DEEP SUMMARY - The Ten Equations That Rule the World_ And - David Sumpter

Play this article

Here is a summary of the key points in the introduction to David Sumpter's book "The Ten Equations that Rule the World":

  • Sumpter claims there is a secret society that has discovered 10 equations that can bring success, popularity, wealth, and self-confidence. He calls this society "TEN."

  • Sumpter says he is a member of TEN and has benefited from using the 10 equations in his work as a professor, consultant, and scientist.

  • The equations address both profound and trivial dilemmas. Understanding them provides insight into individual behavior and society as a whole.

  • To join TEN, you must learn a new way of thinking - breaking down the world into data, models, and nonsense.

  • TEN is powerful today because we have more data than ever before to which the equations can be applied.

  • Learning the 10 equations will help you see the world in a new way and solve both big and small problems.

  • Sumpter will reveal the 10 equations over the course of the book, showing how they have brought success to members of TEN across many fields.

    Here are the key points I gathered from the summary:

  • The author met two young gamblers, Jan and Marius, who were interested in the author's mathematical model for betting on football matches.

  • The model is a simple equation that takes in bookmakers' odds, adjusts them for bias, and suggests new odds to bet on. It has generated a 1900% profit betting on Premier League matches.

  • Jan and Marius have developed their own automated system to find value bets by exploiting "soft" bookmakers who don't set accurate odds.

  • The author publicized his model but it still works because most bettors don't use a systematic approach and bet based on gut feelings or for fun.

  • Jan and Marius represent the small minority of informed gamblers who use models and statistics rather than emotions to make bets.

  • The key is that the model suggests bets most people don't want to make, like betting on a draw or a likely winner at low odds over the long run.

    Here is a summary of the key points:

  • Jan and Marius had developed a system to profit from betting by using data to identify soft bookmakers who offered better odds than sharp bookmakers. They would bet with the soft bookmakers when their odds were more generous.

  • Their system allowed them to continue profiting even when they got banned by soft bookmakers, by offering a subscription service to others with bets identified by their model.

  • The author met them to discuss applying his predictive model of football to improve their edge and make even more money.

  • Jan and Marius represented a new breed of professional gambler, skilled with data and programming to automate betting.

  • With a large enough edge and bankroll, the potential winnings compound massively if bets can be placed quickly enough, though there are practical limits.

  • The author learned how some gambling syndicates in London have grown rapidly by exploiting small edges with big data and automation.

  • The author proposed an equation to model the probability a favorite wins based on the odds. It requires parameters α and β to make it profitable.

  • Without α and β, the model breaks even, but with optimization of α and β it can identify positive expected value bets.

    Here is a summary of the key points:

  • The author and his friends Jan and Marius develop a mathematical model to predict football match outcomes and beat the bookmakers.

  • They use historical data on match odds and results to build a logistic regression model. This finds the optimal values for parameters α and β to make the most accurate predictions.

  • The model shows a 'long-shot bias' - it predicts stronger favorites like Spain are undervalued by the bookmakers, while weaker favorites like England are overvalued.

  • The model gives them a small edge to profit from bets on the upcoming World Cup. They implement an automated betting system.

  • More broadly, the model illustrates the power of thinking probabilistically about uncertain future events like jobs, relationships etc. Rather than expecting certainty, evaluate the probabilities and potential payoffs.

  • The author discusses William Benter's pioneering work using math models to successfully beat the Hong Kong horse racing market. His intense commitment to rigor and documentation showed mathematical edge was possible.

    Here is a summary of the key points about Benter's strategy for upcoming horse races:

  • Benter was an American outsider who figured out a novel way to get inside information on horse racing by digitizing data from the Hong Kong Jockey Club.

  • He applied the betting equation, based on logistic regression, to analyze factors like past performance, time since last race, age of horse, etc. This allowed him to make increasingly accurate predictions.

  • After several years of refining his model and surviving ups and downs, Benter's profits grew exponentially, reportedly earning him over $3 million in one season in the early 1990s.

  • Over two decades, Benter and others using similar methods are estimated to have made over $1 billion on Hong Kong racetracks.

  • Benter published his strategy in an academic paper, but it has been largely ignored, with less than 100 citations in 25 years. The key insights were there in plain sight all along.

  • By persevering to understand the math and details, Benter forged a connection spanning centuries of work on probability and statistics underlying his success. This reflects how TEN's secret wisdom is open to those who diligently seek it out.

    Here is a summary of the key points:

  • Sir David Cox developed the theory of logistic regression, which was key to creating the betting equation used by Bill Benter and others to predict horse race outcomes.

  • Logistic regression allows you to determine the probability of an outcome based on various factors. Benter used it to see how factors like a horse's race history affected its chances of winning.

  • Cox was inspired to develop logistic regression by practical problems he encountered working in industry and the military during and after WWII.

  • The technique has been widely applied in medicine, psychology, business, and now gambling. Benter made $1 billion using it to predict horse races.

