Takeaway
- People tell Google stuff they would never admit in real life.
- By looking at what people search on Google, we can have an actual picture of people’s minds.
- Google can tell us what people fantasize about, obsess over, dream, are scared of, and who they don’t like.
- Indian men are obsessed with the idea of their wives breastfeeding them. !!!
Table of Contents
Introduction: The Outlines of a Revolution
Chapter 1: Your Faulty Gut
Chapter 2: Was Freud Right?
Chapter 3: Data Reimagined
Chapter 4: Digital Truth Serum:
Chapter 5: Zooming In
Chapter 6: All the World’s a Lab
Chapter 7: Big Data, Big Schmata? What It Cannot Do
Chapter 8: Mo Data, Mo Problems? What We Shouldn’t Do
Conclusion: How Many People Finish Books?
What Everybody Lies Talks About
Everybody Lies is a book written by Seth Stephens-Davidowitz. It’s an account of what people secretly search on Google and other search engines and what it really says about society. You will learn about people’s weird sexual fantasies, what people dream about the most, and what men and women are the most self-conscious about when it comes to their bodies.
Jump to the odd and crispy facts directly if you don’t want to read the summary.
The beginning of the book got me incredibly excited. The end bored me a little.
Still, Seth said stuff that few other people ever dared to say – and he backed them up with data!
As a result, I’ll give this book a 9/10.
Summary of Everybody Lies Written by Seth Stephens-Davidowitz
Introduction: The Outlines of a Revolution
Women say they have sex 55 times a year with condoms, which would mean that the number of condoms used is 1.1 billion. For men, it’d be at 1.6 billion.
The truth is that 600 million condoms are sold each year in the United States. So people lie about using condoms, or about the number of times they have sex. Or both.
Everyone is lying, and this is something we can verify with Google data. People write stuff in Google they would never dare tell anyone.
For example, the author uncovered that when Obama was elected, 1/100 searches on Google were racist. He subsequently found out that the place where the searches came from had voted less for Obama than they did for John Kerry, the democrat that was running for the White House four years later which confirmed that the biggest differentiator, was indeed, race.
The US is far more racist than thought originally.
When Donald Trump was elected, the states that massively voted for him were also the states with the most racist queries.
This, in a nutshell, is what Big Data does. It helps us penetrate deep inside the human psyche.
Chapter 1: Your Faulty Gut
Data science isn’t complicated. It is about spotting patterns and predicting how one variable will affect another, which is something everyone does all the time.
While humans are generally good at linking variables to each other, Big Data is interesting when it shows us stuff we’d never thought of linking together before.
It often shows results that are counter-intuitive.
Chapter 2: Was Freud Right?
The author analyzed porn search results and found that many people looked for incest porn.
The top searched roles in porn movies for men are teachers and babysitters. This shows how the early years in the life of a man shape his future sexual fantasies.
All of this data did not exist 10 years ago.
Big data is important for four reasons:
- Big Data offers a new type of data, things people never told anyone but search on the Internet
- Big Data is honest
- Big data enables us to zoom and focus on a small population searching for specific things.
- Big Data enables to test for causality
Chapter 3: Data Reimagined
The value of data, as we can see, does not depend on its size, but on what we can do with it.
Eg: Places where unemployment is high also have high queries for porn and…the spider solitaire game. This means that people that don’t work mainly play and watch porn.
Let’s talk about horse racing. It used to be difficult to predict how well a horse would perform in a race.
Until came Seder. Seder was a Harvard graduate who compiled data to find signs that would help predict horse performances.
He found out that the size of the spleens, heart, and left ventricle were the best predictors. The bigger they were, the faster the horse was.
Another interesting dataset to look at is words.
After Google began digitalizing books, two biologists built a search engine that analyzed how many times a word was written at a given period. They found out that sausage spread slowly, but that pizza spread quickly.
They also asked people that went on dates to record themselves in order to analyze the transcript and conversation later.
Men speak with a monotone voice and control their pitch to signal they are interested.
