Biases

Rolf Dobelli has compiled a list of cognitive biases in his book “The Art of Clear Thinking”. Kahnemann’s Thinking Fast and Slow also explains many of them, in the voice of the researcher that discovered them in the first place, and Cialdini’s Influence highlights how social biases are leveraged to manipulate us. Being aware of these biases is important for decision making. And because of this, they crop up in just about every management book, with the same stories and experiments told. This is my take and categorization on such a list of biases, together with ways to remember them, and the experiments that established them.

Social Biases

Biases that are created from our need to fit in, on social pressure, and inhibits our ability to think independently and fact based.

Authority
Halo Effect
Liking Bias
Reciprocity

Scarcity Error
Social Loafing
Social Proof

Authority: if an expert in a white lab coat says something, or an expert on television, then it has to be true. No, it has not. From: Cialdini Influences. Experiment: Stanely Milgram Electroshocks, 1961. Related: Halo Effect.

Halo Effect: one dominant attribute influences your perception of others. For example, good looking people are perceived to be more intelligent than they are. If you like someone, you tend to interpret their behaviour more favourably. Related: Liking Bias. 

Liking Bias. The principle of mutuality. The more we like someone, the more we allow us to be influenced by them. Liking is driven by physical attractiveness, similarity in background and attitudes to ourselves, and by feeling we are being liked by them. It has been shown that more than anything they say, looks are a predictor of success in elections. (Where?)  Conversely – nothing works better to make people like you than letting them feel that you like them. Antidote: think of the deal without the person, or with someone you don’t like.

Reciprocity: come in good sir, here take a cup of tea. Let me show you my beautiful carpets. If we get something, we want to give something back. We feel indebted. We do not want to be in debt. Even if that thing is small and insignificant in cost, like a free cup of tee, a flower or other gesture. It makes us willing to reciprocate, often in a much larger an more expensive buy. Beware of vendors bearing gifts. From: Cialdini Influences.


Scarcity Error. rara sunt cara. we find scarce things more attractive, even if the scarcity is created artificially. A form of Social Proof, if others want it too, it must be good. From: Cialdini Influences.

Social Loafing. Teams are lazy, as no on is accountable, and giving your all will mostly go unrewarded, not doing so, undetected. At about 20 people in the group, the effect maxes out. Also called Diffusion of Responsibility. In larger management teams, nobody is responsible for a decision alone, and failure will not hit him in particular. So most will not put their best effort in to think it through, and the group will take larger risks, the so called “risky shift“. Antidote: make individual contributions visible.

Social Proof: all the others are doing it, I must be wrong, even if I do not understand it. This is especially compelling if those others are similar to you. This urge is natural, because you can err, and what everyone does is in many cases an excellent heuristic for what the safest thing to do is. But in investing, and in many other cases this drives bubbles, and doing what everyone does is the worst thing you can do. The reason why you trot out customer stories, lists of customers, and testimonials. From: Cialdini Influences.

Ego Biases

Biases that originate in our need to have a positive self-image, or overvalue our power. To look good in front of others, and ourselves. These are difficult, but can be overcome with humility.

• Action Bias
• Beginner’s Luck
• Cognitive Dissonance
• Confirmation Bias
• Consistency bias
• Control Illusion
• Endowment Effect
• Overconfidence Effect
• Self-serving Bias
• Sunk Cost Fallacy
• Winner’s Curse

Action Bias. We do not get paid for activity, we get paid for being right (Buffet). People rather do something than nothing. You do not look as if you are in charge if you do nothing, even if that is the right strategy, for example for investing. It pays to pass on all the mediocre opportunities and wait for the big ones. Its harder to do though, if you are sitting on a pile of cash and feel you have to act. Also: being busy may satisfy your conscience, in telling yourself you do what you can, but does not mean you achieve something. It’s the good boy’s procrastination. Results count, not your activity. Related: Omission Bias, Status Quo Bias.