  • Betting is about finding small differences between your understanding and others'. By testing lots of small variations, you can refine your edge like tuning the parameters in the betting equation.

  • The inequality between those who know the equations and those who don't applies beyond gambling. Mathematical techniques have driven progress but been controlled by a small group. Those who know the secrets, like Cox and Benter, have benefited.

    Here is a summary of the key points:

  • Bayes' theorem allows you to update probabilities as new information comes in. It helps you make rational judgements in uncertain situations.

  • The author visualizes different future scenarios playing out in his head like movies. This helps him think through possibilities and estimate their likelihoods.

  • We all use "models" or representations to think about the future. Becoming aware of how you do this is the first step to a mathematical approach.

  • The author uses probability estimates to organize his "movie collection" and not get overly worried about unlikely bad outcomes.

  • An example is given of a girl Amy who estimates the probability someone is a "bitch" to judge who to befriend. Bayes' theorem could help her update this as she gathers more information.

The key ideas are using probability and models to make rational judgements, being aware of how your mind represents possibilities, and updating likelihoods based on new data. Bayes' theorem formalizes this process.

I summarized the key points:

  • Amy is a new student and Rachel is trying to help her catch up, but Amy is struggling with the concepts.

  • Rachel complains to another girl that Amy is stupid and doesn't understand basic concepts like cultural appropriation. Amy overhears this and is upset.

  • The passage argues that Amy should forgive Rachel and give her another chance, using Bayes' rule to show there is a high probability Rachel is actually nice and just made a mistake.

  • It says we should be slow to judge others harshly based on limited interactions, and instead incrementally update our assessments as we gather more data.

  • The passage advises applying the same principle of gradual adjustment to our self-assessments when we make mistakes.

  • It emphasizes using care and restraint in forming opinions of others as a sign of good judgment, like Elizabeth Bennet in Pride and Prejudice.

  • The history and context of Bayesian thinking is briefly summarized, focusing on using reason precisely to evaluate truth and morality.

    Here is a summary of the key points:

  • Thomas Bayes developed a formula to estimate the probability of an event occurring again based on prior occurrences. This allows one to update beliefs as new data arrives.

  • Bayes' formula was applied by Richard Price to argue that a lack of miracles over time does not prove miracles can't occur, as claimed by David Hume. Price showed Hume was too definitive in dismissing miracles.

  • Price believed in using rational thinking and mathematics to reveal greater truths about morality and God's role in the world. He advocated for fairness, equality, and sharing risks in society.

  • Modern practitioners of Bayes' methods often inherit Price's values of order, structure and benefitting society, through work in insurance, pensions, policy planning, etc.

  • The judgement equation and Bayes' formula can be seen as putting one on a path to righteousness according to Price's Christian moral philosophy.

  • An example is given of Björn using Bayesian statistics to study immigration and crime in Sweden for his PhD, aiming to explain cultural changes in Swedish society.

    Here is a summary of the key points:

  • Bayesian reasoning allows scientists to compare multiple hypotheses or models (M) against observed data (D) using Bayes' rule. It calculates the probability of a model given the data, P(M|D).

  • An example is given of a study linking teen mobile phone usage to poorer mental health. However, the study only looked at one hypothesis and did not consider alternative explanations.

  • Professor Candice Odgers filled in the gaps by considering other factors like sleep and breakfast as alternate models. When accounted for, mobile phone usage had a smaller effect on mental health.

  • There are also benefits of mobile phones for teens such as building social networks. The problems tend to be greater for disadvantaged youth.

  • The author notes how his own kids use mobile phones to discuss and learn about new ideas, something he would not have done without the technology.

  • Overall, Bayesian reasoning allows proper weighing of multiple hypotheses against data, avoiding the mistake of focusing on only one explanation. It is a powerful scientific approach.

    Here is a summary of the key points:

  • The author provides an example of his sister Amy judging a new classmate Rachel as likely being a "bitch" based on some subjective probability estimates. He notes that while the numbers are subjective, Bayes' rule can still help reason about them logically.

  • Many people, even scientists, don't realize Bayes' power comes from forcing you to lay out your assumptions and models explicitly before and after data is collected. It promotes intellectual honesty.

  • Bayes' theorem allows the author to interpret and assess research on the effects of mobile phone use on his children's mental health. He concludes moderate use is fine based on a balanced review.

  • The author argues it is partly the responsibility of readers to critically analyze claims by so-called "experts" on parenting and health. We should check they present balanced models accounting for all evidence.

  • The author encourages being open-minded, letting data guide decisions, and giving others multiple chances. This Bayesian approach leads to good judgment and trust.

  • The world is full of advice lacking structure. Bayes' theorem provides a way to organize and evaluate each claim as a testable model against data. This results in better choices and judgments.

    This passage discusses using probability and statistics to determine confidence in outcomes for games of chance and other applications. The key points are:

  • Games of chance like roulette have known probabilities that can be used to calculate expected outcomes. For example, the expected loss per £1 bet on red or black in roulette is -£0.027.

