Book Summary: Think like a freak by Steven Levitt and Stephen Dubner

The book is an impressive collection of unusual stories aimed at promoting a non-conventional way of approaching problems and solving them.

What does it mean to think like a freak?

Saving one’s own reputation takes priority over maximizing common good. For example, during a penalty shootout, a goalkeeper will jump towards either side with 98% probability, so, it’s better for the soccer player to kick right into the goalkeeper. But the 2% situation will be embarrassing, the player would be considered lousy for not trying hard enough to kick towards the corner. But if he kicks towards a corner and the keeper catches it, he would be still appreciated for a good attempt. Most players prefer saving their reputation by kicking on sides rather than dead center. The author does mention that other reasons for not kicking straight into the center are that it’s counter-intuitive and the keeper will adapt, but he argues that fear of shame for not trying hard enough is the most important one.

Sometimes in life, going straight up in the middle (taking a traditional route) is the boldest option, except the person would be considered lousy for not trying hard.

Married people are happy, not because marriage causes happiness but because happy people are more likely to be married (who would want to marry a grumpy person?).

The three hardest words in The English language

Smart people/experts suffer from dogmatism and find it difficult to admit that “they don’t know”. In 1998, Krugman claimed, the Internet’s impact on the economy will be no greater than fax machines. Another problem which experts suffer from is ultracrepidarianism, which is the habit of giving advice outside of one’s domain. One problem with saying “I don’t know” is that the private cost of that is much higher than making wrong predictions. For example, no one was punished for making the incorrect claim about Iraq’s WMD.

Saying “I don’t know” is the equivalent of not playing a game where downsides are minimal. Very few people go back and point out incorrect predictions, and the upsides are high since the predictor will be celebrated if the prediction is correct. Moral compass just like a real compass can sometimes give incorrect readings, it’s important to set it aside from time to time, or else, the moral compass can lead to incorrect solutions/conclusions.

Once admitted, “I don’t know”, the key to learning is experimentation and feedback. Doing experiments would imply that a lack of theoretical understanding which experts do not want to admit. It’s important to do experiments in natural settings. Unlike hard sciences, the objects (humans) can behave differently in lab settings. Robin Goldstein did multiple experiments to prove that more expensive wines do not taste better than the cheaper ones.

What’s your problem

It’s important to phrase the problem correctly, the way one tries to find its solution relies heavily on that. If the problem is framed incorrectly, one might end up attacking the noisy part of the problem. Most education reforms try to focus on improving schools while parental influence has been determined to be a stronger factor.

“How can I eat more hot dogs?” is incorrect phrasing while “How can I make hot dogs easier to eat?” is better phrasing. Better phrasing leads Kobi, a Japanese guy to figure out that eating dog (by splitting it into two) and bun separately and wetting bun in the water while eating the dog was a successful strategy using which he broke the previous record of 25.125 and made a new record of 50. He figured out these techniques through repeated experimentation and feedback by videotaping his sessions. He also refused to accept the current limit of 25.125  as a real barrier.
It’s important not to limit oneself with artificial barriers.

Like a bad dye job, the truth is in the roots

We usually look for the proximate cause of the problem and most of the times, it’s correct. Except, for complicated problems, the root cause is usually more complicated. For example, legalization of abortions in the 1970s caused a drop in crime rate in 1990s since fewer children were being brought up in undesirable environments. In Germany, Protestants make higher money than Catholics since they work more hours, more likely to be self-employed and their women are more likely to be working full-time. African-Americans in the USA have a higher rate of heart diseases since they contain a higher percentage of salt (since slaves selected from Africa have a higher chance of surviving passage to the USA in ships) – the evidence is still being debated. Ulcer treatment market ($8B in 1994) was satisfied with treating symptoms of ulcer rather than showing interest in solving the cause. The conventional wisdom was the no one bacteria can survive in the stomach. The researchers were ignoring the counter-evidence in front of their eyes. Barry Marshall challenged the status quo, isolated the bacteria and was able to cause ulcers to himself by swallowing the bacteria (Helicobacter pylori). Now, it is accepted that gut bacteria have huge implications on the health and fecal transplant is used for curing intestinal infections.

Think like a child

Small questions are usually less explored and hence, are virgin territories of exploration. It’s easy to have more certainty of the prediction when aiming for smaller problems, triggering changes to study behavior in case of smaller problems is much easier. Raw talent is overrated, the best performers are the one who enjoys the job and improves due to relentless practice. People spend more time doing fun activities like spending, buying lottery tickets, and not boring things like saving. Prize-linked-savings makes the whole process of savings fun by adding a lottery component to it. Having fun, looking at small things and not fearing the obvious are all childlike behaviors which have been frowned upon in adults.