Women speak less when they like a guy. The more they talk about themselves, the more they are interested. “I mean” and “you know” are positive signals for the guy.
“Probably”, and “I guess” are negative signals. Likewise, a conversation with a lot of questions is unlikely to lead to a second date.
Women like men who follow their lead. They like when the guy laughs at their jokes and talks about topics they introduced. They like it when the guy validates their feelings, saying things like “that’s awesome”, “that’s really cool”, “that’s tough”, or “you must be sad.”
Guys are not focused as much on the conversation as on the girls’ appearances.
The word cloud below indicates words that are predominantly spoken by men. The bigger the word, the fewer girls say it.
The following word cloud is the same, but for women.
The words people use also change as they age.
A team of scientists analyzed the mood of books and movies as they progress. They found out there were six types of stories.
- Rags to Riches. The character starts low, then rises.
- Riches to Rags. The character starts high, then falls.
- Man in a Hole. The character falls, then rises.
- Icarus. The character rises, then falls.
- Cinderella. The character rises, then falls, then rises again.
- Oedipus. The character falls, then rises, then falls again.
They also found out that:
- Viral stories are in majority, happy stories.
- While on average, newspapers lean left, this is because on average, readers lean left (newspapers only deliver the content readers want to read.)
- People weren’t smiling in pictures because they compared pics to paintings, and no one ever smiled on a painting due to the impossibility of smiling for a long time.
As time went by, people started smiling more and more due to business, marketing, and advertising.
Chapter 4: Digital Truth Serum
Warning: this chapter wanders through the darkest corners of the human psyche. Do not read if you’re sensitive.
When surveys polled people in the 2016 election, many said they would vote Democrat when they’d vote Republican, and many said they were undecided when in fact, they were decided.
People lie roughly one-third of the time.
When you ask them about their grades at school or who they’re going to vote for, they lie.
Part of the reason is that they want to make a good impression. Another part is that they’re not lying to others – they are also lying to themselves.
Eg: More than 40 percent of engineers think they are in the top 5 percent of all engineers. And more than 90 percent of university professors think they do above-average work.
People say online stuff they’d never say anywhere else, like being gay, for example. When looking at the data from openly gay people in tolerant places and searches for gay porn in intolerant places, the number is the same: 5%.
That means that if a certain region has less than 5% openly gay people, a lot of these people hide they’re gay.
Let’s talk about sex.
Sex
Regarding heterosexual people, there are twice as many complaints online from women saying that their boyfriend won’t have sex with them than the opposite.
It may be due to insecurities. Men are obsessed over their penis size. They google more about their penis than any other organs, and men’s worries related to taking steroids or aging concern…their penis size.
The weird thing is that men google 170 times more about their penis size than women do. And when women complain about penis size, 40% of these complaints say that it is too big. Yet, only 1% of men’s searches about penis size seek to make it smaller.
Conclusion: penis size doesn’t matter to women as much as men think.
The second insecurity of men is about how long they can last in bed. Yet once again, women search more about how to make their boyfriends climax quicker than slower.
They are also more worried about their boyfriends not having orgasms, than the opposite.
Women make up 62% of queries about how to improve how their body looks (and men, 38%).
Women’s first insecurity is about their breasts, then their butts.
Prior to 2010, women sought to make their butts smaller. After 2010 (understand: Kardashian), women started to search for how to make their butts bigger, which correlates to men’s desires (queries for “big butts” in P*rnhub are high).
Men, in general, want a girlfriend with big breasts.
Let’s have a look under the belt.
Women are as worried about their genitalia as men are. Their vaginas’ smell is their top concern. They complain on Google that they smell like (in volume order):
- Fish
- Vinegar
- Onions
- Ammonia
- Garlic
- Cheese
- Body odor
- Urine
- Bread
- Bleach
- Feces
- Sweat
- Metal
- Feet
- Garbage
- Rotten meat
Other questions about their pubis concern how to shave it, tighten it, or…make it taste better.
Men make as many searches about their partners’ vaginas as women do about penises (that is, not many).