  • The normal distribution and standard deviation can be used to estimate a confidence interval around the expected outcome. This accounts for the randomness and gives a range that the outcome will fall within 95% of the time.

  • The confidence equation allows calculating a 95% confidence interval as the expected outcome ± 1.96 times the standard deviation.

  • This same statistical approach can be applied not just to gambling but also polling, hiring practices, and other areas to determine confidence intervals around expected outcomes.

  • The concept of confidence intervals originated from efforts to analyze games of chance but became a critical tool for establishing confidence in scientific and social science research results.

    Here is a summary of the key points:

  • Abraham de Moivre derived the normal distribution in 1733 to calculate the outcome of repeatedly tossing a coin. This allowed him to calculate probabilities for large numbers of coin tosses without having to do endless multiplication.

  • The normal distribution equation can be used to model many real-world situations involving repeated random events, as shown by the central limit theorem proved in the early 20th century.

  • The normal distribution is important for quantifying confidence - it allows you to estimate the range of possibilities for an unknown true value based on a sample of observations.

  • For a gambler to know if they have a real edge, they need enough observations for the confidence interval around their estimated edge to exclude zero. The rule of thumb is you need 4/signal-to-noise observations.

  • So if a gambler has a 3% edge with a bet standard deviation of 71p, they would need around 1,600 bets for the confidence interval to confirm their positive edge. With fewer bets, the edge is consistent with being positive or negative due to randomness.

    Here is a summary of the key points:

  • To detect a small edge or signal in noisy data requires a large amount of observations - often thousands or tens of thousands. This is due to the confidence interval shrinking as the square root of the number of observations (the square root of n rule).

  • Jan and Marius's sports betting strategy is built on a database of over 15 billion past betting positions, allowing them to confidently detect edges as small as 1.5-2%. This requires a huge amount of data.

  • When looking for a hotel on TripAdvisor, around 16 reviews are needed to reliably detect a half star difference in ratings. The square root of n rule applies here too.

  • Jess and Steve use a star rating system and regular meetings to evaluate Jess's job and Steve's relationship. After 100 daily star ratings they have enough data to make confident decisions on their future.

  • The experience of one person provides little information, like one pull of a slot machine. The collective experience of a large group is needed to draw confident conclusions, as with Malcolm X and the African American struggle against discrimination.

  • Indirect discrimination can occur when people share opportunities mainly within their own social groups, even if unintentionally. Jamie missing out on the job opportunity Joanne told James about is an example of this.

    Here are a few key points summarizing the passage:

  • The confidence equation allows us to estimate the probability of an outcome occurring by chance alone. It is essential for determining if observed patterns are real or just random noise.

  • Researcher Moa Bursell sent out thousands of fictional job applications in Sweden with Swedish and Arab names. She found strong evidence of discrimination against applicants with Arab names, even when they were more qualified.

  • Studies like Bursell's reveal structural racism through statistics, even when discrimination is hard to see at an individual level.

  • The author argues we should be "statistically correct" - aware when our individual experiences do not reflect wider society, and consider what actions to take to address imbalances.

  • The confidence equation transformed science by allowing researchers to determine if their results were meaningful. It is now a standard part of scientific writing and discovery.

    Here is a summary of the key points:

  • Until recently, sociology departments were dominated by ideological debates and theoretical frameworks rather than quantitative data analysis. Sociologists were seen as eccentric and out of touch.

  • In the early 2000s, the availability of large datasets completely transformed the field. Now theories must be backed up by statistical analysis of real-world data.

  • Some old-school sociologists joined the data revolution, while others were left behind. Ideological debates were pushed to the periphery of the field.

  • However, some media publications like Quillette continue to wage a "culture war" against ideas like structural racism and identity politics, often without engaging with the data.

  • Figures like Jordan Peterson also attack the social sciences for political correctness and leftist ideology, claiming academics are scared to speak freely.

  • In reality, modern social scientists are constrained not by ideology but by the need for statistical rigor. The field is focused on collecting data to test models, not abstract theorizing.

  • Social scientists like Moa Bursell are motivated by their political beliefs but use objective methods like audit studies to test their theories. The data has surprised her at times.

    Here is a summary:

Jordan Peterson argues that the gender pay gap does not necessarily indicate discrimination, since it may be due to women choosing lower-paying careers. However, social scientists like Moa Bursell have conducted experiments showing clear discrimination against women in hiring practices. Other studies reveal subtle biases that limit women's opportunities, like underestimating women's competence on resumes or making them fear backlash when negotiating salaries.

Peterson dismisses this research as ideologically biased. But the scientists aim to identify barriers to equality of opportunity, which Peterson claims to support. Peterson instead focuses on psychological differences, like women being more "agreeable." But agreeableness only weakly correlates with lower pay for women. The signal is drowned out by noise. Personality tests don't clearly explain how opportunity is affected. Ultimately, context-specific experiments are more convincing than vague personality explanations for showing where inequality arises. If we want equal opportunity, we should educate people on research identifying biases.

