Game Theory, Bias-Proofing, and How to Judge Your Own Judgment

Picture two rival coffee shops on the same street. Both would earn more by keeping prices normal, yet both slash prices anyway and end up worse off. Neither owner is stupid. They are trapped by logic. This is the part of decision-making where your smartest move depends on what someone else does, where your own brain quietly sabotages you, and where the only way to win is to build a process that is smarter than you are.

You may already know that a good decision is not the same as a good outcome. Here we pick up there and push into the deep end: deciding well when other people are deciding too, when your mind works against you, and when you have to judge your own judgment over a lifetime.

Why this matters

Most expensive life mistakes are not knowledge failures. They are process failures.

You did not lose money because you lacked a fact. You lost it because you chased a sunk cost, anchored on the wrong number, trusted your gut in the one situation that punished gut feeling, or judged a decision by how it happened to turn out.

The good news hidden in that sentence: process is fixable, and it transfers. The same loop that helps you negotiate a salary helps you choose a treatment, launch a product, or decide whether to quit. Master it once and you own something rarer than intelligence: durable, improvable judgment.

Three questions sit at the heart of it:

1. What do I do when the right move depends on what someone else does? That is game theory.
2. Knowing my brain is biased, how do I build a process that defends against it? That is bias-proofing.
3. How do I get measurably better at judgment over years, not just feel better? That is calibration.

We will take them in that order.

When your best move depends on theirs: game theory

So far you may have treated the world as something you bet against: a coin, a market, a disease. But many of life’s biggest decisions are made against other thinking people who are also trying to win. A competitor sets a price knowing you will react. A negotiator hides their bottom line. Two drivers approach a merge.

Game theory is the study of decisions where your best choice depends on what others choose, and theirs depends on yours. “Game” just means a situation of interdependent choice. It does not mean trivial or fun.

Think of rock-paper-scissors. There is no best move you can pick in a vacuum. The whole problem is that the other person is reasoning about you while you reason about them.

The easy case: a dominant strategy

A dominant strategy is a move that is best for you no matter what the other side does. When you have one, stop overthinking and take it.

Wearing a seatbelt is the everyday version. You are better off whether or not you crash. You do not need to predict the road, because the move wins in every scenario.

Dominant strategies are wonderful because they shrink a hard interdependent problem down to a simple one. The trouble starts when the dominant move for each person leads everyone somewhere bad. That is the most famous result in the field.

The Prisoner’s Dilemma

Back to our two coffee shops. Each must decide: keep prices normal (call it “cooperate”) or slash them to steal the other’s customers (call it “defect”).

	B keeps prices normal	B cuts prices
A keeps normal	Both do well (+5, +5)	A loses, B wins (-2, +8)
A cuts prices	A wins, B loses (+8, -2)	Price war (+1, +1)

Look at it from Shop A’s seat. If B keeps prices normal, A earns +8 by cutting versus +5 by staying. Cutting wins. If B cuts, A earns +1 by cutting versus -2 by staying. Cutting still wins. So cutting is A’s dominant strategy, and by the same logic it is B’s too. Both cut. Both land on (+1, +1), worse than if both had simply stayed normal at (+5, +5).

That bottom-right box is a Nash equilibrium: a state where no single player can do better by changing only their own move. It is stable. Nobody wants to be the one to break ranks.

Here is the lesson that trips up almost everyone.

A Nash equilibrium is not “the best outcome.” It is the stable outcome, the one nobody can escape alone. The Prisoner’s Dilemma proves a stable outcome can be worse for everyone than an unstable one.

Price wars, arms races, and overfishing are all real-world equilibria that are stable and collectively terrible. Once you can spot a Prisoner’s Dilemma, you can sometimes change the game instead of just playing it badly, which leads to the most hopeful result in the field.

Repeated games and Tit-for-Tat

The grim logic above assumes the game is played once. Most real relationships repeat: suppliers, colleagues, customers, neighbors. When the game repeats, cooperation can survive, because today’s betrayal can be punished tomorrow.

