You remember what the other person contributed in low resolution, if at all. Their efforts are facts you vaguely know. Yours are feelings you carry.

The result: you genuinely believe the balance is unfair. Maybe it is. But your accounting system is so biased that you can't actually tell. And when you act on that sense of unfairness — pulling back, getting sharp, keeping score out loud — the other person experiences it as an attack from nowhere. From their equally biased ledger, they've been doing their share.

Two people, both feeling shortchanged, both responding to grievances the other person can't see. This is how cooperative relationships slide into mutual resentment without either person intending it.

2. The precedent blindspot

People think in incidents. Games think in patterns.

When you say "I'll let it go just this once," you're thinking about this particular moment. The game registers something different: that behaviour had no cost. When you cave on something you said mattered, the game registers: stated standards don't hold.

None of these feel like decisions. They feel like non-decisions — you're just getting through the moment. But in a repeated game, not responding IS a response. Silence is data. Tolerance is instruction. Every non-action sets a precedent as clearly as every action.

The most expensive patterns in people's lives were almost never deliberately chosen. They were allowed, one "just this once" at a time, until the pattern was entrenched and changing it felt like starting a war.

3. Relief addiction

Your brain rewards you for reducing discomfort. That's not a moral failing — it's neurochemistry. When tension, conflict, or guilt shows up, your system is motivated to make it stop. Agreeing, apologising, withdrawing, snapping — anything that drops the pressure in the next thirty seconds gets reinforced.

The problem: what works in thirty seconds often fails over thirty days. Agreeing to keep the peace means resentment later. Snapping to release frustration means repair later. Avoiding the hard conversation means a harder one coming.

Each relief-seeking move works. Briefly. And the brief relief is enough to reinforce the habit. So you do it again. And again. Until you have a chronic pattern of trading long-term position for short-term comfort — and the long-term discomfort is now so large that no single conversation can fix it.

4. The fairness trap

"They did it first" is the most common justification for pulling back in a repeated game. And it feels completely valid. If someone treats you badly, responding in kind feels proportional. Fair. Even necessary.

Here's the problem: your version of "proportional" is almost always heavier than theirs. You felt the full weight of what they did. You experienced a fraction of what drove it. So your response lands on them as an escalation. Their response to that lands on you as a further escalation. Both people genuinely believe they're just responding. Both are driving the game downward.

The fairness trap is especially destructive because it feels righteous. You're not being petty — you're being fair. The emotional logic is perfect. The strategic result is catastrophic. In repeated games, fairness isn't determined by matching the other person's worst move. It's determined by whether your move makes the next round better or worse.

5. Invisible defection

Most people recognise defection when it's loud: lying, breaking promises, shouting. But the most common forms in daily life are quiet, and the person doing them usually doesn't see it.

Going silent instead of saying what's wrong. Agreeing in the conversation and doing something different afterwards. "Forgetting" a commitment you never wanted to make. Being physically present and emotionally checked out. Doing the minimum while technically complying.

These are all defection. They all grab short-term comfort at the other person's expense — because the other person is now operating on false information. They can't fix what they can't see. They can't plan around intentions you haven't shared. The prisoner's dilemma post goes deeper into what happens when both people are doing this simultaneously — the mutual guardedness that drains everything while both people believe they're just protecting themselves.

If you want to know whether you're defecting, ask one question: "Does the other person have an accurate picture of where I actually stand?" If the answer is no, and that's because you haven't told them, you're defecting. Quietly, politely, with full deniability — but defecting.

The gap between knowing and doing isn't closed by more knowledge. It's closed by recognising which trap hijacks you — emotional accounting, precedent blindspot, relief addiction, fairness trap, or invisible defection — and having a structured response ready before the pressure arrives.

Diagnostic: Eight Ways This Shows Up

This is a mirror, not a lecture. If you recognise yourself in several of these, that's normal. Everyone defaults to at least two or three under pressure.

Common Failure Modes

The Repeated Game Reset

Practical Framework

Six Steps to Change the Pattern

Use this when you're stuck in a recurring dynamic — with a partner, a colleague, a friend, a family member, or yourself — and the usual approach keeps producing the same result.

  1. Name the game. Is this a one-off, or will you interact again? Who will remember this move? If you're going to see this person next week, next month, or next year, you're in a repeated game, and short-term moves cost more than they feel like they do.
  2. Define the real payoff. What do you actually want in six months? Not what feels satisfying right now — what you'd choose if you were calm. Trust? Respect? Better work? Peace? If your current behaviour is working against that goal, you have your answer.
  3. Separate emotion from strategy. Two questions. First: "What am I feeling right now?" Name it honestly — anger, hurt, frustration, exhaustion. Second: "What would give me relief, and what would actually improve the payoff?" Write both down. They're almost never the same thing. Choose the second.
  4. Choose one move. A cooperative bid, a boundary, a consequence, a clarification, a pause, or an exit. Not all of them. One clear move that serves the payoff from Step 2. If you can't decide, pause. Pausing is always better than reacting.
  5. Make the move credible. What will you do, not just say? Actions are the currency of repeated games. Words that aren't backed by behaviour become noise — and noise trains people to stop listening. This is where commitment devices become useful: structures that make your move hard to reverse.
  6. Review the pattern. After the move, check: did cooperation increase? Did defection repeat? Is the game improving, or are you running the same cycle with better vocabulary? If the pattern hasn't shifted after several clear, consistent rounds, the game may need structural redesign — or you may need to leave it.