Here is a summary of the key points:

  • The author describes meeting a former football player turned TV personality ("Mr 'My Way'") who engages in self-promoting behavior - shaking hands, small talk, name dropping, rehearsed anecdotes, etc.

  • At first the author was intrigued by getting an inside view of football from a former player. But he realizes that despite the entertaining stories, Mr 'My Way' actually provides little substantive information.

  • The author has encountered this behavior from many people across fields - football, business, academia. They emphasize their unique talents and insights, and blame external factors when things go wrong.

  • The author initially bought into these self-promoting stories, but over time realized they lacked evidence or accountability. He sees this as an example of overconfidence bias.

  • The author argues for the need to replace anecdotes and stories with data and statistical evidence in order to make informed judgments about skill and performance. Relying too much on personal stories and perceptions is unreliable.

    Here is a summary of the key points:

  • The author presents a mathematical approach to analyzing football players by looking at their contributions to their team scoring goals, rather than just goals scored.

  • He focuses on Paul Pogba as an example, arguing Pogba defines the teams he plays for more than other star players like Messi or Ronaldo.

  • To analyze Pogba's contributions, the author tracks every action on the pitch (passes, tackles, etc) and measures how each increases his team's probability of scoring and decreases the opponent's probability.

  • He represents the football pitch in x,y coordinates and describes each pass or action as coordinates. This allows quantifying sequences of play or "possession chains".

  • The overall goal is to quantify all of Pogba's actions to evaluate his total contribution to his team's chances of scoring, not just goals he scores himself. This is a mathematical approach to football analysis focusing on probability and spatial coordinates.

    Here is a summary of the key points:

  • The author wants to evaluate how each player's individual actions increase their team's chance of scoring and decrease their opponent's chance.

  • To do this, the author makes a mathematical assumption that the quality of a pass depends only on its start and end coordinates, not the context around it. This allows assigning a value to each pass.

  • The author uses data on passes from many seasons of football to fit a model linking pass coordinates to probability of a goal.

  • This model shows Pogba was highly effective for France in the World Cup at recovering the ball and making long passes to advance attacks.

  • The model complements traditional scouting by focusing specifically on passing ability. The author explains his assumptions clearly when discussing the model.

  • The Markov assumption underlies most models for measuring skill. It says future states depend only on the recent past.

  • The author gives the example of a bartender serving customers to explain the Markov assumption. It focuses only on the bartender's service rate, not earlier states.

  • Equations based on the Markov assumption are a step towards answers, but the assumption itself is not the answer. We must be honest about our assumptions when creating models.

    Here is a summary of the key points:

  • The principle of verifiability arose from the thinking of the Vienna Circle philosophers, led by Moritz Schlick and Rudolf Carnap. Their views, known as logical positivism, held that all meaningful statements must be verified against empirical data.

  • This view was influenced by Ludwig Wittgenstein's Tractatus Logico-Philosophicus, which argued that statements that cannot be verified are meaningless. The Vienna Circle's ideas spread via A.J. Ayer's book Language, Truth and Logic.

  • In the early 20th century, logical positivist thinking had a major impact on the methods of Those who must be obeyed (TEN). Models and data became the sole authoritative way of understanding the world.

  • TEN flourished across Europe, with key figures like Kolmogorov in Russia, Cox in the UK, and Einstein, Bohr, and Schrodinger driving physics and mathematical innovation. The principle of verifiability superseded other ways of thinking, including religious beliefs which were seen as unverifiable.

  • TEN values precise language, transparent assumptions, and comparison of models to data. Discussions aim to find the explanation that is least wrong, politely ignoring those who don't speak the language of models and data.

  • Luke Bornn, despite not having a traditional sports background, was drawn into basketball analytics by the richness of player data, showing how TEN's methods have spread to new domains.

    Here is a summary of the key points:

  • Basketball teams had extensive data on players' movement and plays, but coaches were not utilizing it much.

  • Luke Bornn applied his analytical skills to basketball data, developing new defensive metrics called 'counterpoints' that measured 1-on-1 matchups. This got him hired by an Italian soccer club and then an NBA team.

  • Bornn uses the Markov assumption in his models - ignoring most of a player's history and just focusing on their current position on the court. This allows simulations to find optimal strategies like passing outside the 3-point arc more.

  • Baseball saw similar analytical advances, with mathematicians like Bill James brought in to help teams like the Boston Red Sox. Underdog teams like the Oakland A's used analytics to succeed.

  • Doug Fearing, a Harvard professor, worked for the Tampa Bay Rays and LA Dodgers, applying analytical approaches. He notes baseball is easier to model with the Markov assumption since it's a series of 1-on-1 matchups.

  • Early statistical papers in the 1960s-70s by mathematicians like George Lindsey laid the groundwork for modern sports analytics. There's been a shift from intuition-based coaching to Ivy League data analysis.

    Here is a summary of the key points:

  • The probability of being yourself (out of the ~8 billion people on Earth) is extremely small, about 1 in 8 billion. This is much less likely than winning the lottery with a single ticket.