In 1980 the political scientist Robert Axelrod ran a famous tournament. He invited experts to submit computer strategies to play the repeated Prisoner’s Dilemma against each other over thousands of rounds. The winner was almost embarrassingly simple. It was called Tit-for-Tat: cooperate on the first move, then just copy whatever the other player did last time.

Be nice first. Punish a betrayal once. Forgive the moment they cooperate again.

The strongest strategies shared four traits:

• Nice - never defect first. Extend trust at the start of a relationship.
• Retaliatory - hit back immediately when cheated. Do not be a pushover.
• Forgiving - return to cooperation as soon as they do. Do not poison a recoverable relationship with a grudge.
• Clear - be easy to read, so partners can learn to trust you.

A small business with a long-term supplier is Tit-for-Tat in a suit and tie. Pay on time and be fair (nice). If they shortchange you on a shipment, raise it firmly and adjust terms (retaliatory). When they fix it, return to the normal relationship rather than punishing forever (forgiving). Keep your own dealings consistent so they know what to expect (clear). This is exactly why reputation is one of the most valuable assets in business.

The single best way to escape a Prisoner’s Dilemma is to make the game repeat.

A quick game-theory checklist for negotiations

• Is this one-shot or repeated? Your generosity should scale with how often you will meet this person again.
• Find your dominant move first. If a choice is best regardless of their response, you have saved yourself a lot of mind-reading.
• Decide your walk-away number before the talk. The first number named drags the final price toward it, so set your anchor in advance and their opening offer cannot reset your reference point.
• Change the payoffs, not just your play. Contracts, escrow, warranties, and public commitments all turn “defect” from tempting into costly. That is how you upgrade a bad equilibrium.

How you really choose: prospect theory

There is a textbook ideal of rational choice called Expected Utility Theory. It says that if your preferences are internally consistent, you behave as if you maximize expected utility and nobody can trick you into a money-losing loop of trades. It is a beautiful description of how we should decide.

It is also a poor description of how we do decide. Real humans break it in predictable ways. Two classic puzzles, the Allais paradox and the Ellsberg paradox, showed that people reliably prefer certainty and known odds in ways no single consistent utility function allows. We dislike not knowing the probabilities themselves, not just disliking risk.

So Daniel Kahneman and Amos Tversky built a model of how we actually choose, called prospect theory. It rests on three pillars.

1. We feel changes, not absolute wealth. A €60,000 salary feels like a triumph to someone who earned €40,000 last year and a punishment to someone who earned €90,000. Same number, opposite feeling. Your reference point is everything.
2. Losses hurt about twice as much as equal gains feel good. This is loss aversion, and it quietly drives a huge share of bad financial decisions.
3. We flip our risk attitude. We play it safe to protect a gain (take the sure win) but turn into gamblers to avoid a loss (roll the dice rather than accept a sure hit).

Watch an investor holding a stock that is down 30%. Selling means locking in a sure loss, which feels unbearable, so they hold and even buy more, gambling to “get back to even.” The same investor sells their winners too early to lock in a sure gain. Risk-seeking in the loss zone, risk-averse in the gain zone. It is the engine behind panic-selling, refusing to sell a losing house, and doubling down at the casino.

Why we buy lottery tickets and insurance

Prospect theory adds one more twist. We do not treat probabilities at face value. We overweight tiny probabilities and underweight moderate-to-large ones.

That is how the same person buys a lottery ticket (overweighting a near-zero chance of winning) and home insurance (overweighting a near-zero chance of fire). To a cold expected-value machine those look contradictory. A tiny probability gets blown up in our minds, so we both chase the jackpot and dread the catastrophe.

Knowing your own default settings is the first step to overriding them.

Common misconceptions

• “A Nash equilibrium is the best result.” No. It is the stable result. Stable and terrible coexist constantly.
• “Smart or experienced people are immune to bias.” Experts are often more overconfident, not less. Credentials do not switch off the illusion.
• “Learning the names of biases protects you.” It barely helps. You can know two lines are the same length and still see one as longer. The fix is process, not awareness.
• “A rational agent would do X, so people will do X.” Expected Utility Theory predicts the idealized agent, not your customers, colleagues, or you. Never use it to forecast real behavior.
• “Intuition is the enemy.” Not always. Gerd Gigerenzer’s research shows simple “fast and frugal” rules can beat complex models when information is scarce and time is short. The skill is knowing which situations reward a snap judgment and which demand slow analysis.