Conclusion: we’re all so busy thinking about our own insecurities that we don’t focus on our partners’ bodies.
There are twice as many women looking for oral sex tips than there are men. And when men search for oral sex tips, half of it is to give oral sex…to themselves. The other half is for their girlfriends.
That concludes the part about sex.
Let’s now talk about race.
Race
Many Americans are still racist.
The table below shows the top five insults they search on Google, with the corresponding racial group.
After an Islamist terrorist attack in 2015, half of the searches with the word “Muslim” were about…killing them.
Four days later, Obama gave a speech calling on everyone in the country to reject hate and discrimination.
The speech was considered a major success by the mainstream media. In practice, it was a total failure.
Hateful and discriminatory queries more or less tripled during and after the speech.
Every year, 7 million US searches have the word “n*gg*r”. “N*gg*r jokes” is particularly popular, and the volume of the query increases every time Black people are in the news (MLK day, Obama election, etc).
These searches come from places where Black people are underpaid and that massively voted for Donald Trump.
The author also looked at Stormfront, the most popular hate website in the US, visited roughly 400 thousand times per month (the website has now been closed off).
Members of Stormfront are rather young, born in the 90s, and 70% of members are male. And weirdly, they love the New York Times.
In fact, they’re twice more likely to read the Times than people that read Yahoo News are.
-> Despite what was thought, a huge chunk of liberals are in fact, racists.
Memberships in Stormfront increased directly after Obama’s election, but did not increase under Donald Trump.
The author suggests that Obama created a wave of white nationalism in the country, that Trump subsequently rode up to the White House.
Let’s now speak about the Internet.
The Internet
The general view is that people only read what they agree with online and are not exposed to opinions they disagree with due to the filter bubble.
This is not true.
There is 45% chance that two people with different political views visit the same website. In life, there are only 34% chances that your friend has a different political opinion than you do.
So real life is much more segregated than the Internet.
Why?
For two reasons.
- A few websites dominate the majority of online news. Everyone read the same stuff.
- People love reading news they disagree with for the mere pleasure of getting angry.
Let’s talk about child abuse and abortion.
Child Abuse and Abortion
We can measure child abuse because when kids are in trouble with their parents, they search for “my dad hit me” on Google.
These queries skyrocketed during the 2008 downturn. The highest number of searches came from places where the unemployment rate was the highest.
Likewise, states that make it difficult to get an abortion have spikes of women looking for ways to do it themselves compared to states where abortion is legal and easy.
Regarding discrimination, many are led against young girls. Parents are twice more likely to ask Google if their boy is a genius, than if their girls are despite that at the same age, girls are more developed than boys.
Likewise:
- They’re twice as likely to google “is my daughter overweight” than “is my son overweight”, despite that more boys are overweight than girls.
- They are 1.5 times more likely to ask “is my daughter beautiful” than they are for boys.
- They are three times more likely to ask if their daughter is ugly than they are for boys.
While Google mirrors the truth, Facebook data is mostly fake because people portray themselves in a way they’re not. On Facebook, people are not honest.
Eg: While The Atlantic and the gossip magazine National Enquirer have the same number of readers on Google data, the former has 30 times more likes on its Facebook page than the latter.
But that doesn’t mean that Facebook itself has fake data.
When Zuckerberg introduced the Facebook newsfeed, people were outraged and wanted it off the platform. But Zuck didn’t budge. He noticed that not only people clicked more on the newsfeed, they also spent more time on Facebook.
He understood there was a huge difference between what people say they want, and what they really want.
Eg: Women would be quick to say they don’t enjoy BDSM stories between a young innocent college graduate and a dark business guy, but 50 Shades of Grey sells 125 million copies.
Everybody lies.
Chapter 5: Zooming In
Why are some boys fans of certain sports teams, and others aren’t?
It appears that young boys develop a fandom for a certain sport (team) at around eight years old.
Eg: if a basketball team does particularly well when the boy is eight, he will likely be a basketball fan of that team for his entire life.
If it’s a football, or baseball team, then he will be passionate about football or baseball.