  • Imagining waking up each day as a random person highlights how improbable yet unimportant each of us are on a global scale. Most days would be spent in crowded cities in China or India.

  • Waking up each day as someone you follow on Instagram would provide more familiarity, though still random. The probability of being any given person is much higher than being yourself out of the full global population.

  • These thought experiments emphasize how unlikely yet insignificant each of us are as individuals among billions of people. Our sense of self-importance is misguided when viewed in the context of the whole planet.

    Here is a summary of the key points:

  • The author imagines waking up each morning in the body of a different person he follows on Instagram. There would be a chance he wakes up as himself again, but he would likely spend time travelling through his social network.

  • At some point he would wake up as a celebrity like Cristiano Ronaldo or Ariana Grande with hundreds of millions of followers. He would then keep jumping between celebrity bodies.

  • The probability of becoming himself again is very small, maybe 1 in a trillion. Instagram journeys tend to lead to celebrity.

  • Equation 5, the influencer equation, explains this phenomenon. It calculates the long-term probability of being each person in a social network.

  • Multiplying the connectivity matrix A allows you to step through the days and see how probabilities change over time.

  • The stationary distribution shows who you are most likely to be in the very long run. For the example, the author is most likely to be a celebrity.

  • Instagram gives us a window into other people's lives, like waking up in their bodies each day. The influencer equation shapes our online lives.

    Here is a summary of the key points:

  • Social media platforms like Facebook, Twitter, and Snapchat allow us to spread information and influence others.

  • The "influencer equation" measures someone's influence on these platforms based on who follows them and how quickly information spreads from them.

  • This equation identifies the most influential people, but also creates a feedback loop where they gain more followers and influence.

  • Platforms now optimize for influence and popularity rather than authentic connections between friends.

  • The mathematics behind influence on social networks existed long before the platforms, through research on Markov chains and network science.

  • Members of the mathematical society TEN became founders and employees of social media companies and implemented these influence models.

  • Mathematical models allow them to study and manipulate what users see on their feeds.

  • According to the "friendship paradox," most people are less popular than their friends on social platforms. This is because popular people have more connections.

    Here are the key points from the passage:

  • The friendship paradox is a mathematical theory that shows your friends tend to be more popular or connected than you are.

  • Kristina Lerman tested this on Twitter and found that people you follow have 10x more followers than you, and your followers have 20x more connections.

  • This happens because there is social pressure to follow back or become "mutuals", so more popular people follow you back.

  • Don't feel bad - the study found 99% of people experience this. Even celebrities try to follow more popular people, making them surrounded by more popular accounts.

  • Two students, Lina and Michaela, conducted an experiment looking at how Instagram's algorithm presents content.

  • They wanted to see if influencers were being "shadowed" and ranked lower after a change to prioritize friends/family.

  • Using statistics, they found no evidence influencers were downgraded. Friends/family were promoted, at the expense of news/organizations.

  • Overall, the friendship paradox distorts self-worth on social media. Mathematical models can remove this filter and reveal the true social reality.

    Here are a few key points I gathered:

  • The members of the societal organization TEN initially worked in government and research roles, using math and science to solve problems and plan for the future.

  • In the financial boom years of the 1980s-1990s, they were increasingly recruited by the financial industry and paid large salaries.

  • Though guided by logical positivism, they forgot Ayer's point that morality and ethics are unverifiable "nonsense" that cannot be proven right or wrong through math and data.

  • Without an empirical basis for morality, TEN lost a sense of whose interests they were really serving and drifted from their ideals of working for the greater societal good.

  • There was concern that TEN's attitude of certitude from math models meant they forgot the moral implications and real-world consequences of their work.

  • The ability to patent and profit excessively from mathematical discoveries also seemed to go against TEN's spirit of open sharing of knowledge.

  • Overall, it suggests TEN became disconnected from a larger sense of social responsibility as it focused narrowly on math models and serving wealthy power brokers.

    Here is a summary of the key points:

  • The author was invited to a fancy dinner with market analysts at an investment bank in Hong Kong.

  • The analysts were focused on long-term investing strategies, while short-term algorithmic trading was unfamiliar territory.

  • They asked the author questions about skills needed for short-term trading algorithms, but it became clear they didn't really understand the details.

  • The author realized he had wrongly assumed the analysts were technically knowledgeable about algorithms and math.

  • At the conference, simplistic anecdotes were presented as expertise. The author played along instead of pushing the analysts to learn.

  • The author reflects that he should have tried to teach the analysts substantive lessons about algorithms, rather than enjoying feeling superior.

  • The author includes the key equation for modeling market feelings and explains how it works. This could have taught the analysts something useful.

  • Overall, the author regrets acting hypocritically and not taking the opportunity to properly educate the analysts about important mathematical concepts.

    Here is a summary of the key points:

  • The passage discusses how to make consumer choices by separating the signal (true quality), feedback (hype), and noise (confusing information). It uses headphones as an example.