Beginner’s Luck. were you only lucky, or are you good? If you at first succeed, you run the risk to overestimate your ability and put too much on the line. Related: Survivorship Bias.

Cognitive Dissonance. Who would want stupid grapes, said the fox, when he could not reach them. Like in Aesop’s fable, instead of being honest with ourselves if we cannot achieve something, or made a mistake, we rather tell ourselves a lie to make us feel better about it. That means we do not take action to achieve it, or repair it. Experiment: Festinger/Carlsmith at Stanford, Students paid less to lie about a boring works attractiveness, found it less boring — presumably because they had to justify having done it to themselves. 

Confirmation Bias: one of the worst and most fundamental biases. And one of the hardest to avoid: if we believe something is right, we become invested in that idea. Instead of looking for possible ways to invalidate it, and thereby harden it when it survives these tests, we look for evidence that supports our idea, and even ignore or reinterpret evidence against it. Antidote: Write down disconfirming evidence, otherwise your brain will forget it. Look for it actively. Have alarm bells go of in your head if you hear something that contradicts your plan or theory labelled as “special case”. Related: Survivorship Bias.

Consistency Bias (also: Hobgoblin's): a foolish consistency is the hobgoblin of little minds (Thoreau). We desire to be internally consistent and stick to positions we took. We do not want to admit to having made a mistake, to ourselves or others. We want to save face. So we continue with an erroneous judgment, instead of admitting to be wrong and doing the right thing. Are you secure enough to admit to having made the mistake? Related: Sunk Cost Fallacy.

Control Illusion: you think your actions influence the outcome more than they do. You also think in hindsight success was due to your effort, while you happily attribute failure to luck.  Antidote: focus on things really under your control.  Anecdotes: placebo buttons on lifts, red lights for crossing the street in NYC, office air conditioners. Related: Self-serving Bias.

Endowment effect. Just take it for a test drive. As soon as you possess something, you do not like to give it up again.  That is why car salesmen want you to drive the car. Why pet stores let you take the puppy home. Also: once we buy and own something, we value it higher. Think trading cards – what you trade away for a card is much less then what you need to give up that card again. This also works against you in auctions – you feel as if it is yours already, and you defend it against other bidders. Antidote: Buddhism or Stoicism. Do not hang your heart onto things.

Overconfidence Effect The tendency to overestimate your own ability to estimate future outcomes. When estimating the outcome of future events, you often underestimate the breadth of possibilities due to a host of related biases. Established: Marc Alpert, Howard Raffia.

Self-Serving Bias. If I win, I’m to credit, if I lose, the circumstances are to blame. People like to believe that success is made by them, but failure is not their fault. It’s natural, you do not want to face that you are to blame, or did not really contribute that much. This is weak, as you do not recognize areas where you could improve. Think Magic. There is no point to blame bad luck. Analyze your game and eradicate your mistakes. Being self-righteous means you never improve. Antidote: good friends, or at least an enemy. Related: Control Illusion.

Sunk Cost Fallacy: I cannot stop now, or I’ll lose all the money and ego I already have invested in this. People have a hard time admitting they made a mistake. You want to be consistent in your behaviour, believable and trustworthy. That is what makes it hard to revert a course of action. You are loss averse and do not want to admit the money is lost, and nothing gained. But it is lost, if you put in more or not. You should only evaluate the current situation, and decide based on it what the right course of action is. Don’t throw more good money after bad. Related: Consistency Bias.

Status Quo Bias. We tend to overvalue the situation as it is. One example is, you wonder if you would move to another city for a job that pays $50,000 more, giving up your social envrionment. But if you picture yourself already being there, would you accept the $50,000 paycut to move back home into your current environment? Antidote: imagine the alternative to be the current situation Related: Endowment Effect.

Intuition Biases

Biases that originate in our intuitive thinking patterns that overvalue visual and story, and undervalue things that require require probability thinking, consideration of invisible options or that need projections of math.