Like giving candy to a baby (incentives)

Humans respond to incentives. Different forms of incentives are financial, moral, social and herd mentality. Though people hate admitting, herd-mentality works the best.
A financial incentive can produce cobra effect. Appealing to moral outrage (“our forests are being destroyed by the theft of petrified wood ~14 tons a year,  mostly small pieces at a time”) encourages people to do something since everyone else is doing it too, I should get my share too (private benefit prioritized over communal good). Brian Mullaney, the founder of Smile Train (fixing cleft lip at the time of birth in developing countries), gave an option to donors that if they make one donation, they can opt for never being asked again for another donation. Most didn’t choose that option, but they preferred the fact they were given an option to opt-out, and they preferred to remained opt-in, and those who opted out saved costs of future mailing for the NGO.
Frames of relationships – financial (buyer and seller), us-vs-them (sports, politics), loved-one (in friends and family, when things are going smoothly, it becomes us-vs-them, when things are rough) and authority-figure (parents, teachers, bosses). All relationships expect different incentives (eg. paying in financial one and thanking a person in case of loved-one). What’s incredibly useful is to develop a relationship in one frame and nudge it into the next one.
Zappos converted a financial relationship to a quasi-friendly relationship with its customers by having 24×7 customer service, posting the phone number on every page of the site and providing unusually good customer service. Carbon credits were offered to factories to destroy greenhouse gases, a lot of factories in China started producing HFC-23 just to get carbon credits for destroying them when UN changed the rules, these factories might just end up releasing their gas store into atmosphere leading to even more emissions.

Why most incentives backfire?

1. People scheming to game the incentives are way more numerous and way smarter.
2. People think differently, therefore, a mental model of what behavior change it would produce could be different than expected.
3. The assumption that people’s behavior will not change once the incentive is provided is an incorrect one, in fact, behavior can change in an unexpected fashion.

Getting the right incentive scheme

1. Figure out what people care about (revealed preferences) not what they say they care about (declared preferences).
2. Incentivize on dimensions which are valuable for them but cheap for you.
3. Keep an eye on feedback (pay attention to responses) and be ready to change, if responses are unexpected.
4. Switch the framework from adversarial to cooperative (if possible).
5. Never assume that people will do what’s the right thing to do.
6. Some will still be able to game the system, applaud their ingenuity rather than cursing the greed.

What do King Solomon and David Lee Roth have in common?

(Ans: Strong understanding of game theory).

A pooling equilibrium is a situation where different types of people (say liar and honest guy) will act similarly, an incentive can cause them to respond differently and, therefore, leading to separating equilibrium. Rather than making job applications easier, employees can make them harder (just like colleges), which will make them easier to weed out candidates. Zappos implies a different way by offering 2000$ to any employee who wants to quit (and becomes ineligible for future employment). Since bad employees cost about $25, 000, so, Zappos saved a lot by doing that. Nigerian scam emails still use the name of Nigeria to weed out everyone who knows about these scams and only get responses from the most gullible ones. By using the name of Nigeria and writing unrealistic sums of money, they create the incentive to separate the gullible from non-gullible ones. One way to stop them is to write an automated bot which replies to these emails and wastes scammers time. In Superfreakonomics, it was mentioned that suicide bombers in Britain don’t buy life insurance since monthly payments would be a waste of money, by publishing this openly, it caused potential terrorists to buy insurance from their banks (which is a convenient way to an attempt to hide their tracks) except no one buys insurance from their banks (they buy from broker or insurer where it’s cheaper), revealing this in Superfreakonomics was a scheme to separate potential terrorists from normal people (who would by insurance).

How to persuade people who don’t want to be persuaded

If possible, avoid that, it’s much harder than thinking about problems and solving them.

1. Understand its really hard (even harder if a person is more educated since s/he is more confident and dogmatic).
2. It’s important to see how the argument will be perceived rather than what’s being said – if it does not resonate with the listener, it does not matter how indisputable and logically correct it is.
3. Never pretend your argument is perfect – since that makes it sound unrealistic.
4. Acknowledge the strength of an opponent’s argument – if you ignore his argument, he might decide not to engage at all.
5. Do not insult the opponent – since it will strengthen the adversarial situation.
6. Tell the stories (not anecdotes) – since humans relate to them and they can create long-lasting impressions.

The upside of quitting

People avoid quitting due to three reasons

1. They think its a sign of failure
2. Sunk costs – since already so much of resources(time and money) is invested, people invest more to get value out of already invested resources.
3. Opportunity cost – investing resources implies that they cannot be invested somewhere else, rarely people think about that missed opportunity.

Winson Churchill said, “Never give in” (to Harrow school boys) and stood strongly against Nazis but he himself gave up a lot, he moved across parties, quit govt, spent years in the political wilderness and returned only when policy failure lead to all-out war. Premortem is a good practice where an institute before launching a project, thinks about all the ways it can fail. This prevents flushing out the doubts which no one is willing to say out loud.