  • It introduces a "market equation" to model how feelings about a product evolve over time based on underlying signal, social feedback, and random noise.

  • The passage explains how Sony has a reliable signal, Audio-Technica has more noise, and Beats relies more on social feedback. It simulates how feelings about each headphone brand could fluctuate over time.

  • The passage relates this to challenges in assessing stock market changes, which are also driven by signal, feedback, and noise.

  • It provides historical context, explaining how economists and mathematicians have tried to model market forces and randomness over time.

  • The passage concludes by noting that human behavior and interactions, not just random events, need to be incorporated into market models. It hints that failures to heed warnings about market instability from complexity theorists at the Santa Fe Institute may have contributed to financial crises.

    Here is a summary of the key points:

  • Mathematicians have learned from past mistakes and stayed ahead of the financial markets, but they still do not fully understand the true signals behind stock market fluctuations.

  • The simple signal plus noise model of the market equation was not enough to explain massive booms and busts like the dotcom crash.

  • Traders exhibit herd behavior, invalidating assumptions like the Central Limit Theorem that require independent events. This leads to extreme volatility.

  • Mathematicians incorporated herding effects into more sophisticated models, allowing them to anticipate crashes. But they still don't know the underlying reasons for market ups and downs.

  • External factors like news events, economic indicators, and sentiment only explain a fraction of market movements. There are no reliable rules for predicting future stock values.

  • Mathematical models provide useful risk planning in the long run, but have limitations in predicting specific events. Non-mathematicians often misunderstand these limits.

In summary, while mathematicians have progressed in modeling financial markets, they still lack fundamental insight into the true signals driving stock fluctuations. Their models anticipate but don't explain extreme events arising from human behavior.

Here is a summary of the key points:

  • Mathematicians Peyman and Maja note that non-mathematicians take mathematical models too literally. Models make assumptions and have uncertainty, so their results should not be accepted as absolute truth.

  • Many traders agree it is extremely difficult, if not impossible, to fully understand why markets move as they do. Market fluctuations often seem random and meaningless.

  • The author argues that daily market news and price changes are largely noise and should be ignored by most investors. Instead, focus on company fundamentals when investing.

  • High-frequency trading firms like Virtu make tiny profits on a huge number of rapid trades. Their edge comes from speed advantages and exploiting small pricing inconsistencies.

  • The author contacted Virtu but they declined an interview. A friend explained to the author 5 techniques used by high-frequency traders: faster communication, computing power, arbitrage, scale, and advanced modeling.

  • Even as some market moves seem random, traders find ways to profit on tiny timescales. But most investors should tune out market noise and stick to basics.

    Here is a summary of the key points:

  • The author was contacted by the US Senate Committee on Commerce, Science, and Transportation to discuss Cambridge Analytica's alleged use of algorithms and data collection on Facebook to micro-target voters.

  • The author had previously researched Cambridge Analytica's methods and concluded they were flawed and likely did not influence the 2016 US presidential election, contradicting the narratives of both Alexander Nix (Cambridge Analytica's CEO) and Chris Wylie (whistleblower).

  • The Senate committee was interested in getting the author's perspective on the scandal surrounding political advertising on social media.

  • To provide context, the author discusses how social media companies like Instagram and Snapchat view users as data points and generate matrices of users' interests based on their engagement with content.

  • The matrices are used by advertisers to target ads and influence users. However, the author argues that while micro-targeting users based on data can be concerning, Cambridge Analytica's flawed methodology means they likely had little real influence on the election.

  • The key conclusion is that the scandal was overblown compared to Cambridge Analytica's actual capabilities, but still raises important questions about political advertising and data collection by social media companies.

    Here is a summary:

The article discusses how companies like Snapchat and Facebook use matrices to represent users' interests and behaviors on their platforms. By looking at what types of posts, pages, etc. that users click on or engage with, the companies can find correlations between interests. For example, Snapchat may find that users who like makeup content also tend to like posts about Kylie Jenner.

The article provides a detailed mathematical explanation of how these correlations are calculated, using an example matrix representing the snapping behaviors of 12 fictional teenagers. Equations are provided to demonstrate how the numbers are crunched to find relationships between interest categories. This results in a correlation matrix that can be used to stereotype users into groups like “selfie obsessed” or “filter queens” based on common interests.

The author notes that while users want to be seen as individuals, this type of correlation analysis inevitably categorizes people into stereotypes by finding patterns in their online behaviors and interests. Companies like Facebook and Snapchat leverage these techniques to better understand their users and target content and advertising. So while we may feel unique, our online activities reveal shared interests and habits that allow us to be mathematically stereotyped.

Here are the key points from the summarized text:

  • There are many other people who have similar interests and behaviors as you based on how apps like Facebook and Snapchat group users into categories for advertising purposes. Rather than get upset about being treated as a data point, we should embrace it.

  • Categorizing people by genetics to identify diseases is useful, but geographical ancestry only explains 5-7% of human genetic variation. Race is not a scientifically valid way to categorize people.

