The Big Biases II: Loss Aversion, Sunk Cost, Confirmation & Overconfidence

By Pritesh Yadav June 22, 2026 12 min read —

In the last chapter we met biases that distort how we judge the world — how likely something is, how big a number should be. This chapter is about a different and more painful family of errors: the ones that make us hold on too long, defend bad ideas, and overrate ourselves. These are the biases behind the failing project nobody will cancel, the argument where neither side ever changes its mind, and the plan that always runs late. Understanding them is genuinely life-changing, because they shape your money, your relationships, and your work more than almost anything else in psychology.

Before we start, two quick definitions that everything else hangs on.

Bias: A systematic error in thinking — meaning it happens in a predictable direction, again and again, not just randomly. It is not stupidity. Smart, well-informed, expert people make these errors too. They are side effects of normal, efficient thinking.
Loss aversion: The discovery that losses hurt about twice as much as equivalent gains feel good. Losing $100 stings roughly as much as winning $200 pleases. This single fact is the engine underneath several of the biases in this chapter.

7.1 Loss aversion: why losing feels worse than winning feels good

Imagine I offer you a coin flip. Heads, you win $100. Tails, you lose $100. The "expected value" is exactly zero — it's a fair bet. Yet almost nobody takes it. Studies show most people only accept the flip when the win is around $200 or more against a possible $100 loss. The math says the bet is even; your gut says the possible loss is roughly twice as heavy as the possible win.

That ratio — losses weigh about 2× (often written as λ ≈ 2) — was measured by psychologists Daniel Kahneman and Amos Tversky. It explains an enormous amount of human behavior, much of it stuff that looks irrational until you see the loss hiding in it.

Analogy: Loss aversion is like a smoke detector for your wallet and ego. A smoke detector is deliberately tuned to scream loudly at the faintest hint of smoke, because in our ancestors' world, missing a real fire was fatal while a false alarm was just annoying. Your mind is tuned the same way about losses — losing a resource could once mean starving, so the alarm for losses is wired louder than the reward bell for gains.

Example: Free trials are pure loss aversion. Before the trial, signing up for a paid service feels like spending money — a loss you resist. But once you've used the service free for 30 days, it feels like yours. Cancelling now feels like losing something you have. Companies know that flipping you from "thinking about paying" to "afraid of losing access" multiplies how many people stay.

You'll also recognize loss aversion in: gym memberships you keep paying for so you don't "waste" them, the difficulty of selling an old car for less than you "feel" it's worth, and money-back guarantees (returning an item means admitting a loss, so most people just keep it).

7.2 The sunk cost fallacy: throwing good money after bad

A sunk cost is money, time, or effort you've already spent and can never get back. The sunk cost fallacy is letting those unrecoverable costs drive your future decisions — continuing something only because you've already invested in it, not because it's still worth doing.

Example: You buy a $15 cinema ticket. Twenty minutes in, the film is dreadful. The logical move is to leave and reclaim your evening — the $15 is gone either way. But most people sit through the whole thing "because I paid for it." Staying doesn't bring back the $15; it just adds a wasted two hours on top of wasted money.

The same trap keeps people in failing relationships ("we've been together five years"), dying business projects ("we've sunk $2 million into this"), and dead-end careers ("I spent six years training for this"). In one famous study, people were asked about a $10 million project that was 90% finished but had been made pointless by a competitor — about 85% chose to finish it anyway, even though only about 17% would have started a fresh project with the same hopeless outlook.

Why does this happen? Largely because of loss aversion. Walking away forces you to officially write off the investment as a loss — and losses hurt. Continuing lets you pretend the loss hasn't happened yet. There's also a deep human desire to look consistent and to avoid "wasting" things.

Tip: The cure for sunk cost is one clean question: "Knowing what I know now, if I weren't already in this, would I choose to start it today?" If the answer is no, the money or time already spent is irrelevant — it's gone no matter what you do next. Decide only on the future.

Common mistake: People think "not quitting" is always a virtue — grit, persistence, never giving up. But grit applied to a genuinely lost cause is just the sunk cost fallacy wearing a motivational T-shirt. Persistence is admirable when the future payoff is real; it's a trap when you're only continuing because of the past.

7.3 Confirmation bias: hearing only what we already believe

Confirmation bias is the tendency to seek out, notice, and remember information that supports what we already believe — while ignoring, dismissing, or forgetting anything that contradicts it. It is arguably the most consequential bias of all, because it quietly poisons how we learn anything.

The classic demonstration is Peter Wason's "2-4-6 task." People were told the sequence 2-4-6 follows a hidden rule, and they could test their guesses by proposing new triples. Almost everyone guessed the rule was "add 2 each time" and then tested only sequences that fit it — 8-10-12, 20-22-24 — getting a "yes" each time and feeling more and more sure. The real rule was simply "any three increasing numbers." Hardly anyone thought to test a sequence designed to prove themselves wrong, like 1-2-3 or 5-50-500. That refusal to look for disconfirming evidence is confirmation bias in its purest form.

Example: An investor who owns a stock reads only the bullish articles about it and waves away the bearish ones as "noise." A manager convinced her strategy is right notices every data point that supports it and explains away the rest. A person scrolling social media is fed an endless stream that already agrees with them — the "echo chamber." None of these people feel biased; each feels like they're simply seeing the truth.

Analogy: Confirmation bias is like a lawyer who has already decided you're guilty and now only collects evidence for the prosecution. A good scientist does the opposite — actively hunts for the evidence that would prove their own theory wrong. Most of us, most of the time, behave like the biased lawyer about our own beliefs.