• Alternate Path Blindness
• Anchoring
• Availability Bias
• Base-rate Neglect
• Contrast Effect
• Exponentiality Blindness
• Fundamental Attribution Error
• Gambler’s Fallacy
• Memorability Bias
• Outcome bias
• Omission bias
• Probability Bias
• Self-selection Bias
• Story Bias
• Survivorship bias
• Zero-Risk-Bias

Alternate path blindness. My live was full of calamities, of which most never happened. (Montaigne). If you only look at successful outcomes, you ignore the alternative possibilities that never happened. Russian roulette for a big bet, and diligently working half your life have similar expected value, but vastly different alternate paths. Related: Survivorship Bias.

Anchoring. A number we hear influences our estimates, acting at an anchor. Even if it has nothing to do with the question at hand. This is why negotiators start with a very high or low offer — to anchor our response. Experiment: Kahnemann, Tversky Wheel of Fortune. 

Availability Bias: you tend to consider things that are in front of you, and fail to consider things that are not. We overestimate things that are easy to recall. “Don’t wake the sleeping baby” in Japan is advise to not making the negotiation partner aware about something he is not. Antidote: work with people that are very different, and from different backgrounds, that bring other ideas and perspectives. Related:

Base-Rate-Neglect. humans ignore the a base rate prior probability in absence of evidence,  and overestimate the evidence. Imagine an introverted man with glasses — is he a professor of literature or a truck driver? The image we see fits better to the professor. We ignore that there are thousands of times more truck drivers than professors. Also happens often in medical diagnosis -- the base rate has a huge impact, and a common diagnosis a priori has a much higher chance to be right.

Contrast Effect: you do not well judge things in absolute terms, you tend to judge them in comparison to their environment. Thus Tobi’s Law, that if all the girls at a party are bland, the one least so will appear pretty. Thus Harner’s Alpha/Beta theory, why pretty girls have unattractive best friends. Thus Goldilocks Pricing, putting an extra expensive option up, to make people buy the medium priced one, instead of the cheap one. Related: Anchoring

Exponentiality Blindness. The story of the wise man, who asked for a rice seed on each field of a chess board, doubled. Humans suck at projecting and visualising the effect of sustained exponential growth over longer time periods. Antidote: divide 70 by the annual growth rate: this is the number of years needed to double the amount.

Fundamental Attribution Error. No story without face. We like to tell ourselves stories that explain why something happened, even if there is no simple explanation in reality. And stories need actors. Thus the tendency to overestimate the influence of individuals like presidents and CEOs. Antidote: Look for interests and influences, not people, driving developments.

Gambler’s Fallacy. The roulette ball landed so often on red, it must land on black now. Obviously bunk, the probability of independent events is exactly that, independent of the past. In a fair roulette wheel, the probability for red or black is (ignoring the 0) 50%, no matter how often a color has appeared.

Logical and sequencing errors. Biases that confound cause and correlation.

Correlation for Causation
Association Bias
Conjunction Fallacy
Swimmers Body Illusion
Hindsight Bias
Regression to the mean
Hyperbolic discounting
Induction

Traps and dangers. Biases that do not fit the other categories, and may be based on evolutionary ingrained behavior that is not appropriate to today’s world, or on our overall limited capacity to make decisions or process information, or just on fraud due to lack of consequences for the individual.

Framing
Loss Aversion
Incentive superresponse tendency
It gets worse before it gets better
Tragedy of the commons
Selection Paradox
Prognosis Illusion
Hedonic Treadmill
Fake knowledge

List of Biases with Short Explanations

Survivorship Bias: I’ll look at highly successful results to understand what it takes to achieve success. Looking at successful outcomes gives you an incomplete idea if something works. It ignores that success can also be the result of many people trying something, most failing, some succeeding just by random chance, and those are the one you now examine. Antidote: look for the failures that had the same strategy, or survivors that had another strategy. Anecdote: a priest claims the letters of shipwreck survivors as proof of the existence of god, and then a sceptic asks: where are the letters of those that did not survive? Example: Dropping out of college. Survivorship bias may suggest this is a great strategy to get rich — just look at self-made milliardaires like Jobs and Gates who did this, and even studies that show that more of the millardaires dropped out than of the average person. However, if you look at the overall number of such milliardaires, and compare it to the number of peope that dropped out and never got a decent education and now as their job have to sell fries, you can see that for the average person, this may be a horrible strategy. Related: Confirmation Bias