The quiet biases that wreck big decisions

The most expensive biases are the self-flattering ones you never notice. Watch for these:

• Confirmation bias - hunting for evidence that supports what you already believe. The cure is to ask, on purpose, “what would change my mind?” and go find it.
• Sunk cost fallacy - throwing more money or time after a failing project because of what you already spent. The spent resources are gone either way.
• Overconfidence - being more certain than your accuracy warrants, especially giving one number when you should give a range.
• Planning fallacy - underestimating how long things take, how much they cost, and what could go wrong.
• Hindsight bias - “I knew it all along.” After an event we rewrite our memory, which destroys our ability to learn.
• Resulting - judging the quality of a decision by the quality of its outcome. Annie Duke’s term, and maybe the most important reframe in the whole discipline.

Here is the sunk cost test in action. You have spent two years and your savings building a product nobody is buying. The honest question is not “look how much I have invested.” It is: “Knowing what I know now, would I start this project today?” If the answer is no, continuing is throwing good money after bad. The two years are already gone whichever way you choose.

How to use this: bias-proofing your process

Since awareness barely helps, you engineer the environment instead. Each tool below targets a specific bias. This is the “what to actually do Monday morning” layer.

1. Run a premortem

Before you act, imagine it is a year from now and the decision has already failed spectacularly. Then ask everyone: “Why did it fail?” You are writing the autopsy before the death, while you can still prevent it.

A wedding planner does this a month out: “Picture the big day as a total disaster. What happened?” Suddenly someone realizes the caterer was never actually confirmed. The premortem licenses people to voice doubts they would otherwise swallow, and it inverts optimism: instead of “will this work?” (which invites cheerleading) you ask “how did this die?” (which invites honesty).

2. Use a checklist

A checklist is a short, pre-written list of must-check items for a recurring high-stakes decision. Not because you are forgetful, but because everyone’s memory fails under stress, fatigue, and time pressure.

Pilots run one before every takeoff. Surgeons run one before every incision. These are among the most skilled people alive, and they still use a piece of paper, because the cost of forgetting one obvious step is catastrophic. Your version: a hiring checklist, a “before we sign” checklist, a launch checklist.

3. Sort one-way doors from two-way doors

Not every decision deserves the same effort. Jeff Bezos’s framing: some doors are one-way (hard or impossible to undo) and some are two-way (easy to walk back).

	One-way door	Two-way door
Reversible?	No	Yes
Examples	Selling the company, a risky surgery, quitting with no backup	A new pricing page, testing a tool, a reversible hire
How to decide	Slowly, with premortem and checklist	Fast. Bias toward action and learn by doing

The classic mistake is spending weeks agonizing over a two-way door you could simply try and reverse, while rushing a one-way door because you got impatient. Ask “can I undo this?” before you ask “is it right?“

4. Beat groupthink

Teams have a special failure mode: everyone anchors on the first loud voice (often the boss), doubts go unspoken, and the group talks itself into a consensus nobody truly holds. Three fixes:

1. Anonymous estimates first. Have everyone write their number or vote privately, before discussion, so the room cannot echo the first speaker.
2. Assign a devil’s advocate. Make it someone’s explicit job to argue the opposing case, so dissent is a duty rather than an act of courage.
3. Disagree and commit. Air the disagreement fully, then unite behind the decision so the team does not relitigate forever.

Make the bias-resistant version the default, not a special effort. Pre-commit your criteria before you see the options, so you cannot move the goalposts to favor the choice you already like. A good default protects you on the days your discipline is low.

Getting sharper every year: calibration

Everything above improves a single decision. The crown jewel is getting measurably better over a lifetime. That requires turning vague confidence into numbers you can score.