7.4 Hindsight bias: the "I knew it all along" trap

Hindsight bias is the tendency, once you know how something turned out, to believe you "knew it all along" — to feel the outcome was far more predictable than it actually was beforehand. In a classic study, people read about an obscure historical battle; whichever ending they were told was the real one, they rated that ending as having been the obviously likely result the whole time.

You hear this constantly: "The 2008 financial crash was obvious." "Of course that startup failed." It almost never felt obvious before it happened. What's going on is that your memory quietly rewrites itself to fit the known result, so your earlier uncertainty vanishes.

Common mistake: Hindsight bias makes us unfair to decision-makers. We judge a doctor, manager, or coach harshly for a "foreseeable" mistake that wasn't foreseeable at the time. This blurs into outcome bias — judging a decision by how it turned out rather than by whether it was sensible given what was known. A good decision can have a bad outcome (you can play a poker hand perfectly and still lose), and a reckless decision can get lucky.

Tip: Keep a decision journal. Before you make a real choice, write down what you predict and why, in plain words. Later, compare it to what actually happened. This is the single most powerful tool against hindsight bias, because it freezes your real uncertainty in writing so your memory can't quietly edit it.

7.5 Overconfidence and the Dunning–Kruger effect

The overconfidence effect is simply having more confidence in your judgments and abilities than your actual accuracy justifies. When experienced managers are asked to give ranges they're "90% sure" contain the right answer, the true value falls inside their range only about 40–60% of the time — not 90%. We are confidently, reliably too sure.

The most famous version is the Dunning–Kruger effect: the people who are worst at a skill tend to most overestimate themselves. In tests of humor, grammar, and logic, the bottom performers (around the 12th percentile) guessed they were around the 60th — well above average. The reason is a cruel double bind: the very knowledge you'd need to be good at something is the same knowledge you'd need to realize you're bad at it. If you can't spot good grammar, you also can't spot your own grammar errors.

Common mistake: The popular cartoon of Dunning–Kruger — a "Mount Stupid" where idiots think they're geniuses — overstates it badly. The real effect is gentler, and part of it is just a statistical pattern (very low scorers have nowhere to guess but upward). Importantly, even top performers tend to overestimate themselves in absolute terms, and they often underrate how they compare to others (assuming "if it's easy for me, it's easy for everyone"). The honest lesson isn't "dumb people are deluded" — it's "almost everyone is a poor judge of their own ability."

7.6 Optimism bias and the planning fallacy

Optimism bias is overestimating good outcomes and underestimating bad ones for yourself specifically. Most people rate their own risk of divorce, cancer, or a car accident as below average — which is statistically impossible for most people to be true. Smokers underestimate their personal risk; new business owners rate their survival odds high even though most new businesses fail.

The most practical offspring of optimism bias is the planning fallacy: we underestimate how long things will take and how much they'll cost — even when we know similar past projects ran over. Students predicting their thesis would take 27 days actually took around 56 days; only about 30% finished within their own predicted time. At grand scale, the Sydney Opera House was estimated at A$7 million and finished at A$102 million — roughly 1,400% over budget and a decade late. A study of 258 transport projects across 20 countries found about 90% ran over budget.

Analogy: Estimating from inside your own plan is like judging a road trip by staring at the map and imagining a perfect drive — no traffic, no rain, no wrong turns, no bathroom stops. The map ("inside view") only shows the best case. The honest estimate comes from the outside view: asking "how long did trips like this actually take last time?" Reality always includes the surprises you can't see in advance.

Tip — reference-class forecasting: To beat the planning fallacy, don't ask "how long will my project take?" Ask "how long did a bunch of similar projects actually take?" — then base your estimate on that real distribution, and add a buffer. This "outside view" method is so reliable it's officially recommended in government project guidelines.

How to apply this — practical debiasing

Knowing about a bias rarely makes it disappear; these errors persist even in people who study them. What actually helps are processes that force slower, more honest thinking:

For sunk cost: ask "Would I start this today, from zero?" Ignore what's already spent.
For confirmation bias: deliberately "consider the opposite." Actively hunt for the evidence that would prove you wrong. Appoint a "devil's advocate" or red team to attack the plan.
For hindsight bias: keep a decision journal; run fair post-mortems that judge the process, not just the result.
For overconfidence and the planning fallacy: run a premortem — before starting, imagine it's a year later and the project failed completely, then write down all the reasons why. This frees people to voice doubts they'd otherwise suppress, and surfaces risks early.
For optimism bias: always check the base rate — the outside view — before trusting your gut about your own odds.

Key takeaway: Loss aversion, sunk cost, confirmation bias, hindsight bias, and overconfidence aren't signs of a weak mind — they're built-in features of a normal one. They share a common thread: we cling to what we have, what we've spent, and what we already believe, and we trust our own judgment far more than it deserves. You can't switch these instincts off, but you can outsmart them with simple habits — ask "would I start today?", actively look for evidence you're wrong, write down predictions before outcomes, and estimate from how similar things really went, not from your hopeful plan.

Continue reading

🎯 Psychology of Decisions

What Psychology Is, and Why "Rational" People Decide Irrationally

🎯 Psychology of Decisions

Two Minds in One Head: Fast Thinking vs Slow Thinking

🎯 Psychology of Decisions

How the Brain Actually Decides: Emotion, Memory & Attention