Swimmers Body Illusion: All swimmers seem to have athletic bodies. Ergo, swimming must be great to get an athletic body. This is a special version of the common mistake of mixing up causation and correlation, the most extreme case of confusing cause end effect. People with a naturally athletic body have a better chance to become successful swimmers, and thus successful swimmers you see all have great bodies. Antidote: ask yourself if you really see causation or just correlation. From: Nassim Nicolas Taleb

It will get worse before it gets better: not really a bias, but a trick that allows consultants of all kind to be always right — if their guidance doesn’t work, they told you so. If it does work, this warning is happily forgotten.

Story Bias: wir merken uns Geschichten, nicht fakten. Und wir merken uns Dinge die eine Begründung haben leichter, als welche die keine haben. Egal ob diese sinnvoll ist.

Hindsight Bias: in hindsight everything seems obvious and easy. We forget, how unclear things were, when we made the decision. Antidote: write down your reckoning and reasons when making the decision, so you can check back.

Fake knowledge: know your circle of competence and stick to it. It may be possible to snow others and fake knowing an area by using soundbites, but it is not possible to make sound decisions that way. Anecdote: Max Planck’s talks on quantum dynamics and his chauffeur.

Incentive Superresponse Tendency. Don’t ask the barber if you need a haircut. People deliver what you incentivize, not what you intend. External incentives lower performance, rather than intrensic incentives of making a good job. And consultants paid by effort will find ways to create unneeded effort. Not really a bias, but a cautionary note.

Regression to the mean. If things are bad, and then get better, or things are good and then get worse, this may be caused by their natural tendency to revert to the mean. You tend to overestimate the influence of your actions on this.

Tragic of the commons. A shared resource without cost is used by each user so heavily that it will be worthless to all. In an anonymous society there is no social shaming as control mechanism to stop selfish agents. Whenever the benefits of something accrue to the individual, but the costs are shared (for example pollution), management and laws are needed. The free market will not fix it.

Outcome bias. Judging decisions by their result, not by the decision making process. You can win a game of magic with luck, even if you made play mistakes. You can lose it even if you play flawlessly. You can succeed in stock picking by pure luck. Don’t judge someone purely based on the result. (Of course, it is hard to objectively judge the decision making process, and easy to see the result).

Selection Paradox. It is harder to decide on something the more options you have. Too many options lead to decision overload, you cannot evaluate them all. In the end you either take no decision, to not take the wrong one, or take one and are dissatisfied because you suspect it was the wrong one. Experiment: Barry Schwartz, Marmalades – customers bought more when there were fewer to choose from.

Memorability Bias. All the lights are red when I drive. We remember annoying things, and forget about all the cases where the lights were green. We have a tendency to igore prior probabilities. If something improbable happens, we tend to ignore the many times the probable happened. 

Neglect of probability: humans tend to ignore probability in assessing the impact of a risk, and only look at severity. Severity is easy to picture, probability isn’t. Experiment: Electroshock experiment — people were as afraid of the shock, wether there was a 100%, 50% or 5% chance to get a shock. Only the strength of the shock influenced the fear level. Only exception: at 0% chance, people were relaxed. 

Zero-Risk-Bias. We overvalue a risk going to zero. Experiment: People preferred a risk going from 1% probability to 0% over 5% to 2%, even though in absolute terms, the latter is a three times larger reduction. 

Winner’s Curse. In auctions, Endowment Effect and Scarcity Error combine powerfully to the bidders disadvantage — the party that ends up winning is often is paying too much, more than the good is worth. If you sell something, like a house or car, the best way to get a good price is to have multiple bidders, and run them against each other. Antidote: Buffett recommends to just not bid in auctions. 