Women become passionate about a sports team when they are twenty-two years old on average.
What about political affiliation, music, or financial habits?
Most Americans will become more republican or democrat depending on the popularity of the president when they’re between fourteen and twenty-four.
If the president is popular, they’ll vote for his party. If he is not, they’ll join the opposite side.
These views, furthermore, may last a lifetime.
Zooming In
The great thing with big datasets is that we can zoom in. In America, some towns help poor kids become part of the 20%, while other towns severely decrease their chances to be so.
Richer Americans live longer. Not necessarily because they’re richer, but because they have better habits.
They smoke less, exercise more, eat healthier, etc.
These rich people, because of their good habits, also help poor people live longer.
It’s true.
The number of rich people in a city has an impact on how long poor people live. The more rich people you have, the longer poor people will live.
Why? We don’t know exactly.
One theory states that habits are contagious. Because rich people have healthier habits, they help poor people adopt them too.
Doppelganger
Internet companies no longer need much data about you. Once they have a few points, they look for doppelgangers – that is, people that like the exact same things you like, and see what else they like so that they can recommend it to you too.
Chapter 6: All the World’s a Lab
Correlation doesn’t mean causation.
The only way to find out if correlation is in fact causation, is to do an A/B test.
An A/B test compels you to test different things with different groups of random people to see which one worked best.
Today, A/B testing is everywhere on the Internet. The problem is that it leads designers to over-optimize, making us hooked to their products.
Big Data also helped understand the value of education.
Do students perform better in life once they attend a prestigious highs-school?
To find out, data scientists measured how well students that barely miss the entry exam for one high school fared in life, compared to those that barely made it to this high school.
The result is that attending the high-school did not influence students’ life one bit. Both ended up having similar trajectories in life.
While going to a good school is important, no need to go to the best school. What makes you successful is your talent and drive – not your school.
Chapter 7: Big Data, Big Schmata? What It Cannot Do
Since we can predict the results of elections based on searches for racist queries, could we do the same with the stock market?
The answer is no.
Big Data can’t help with the stock market because Big Data helps solve problems where data lacks.
Data does not lack when it comes to finance – in fact, it is the opposite!
Many data scientists come up with datasets that they think will help them predict the stock market, but it doesn’t.
These scientists are victims of the curse of dimensionality. The curse of dimensionality happens when one uses a lot of variables to determine only a few predictions (in this case, will the stock market be up or down tomorrow?).
If you test enough things against an observation, you will always find correlations. The problem is that this correlation is rarely relevant.
Eg: while the number of kilos of chocolate that a population eats per year correlates with the number of Nobel prizes from the country, there is no way these two variables influence each other. They’re just random, called spurious correlation.
Side note: this website created various spurious correlations that are funny to look at.
More data doesn’t mean better data, as we have already established. There is a lot of data we can collect but that doesn’t help us understand things better. Sometimes, the data we miss can be assembled with just a dumb survey.
Chapter 8: Mo Data, Mo Problems? What We Shouldn’t Do
Let’s now take a look at ethically questionable things we can do with data.
A team of economists found out that a borrower’s chances of paying back his loan could be predicted based on the words he used in his description of why he needed a loan.
Someone who mentions God was 2.2 times more likely to default. This was among the single highest indicators.
This leads to the following dilemma: should corporations have the right to judge our fitness for their services based on abstract but statistically predictive criteria not directly related to those services?
Employers now scout their potential employees’ social media for certain words or behavior indications.
Casinos use Big Data to find out the pain point of their customers. The paint point is an arbitrary sum of money past which a casino player will not come back to play because it’s too much.
Casinos try to stop their customers before they reach their pain points.
But should all this be allowed?
Conclusion: How Many People Finish Books?
The best conclusions bring to the surface an important point that has been there all along.
For this book, that big point is this: social science is becoming a real science thanks to Big Data.
Odd Data Facts
In Mexico, men search for poems and words of love to tell their wives when they get pregnant. In the US, men google “my wife, is pregnant, now what?” or “my wife is pregnant, what do I do?”.