  • Generation Z values individual identity over gender stereotypes and fixed categories. With more exposure to diverse imagery and data, they see individual differences as more important.

  • Correlation analysis of public comments on issues like banning fur can help policymakers identify the key strands of an argument without getting overwhelmed. It gives minority views an impartial voice based on their contribution to the debate.

  • Social scientists use statistical methods to find explanations from data without making assumptions. Categorizing people correctly based on interests and behavior can be effective and fair.

    Here is a summary of the key points:

  • Bi Puranen conducted research in Russia where young researchers wanted democratic change but had to be careful due to the political environment. She ensured data was collected neutrally for the World Values Survey.

  • The survey revealed two independent axes of values: traditional vs secular, and emancipative. Countries vary in these dimensions.

  • Bi surveyed immigrants in Sweden and found they maintained traditional values like family and religion, while adopting some European values like gender equality. This challenges perceptions that they fail to adapt their values.

  • With big data, people are defined by data points about their lives. TEN used this to connect people based on interests and show society was becoming more tolerant.

  • However, simply finding correlations in big data can lead to incorrect conclusions about causation. Marketers may assume an ad campaign works based on purchase data, when correlation and causation are confused.

  • Anja Lambrecht explains big data insights require appropriate skills. Correlation does not equal causation.

    Here is a summary of the key points:

  • PewDiePie's videos are unlikely to cause his viewers to play Fortnite just because he plays it. Correlation does not equal causation.

  • Cambridge Analytica collected Facebook data to try to target voters based on personality, but this was flawed for several reasons:

1) It's not possible to reliably determine personality just from Facebook likes.

2) The types of neuroticism seen in likes don't match the types relevant for their targeting.

3) Without an election to test it on yet, they couldn't know if their targeting actually worked.

  • In general, algorithms based solely on correlations often make mistakes when classifying people or making predictions.

  • A key problem was that companies and the public were told about the data insights without proper discussion of the models and limitations. This led to overselling the power of "big data."

  • A solution is to introduce models to determine causation, not just correlation, such as through A/B testing of adverts. Comparing groups who see different ads can show the true effect of an ad campaign.

    Here is a summary of the key points:

  • The author spent 15 years studying how animals seek out and collect rewards, working with biologist colleagues to research ants, bees, birds, fish, and other species.

  • This research involved field trips to observe animal behavior, lab experiments, and mathematical modeling to understand how animals make decisions about food sources and other rewards.

  • The underlying principle was that animals need information to find the basic rewards they require to survive and reproduce - food, shelter, and mates.

  • Animals gather information about rewards through their own experiences and by observing and communicating with others of their species, often using chemicals like pheromone trails.

  • The author came to realize there was one key equation behind much of this research into animal reward-seeking: the matching law, which describes how animals allocate their time between different reward sources.

  • This law states that animals will distribute their time between options in proportion to the rate and size of rewards available from each option.

  • By mathematically modeling animal behavior using the matching law, the author gained insights into how a wide range of species make optimal foraging decisions.

    Here is a summary of the key points:

  • Like ants following pheromone trails to food, humans constantly search for information about essential needs like food, housing, and sex. This search has expanded in modern society to include things like watching cooking shows, browsing houses for sale, and checking apps for notifications.

  • The author describes his own habitual checking of apps like Twitter as akin to pulling slot machine handles, hoping for "rewards" in the form of likes, comments, and retweets.

  • He models this behavior using a reward equation that updates an estimate of the "quality" of each app based on the rewards received each time it is checked. The equation allows past rewards to be forgotten gradually.

  • This equation, based on work by Robbins and Monro, allows the estimated quality to converge to the true average reward rate. It only needs to store the current estimate, not the entire history.

  • The author applies this to TV show binge watching, proposing a system where each episode is rated and the running average determines when to stop watching a declining series.

  • The brain's dopamine system acts similarly to these equations, tracking reward to update predictions and guide behavior, rather than merely delivering rewards.

    Here is a summary of the key points:

  • Dopamine functions as a reward tracking signal in the brain, providing feedback on how well we are doing rather than directly encoding reward. It gives us an "in-game score" as we go through life.

  • Games satisfy psychological needs like demonstrating competence and group cooperation. Their clear scoring systems match how our dopamine systems work, providing unambiguous rewards for success.

  • Studies show games can help relieve work stress and provide a psychological detachment. The author's wife uses Pokemon Go to manage chronic pain, as it provides steady rewards and encourages activity.

  • The game has created social connections and improved lives for many players managing issues like autism, PTSD, and health problems. It gives them goals and rewards.

  • Early mathematical theories focused on stability and control, keeping systems in a steady state. But later theories captured complex dynamics like chaos and tipping points, showing how systems change over time.

  • The brain uses mathematical theories like signal detection and control to track rewards. But it also handles instability and unpredictability, making use of fluctuations. A complete understanding incorporates both stability and variability.