Three Repeated Games in Action

Case Study: The Helpful Colleague Who Burns Out

A woman keeps picking up slack for her team — staying late, covering for people, absorbing extra work. She doesn't want to be "difficult." After a year, she's exhausted and resentful. Her team has learned she'll absorb anything, so they keep sending it her way. When she finally pushes back, they're surprised and annoyed. From their perspective, nothing changed.

Why this fails: She trained the game for twelve months and then blamed the players for learning the rules. Her cooperation was unconditional, which means it carried no information about her limits. Everyone assumed she was fine because she never signalled otherwise.

The reset: She starts responding with trade-offs instead of absorption: "I can do this or that this week — which is the priority?" She stops covering without being asked. Within a month, work redistributes. Not because people became more generous. Because the game changed.

The lesson: Unconditional cooperation doesn't build goodwill. It builds expectation. If you never signal a limit, people don't know you have one.

Case Study: The Bedtime Battle

A father tells his daughter bedtime is 8pm. She pushes back. Sometimes he holds firm. Sometimes he caves because he's tired or she's been particularly good that day. She doesn't learn "bedtime is 8pm." She learns "bedtime is negotiable, and the outcome depends on Dad's energy level." So she negotiates every single night. He's exhausted. She's doing exactly what the incentives taught her to do.

Why this fails: Inconsistency doesn't teach the rule. It teaches the loopholes. The daughter isn't being defiant — she's being strategic, based on the data she's been given.

The reset: He picks a time and holds it. Every night. No exceptions based on mood. The first week is harder. By week three, she stops testing — because the test stopped producing results. Boring, predictable enforcement is what actually changed the game. This is the same principle that drives boundary setting in every ongoing relationship — not the initial statement, but the follow-through.

The lesson: Children don't learn rules from words. They learn them from consequences. So does everyone else.

Case Study: The Couple Who Keep Score

A man and woman have been together four years. He tracks every time he does the dishes and she doesn't. She tracks every emotional load she carries that he avoids. Both feel they're carrying more weight. Both bring it up in fights as evidence. The fights aren't about dishes or feelings — they're about the ledger. And each fight adds a new entry.

Why this fails: Both are running biased emotional accounting. Both are technically right from their own records. The ledger doesn't resolve the conflict — it generates it.

The reset: They stop arguing about what's "fair" in the abstract and start making specific, small agreements: "I'll handle dinners Monday to Thursday, you handle Friday to Sunday." Not because fairness needs to be mechanical — but because vague expectations invite biased accounting. When the terms are explicit, the scorekeeping loses its power.

The lesson: Unspoken expectations are almost always tracked with biased ledgers. Making agreements specific doesn't kill the romance. It kills the resentment.

When to Stop Trying

The Reset assumes the other person is capable of responding to a change in your strategy. Sometimes they're not.

If you've run three or four clear rounds — consistent, proportional, well-communicated — and the pattern hasn't moved, you're probably not dealing with a strategy problem. You're dealing with a structural one: the other person's incentives, capacity, or willingness simply aren't there.

Signs: feedback given clearly multiple times with no change. Apologies that prevent consequences but don't produce different behaviour. Compliance when it suits them, defection when it doesn't. Punishment when you set a limit.

At that point, better strategy won't save the game. Your options narrow to: tighten boundaries, reduce exposure, or leave. The primer covers this in more depth under exit strategy. It's not cynicism. It's arithmetic.

Reflection Prompts

Read slowly. Sit with the ones that produce a reaction.

  1. Which of the five traps — emotional accounting, precedent blindspot, relief addiction, fairness trap, invisible defection — is my default under pressure?
  2. Where am I solving for this interaction instead of the pattern?
  3. What behaviour have I been tolerating that I've accidentally made permanent?
  4. Where am I defecting quietly — agreeing on the surface, doing something different underneath?
  5. What would change if I made my actual position visible to the people around me?
  6. Which relationship am I keeping score in — and what would happen if I replaced the ledger with a specific agreement?
  7. Where has my inconsistency taught people to ignore my words?
  8. What one move this week would improve the next ten interactions?
  9. Which repeated game in my life is no longer fixable — and what's keeping me in it?
  10. Am I solving for comfort or payoff?

The Pattern You're Setting Right Now

Every repeated game you're in is producing a pattern. Some of those patterns you chose. Most of them you allowed — through relief-seeking, precedent blindness, biased accounting, or quiet defection you didn't recognise as defection.

The correction isn't complicated. Pick one game. Run the six-step Reset. Change one move and see what shifts.

The thing about repeated games is that they respond to change. Not immediately — there's always a lag while the other person updates their expectations. But if your new move is clear, consistent, and credible, the game adjusts. It has to. Because the other person is also playing the pattern, not just the moment.

In repeated games, behaviour is never just behaviour. It's also a signal — about what you'll accept, what you'll enforce, and what you're actually committed to. The next post in this series explores that directly: how the signals you send, deliberately and accidentally, build or destroy your reputation, and why signalling and trust are the foundation of every repeated game that actually works.

Game Theory Series Next: Signalling, Reputation, and Trust →

Ready to put this into practice? The capstone of this series gives you a complete decision framework built on every concept covered here.

Read the OM Game Theory Decision Framework

This content is educational and does not constitute psychological, business, or medical advice.