In India, the number one search starting with “my husband wants…” is “my husband wants me to breastfeed him”. Porn queries for such practices are very popular in India and Bangladesh.
Summer climate is twice as powerful as antidepressants to solve depression.
Having common friends with your partner (husband, wife, boyfriend, girlfriend) is an indicator that your relationship is less likely to continue. Having separate groups of friends increases the chances of your relationship continuing.
More people die of asthma than tornadoes.
Black men are forty times more likely than white men to reach the NBA.
Middle-class kids are more likely to reach the NBA than poor kids.
The substance that is most dreamed about is water. The top twenty foods include chicken, bread, sandwiches, and rice. The more we eat food, the more we dream about it.
A lot of people on porn sites look for incest porn. Men look for “mom and son” and women look for “father and daughter”.
Whether wine is good or not solely depends on the weather during the growth period.
More than 40 percent of engineers said they are in the top 5% and more than 90 percent of university professors say they do above-average work.
20% of videos watched by women on P*rnhub are lesbian videos.
Sometimes, men search on P*rnhub for:
- fat women
- women with tiny t*ts,
- women with green hair
- bald women
- midget women
- women with no nipples
- shemales
- granny
Women search on P*rnhub for:
- short guys
- pale guys
- ugly guys
- disabled guys
- chubby guy with small d*ck
- fat ugly old man
25% of porn searches for women are related to sex featuring violence against women. In general, violence against women appeals more to women than to men. Keywords are:
- painful anal crying
- public disgrace
- extreme brutal gangbang
5% of them are looking for:
- rape
- forced sex
There are twice as many complaints that a boyfriend won’t have sex than that a girlfriend won’t have sex.
Men are obsessed over their penis size. They google more about their penis than any other organs, and worries related to taking steroids or aging concern…their penis size.
Men google 170 times more about their penis size than women do. And when women complain about penis size, 40% of these complaints say that it is too big. Yet, only 1% of men’s searches about penis size are about making it smaller.
The second insecurity of men is related to how long they can last. Yet women search more about how to make their boyfriends climax more quickly, than slowly. Women are more worried about their boyfriends not having orgasms, than the opposite.
Women make 62% of search volume for queries related to how they can make their bodies look better (and men make the rest lol).
Women feel insecure about their breasts first, then their butts.
Prior to 2010, women sought to make their butts smaller. After 2010 (understand: Kardashian), women started to search for how to make their butts bigger, which suit men’s desires (high volume queries for “big butts” on P*rnhub).
Men, in general, want a girlfriend with big boobs.
Women are as worried about their genitalia as men are. They complain on Google that their vaginas smell like (ordered per volume):
- Fish
- Vinegar
- Onions
- Ammonia
- Garlic
- Cheese
- Body odor
- Urine
- Bread
- Bleach
- Feces
- Sweat
- Metal
- Feet
- Garbage
- Rotten meat
Other questions about their pubis concern how to shave it, tighten it, or…make it taste better.
Men make as many searches about their partners’ vaginas as women do about penises (that is, not many).
There are twice as many women looking for oral sex tips than there are men. And when men search for oral sex tips, half of it is to give oral sex…to themselves. The other half is for their girlfriends.
Crime drop when a popular violent movie is shown at the cinema. This is because young, aggressive men go watch the movie. They don’t drink and are not in a place where crime usually happens.
Google queries for “weather”, “prayer”, and “news” peak before 5h30.
Searches for “suicide” peak at 00h30, and are the lowest at 9h00.
Searches for rolling a joint peak at 1h00-2h00, and searches for big existential questions peak at 2h00-4h00 (likely related).
Neighbors of lottery winners are likely to go bankrupt because they try to equal spending with their now-rich neighbors.
Your university will not have an impact on your future career. What matters is your drive and talent.
Did you enjoy the summary? Get the book here!
Subscribe to my monthly newsletter and I'll send you a list of the articles I wrote during the previous month + insights from the books I am reading + a short bullet list of savvy facts that will expand your mind. I keep the whole thing under three minutes.
How does that sound?