Calibration is the match between your stated confidence and your real hit rate. You are well-calibrated if, across all the times you said “70% sure,” you turned out right about 70% of the time.

A weather forecaster who says “70% chance of rain” and is right on roughly 7 of every 10 such days is genuinely useful. A confident pundit who states everything as a certainty but is right half the time is not, even though he sounds more authoritative. Calibration, not confidence, is the real signal.

Keeping honest score

The Brier score is a single number that measures how good your probability forecasts are. Lower is better: 0 is perfect, 0.5 is no better than a coin flip. It rewards being both accurate and appropriately confident.

Philip Tetlock’s Good Judgment Project ran the landmark experiment, having thousands of trained volunteers forecast real geopolitical events. The headline finding is the most encouraging result in the field:

• Calibration is learnable. The best volunteers, the “superforecasters,” hit Brier scores around 0.20 to 0.25 and routinely out-forecast professional intelligence analysts who had classified material.
• They were not geniuses with secret data. They had better habits.

What did they actually do?

1. Start from the base rate, then adjust. Before predicting “will this startup succeed?” they note that most startups fail, and only then move toward the specifics. Base rate first, evidence second.
2. Update in small steps. Many small revisions as new information arrives beat both stubbornness and flip-flopping on every headline.
3. Think in fine probabilities (“23%,” not “probably not”). Precision forces clear thinking and makes scoring possible.
4. Break big questions into smaller ones they can actually estimate, then recombine.

The highest-leverage habit: a decision journal

You cannot improve a process you cannot see. A decision journal makes your decisions visible and scoreable, and it defeats the two biases that otherwise make learning impossible: hindsight bias (“I knew it all along”) and resulting (“it worked out, so it was a great call”).

For every meaningful decision, write down, at the time, before the outcome is known:

1. What you decided.
2. Why, including your key assumptions.
3. Your probability estimate (“65% chance of working”).
4. What you expect to happen, and by when.
5. How you felt (tired, rushed, emotional all predict bad decisions).

Then review later. Now you can ask the only fair question. Say you bet on a project you gave a 70% chance, and it failed. Resulting says “bad decision.” The journal asks: was 70% a well-reasoned estimate given what you knew at the time? If yes, this was a good decision with a bad outcome. 30% things happen, and punishing yourself for them only teaches you to be a coward. The reverse is just as important: a reckless call that happened to win is a bad decision with a lucky outcome, and celebrating it teaches you to gamble.

Review the journal quarterly. Most people never learn from experience because they never wrote down what they actually expected.

Your personal decision operating system

None of this is a collection of party tricks. It is one repeatable loop you can run on every important choice:

1. Frame. Is this reversible? Match effort to stakes.
2. Outside view. Start from the base rate. What usually happens to choices like this?
3. Think in bets. Give a probability and a range, not a certainty.
4. Who else? Is anyone else deciding? Map the game. One-shot or repeated?
5. Bias-proof. Premortem, checklist, anonymous estimates, reframe gain versus loss.
6. Journal. Record the decision, reasoning, and odds before the outcome.
7. Review and score. Judge the decision, not the outcome. Update your calibration.

Then it feeds back: every reviewed decision sharpens the base rates you bring to the next one. Steps 2 and 3 are about being right. Step 5 admits you are biased and engineers around it. Steps 1, 4, 6, and 7 are about doing better anyway.

Conclusion

If you forget every formula here, keep one sentence: a decision is not its outcome. You make decisions with foresight under uncertainty, and outcomes are judged in hindsight. Judge the process, because the process is the only thing you control, and a good 70% bet that loses was still a good bet.

You cannot command luck, other players, or the future. You can command your framing, your base rates, your bias-proofing, and your journal. Build the operating system, run it on your next real decision, and let good outcomes accumulate as the long-run reward of good decisions.

One question is worth sitting with: if good judgment is this trainable, why do so few people ever get better at it? The answer is not intelligence. It is that almost nobody keeps score. The day you start writing your reasoning down before the result arrives is the day your experience finally begins to teach you something.