Induction. Guessing from past experience at future events. This is actually highly rational but still may go very wrong. Imagine a geese being fattened for Christmas — every day she is fed, life is good. Until that fateful day when suddenly, she instead is killed. Antidote: think about what is driving the observed behavior, and if that still holds. 

Loss aversion. In the past, a mistake often was deadly. So we are primed to be more careful about risks than greedy for benefits. You can easily survive without eating honey (or anything) for quite a while, but one mistaken encounter with a bear, and you are history. We experience losses about twice as strongly as equivalent wins. Argue with avoiding a possible loss, not with a possible win, if you want to convince. Most employees shy away from taking risks, because if they win they may get a small bonus, but if they fail they might lose their job.

Taking correlation for causality. Married people live longer. Ergo, marriage is good for living longer. Or is it maybe, that healthy people have better chances of finding a partner? This is one of the most fundamental and common ones. Antidote: check if you really have data that indicate a cause-effect relationship. 

Halo-Effect. People do judge the book by its cover. Looking only at the most visible attribute, and letting our judgment of all others be influenced by it. We think beautiful people are smarter.

Prognosis Illusion. There are two kinds of people – those that know nothing and those that know nothing and do not know that they know nothing (Galbraith).Experiment: Tetlock showed by reviewing more than 80,000 prognoses, that the prognoses of experts do not differ much from random prediction.

Conjunction Fallacy. We attribute higher probability to stories that seem to “make sense”, even though they logically have to be less probable. For example, Experiment: Kahnemann & Tverski: people asked what was more probable “Oil consumption drops 30%” or “Oil consumption drops 30% due to a steep rise in oil price” preferred the latter prediction as more probable, even though the former one included it, among many other possible reasons.

Framing. It’s not the song, it’s the singer. Depending how the facts are presented, they are seen differently. This is different from Anchoring, as it is not dependent on a context. Famous Experiment: Kahnemann & Tverski: people preferred saving 200 people out of 600 over a 1/3 chance to save all, and at the same time preferred a 1/3 chance to kill no one over killing 400 out of 600. Mathematically they are all equivalent, but we prefer sure gains and are Loss Averse. As we inherently are naturals in imagining probabilities, the probability based versions seem less substantial in either case.

Omission Bias. We tend to overvalue errors of commission, compared to errors of omission. Putting money in the wrong stock feels worse than not putting money in a great stock. Killing someone feels worse than not saving someone from being killed. Related: Action Bias. The difference is that for action bias the situation is less clear. Here it is often clear what one should do, but does not feel so urgent.

Hedonic Treadmill. Buying things only makes us happy for the short term. Soon they become the new normal, and we are back to where we started. True happiness or being content only comes from within, from the ability to be grateful, to enjoy the small things. It does not come from amassing things. Experiment: Dan Gilbert investigated self-reported happiness of people after they won in lotto or lost the use or their legs in an accident. A few months later, they all were back to their base levels. The winners sometimes even less happy. Antidote: avoid chronic negative experiences like noise, long commutes, stress. Don’t expect happiness from things. Be in control of your time, autonomous. Care for others.

Self-Selection Bias. See above – the feeling that you are somehow specially selected or unlucky, in the slow lane all the time. It’s just because only then you start to think about it. If you are in the other group, you do not mind.

Association Bias. Shoot the messenger of bad news. We tend to associate things that appear together, even if they have no causal relation. We are superstitious, and wear our lucky pants.  Experiment: Pawlow’s dog.

Hyperbolic Discounting.Leider geil. We have a hard time resisting instant gratification for something that pays off more in the future. That’s why we party and watch youtube videos, instead of studying or getting our work done. We are more willing to forgo money in a year to get a higher payout in two, than willing to forgo money now to get a higher payout a year from now. Experiment: Walter Mischel, children who had the choice of eating a sweet now, or get two if they wait. The ones who waited, had better career success.