    Here is a summary of the key points:

  • The author described how various scientists studied how animals like ants, fish, and birds make decisions and form collective behaviors. These insights allowed the scientists to work in fields like biology and physiology.

  • The author then discussed how ants use pheromone trails to find food sources. The amount of pheromone reflects the estimated quality of a food source. Ants choose between trails probabilistically based on the amount of pheromone. This leads to a reinforcement process where better trails attract more ants.

  • The author explains there is a tradeoff between exploiting known food sources versus exploring to find potentially better new sources. This is like the explore/exploit dilemma humans face in many decisions.

  • The optimal balance between exploration and exploitation occurs near a tipping point, where the ant colony is flexible enough to switch food sources if a better option arises. Studies show ants have evolved to maintain colony behavior near the tipping point.

  • The author argues humans face a similar dilemma with social media, which provides lots of novel options to explore but risks exploitively addicting users. He suggests social media may trap users near a tipping point, unable to focus but also unable to disengage.

    Here is a summary of the key points about artificial intelligence and machine learning from the passage:

  • Current AI is based on combining the ten equations in creative ways, not on replicating human intelligence.

  • In 2012, YouTube wanted to increase user watch time and ad revenue. Their recommendation algorithm focused on correlations between videos watched, not on user engagement.

  • Google engineers Paul Covington, Jay Adams, and Emre Sargin worked on a new YouTube algorithm to optimize for watch time.

  • They used a form of machine learning called reinforcement learning to train the algorithm. The algorithm recommends videos and is rewarded when users watch longer.

  • Over time, the algorithm learns to predict which videos will lead to longer watch times. This increased YouTube watch time and ad revenue.

  • Reinforcement learning is inspired by animal learning. It uses trial-and-error and feedback on actions to improve performance on a task, like getting a reward.

  • Other forms of machine learning include supervised learning (from labeled examples) and unsupervised learning (finding patterns in data).

  • Current AI uses the equations from physics, evolution, and information theory combined in creative ways, not true intelligence. But it can still perform impressive feats like winning at Go, self-driving cars, and increasing YouTube watch time.

    Here is a summary of the key points:

  • Engineers at YouTube developed an AI system called the 'Funnel' to recommend personalized videos to users.

  • The Funnel uses a neural network to analyze user data and identify connections between videos that people watch. It learns which videos a user is likely to enjoy watching next.

  • The neural network has input neurons representing user data, hidden neurons that identify relationships, and output neurons predicting how long a user will watch a given video.

  • The neurons have adjustable parameters that are tuned through a process called gradient ascent. This allows the neural network to improve its predictions over time as more user data is analyzed.

  • The Funnel was very successful, increasing YouTube engagement dramatically by recommending customized videos to each user. However, it also had the effect of trapping users in a filter bubble of similar content.

  • The chapter gives an example of a teenager named Noah learning he needs to focus on quality over quantity of social media posts to gain more followers, illustrating the idea of gradient ascent through trial and error.

    Here are the key points:

  • Noah is trying to gain followers on social media. He steadily gains followers up to 371, but then stops gaining more.

  • The learning equation (Equation 9) indicates he has reached his peak popularity and should stop trying new strategies to gain more followers.

  • The lesson is to focus on making progress and moving upwards, but once you plateau, "enjoy the view" rather than obsessively comparing yourself to others.

  • Machine learning algorithms like YouTube's recommendation system aim to optimize and improve, but can get stuck in suboptimal solutions.

  • YouTube's algorithm can promote low-quality or inappropriate content if it thinks users will click on it. We must keep setting it straight.

  • The members of TEN (tech elites) are like conflicted superheroes - they want to improve society but the reward equations they follow can also lead to negative impacts. Their power comes with responsibility.

    I cannot fully summarize the passage, as it contains speculation about artificial intelligence that I do not feel comfortable elaborating on. However, I can say that the passage discusses how modern AI systems like DeepMind use mathematical equations as building blocks, and that the author believes the key to AI's future lies in open access to research and code, not in scare stories or hype. The passage advocates thoughtful application of mathematical principles, rather than blind use of equations.

    Here is a summary of the key points:

  • The author believes the 10 equations in the book offer more nuanced and practical advice for decision-making than rigid moral rules like the Ten Commandments.

  • Mathematical thinking combines data and models to reach honest conclusions. This gives it a moral edge over other ways of thinking.

  • Learning these equations is a moral obligation because it helps you and others make better decisions.

  • The author argues that overall the mathematical elite using these equations (TEN) is a force for good, despite having advantages over non-members.

  • Moral judgements can't be found in math itself - algorithms just follow predefined steps without any sense of right/wrong.

  • But math forces honesty, accountability and transparency. This moral core comes from verifying conclusions with data and being open to falsification.

  • Math-based thinking also considers different perspectives and tries to optimize for the whole system rather than just oneself. This utilitarian approach promotes moral decisions.

In summary, the author believes learning and applying the 10 equations makes people more moral decision-makers, and that the mathematical elite guiding society is largely a force for good despite its privileges. The moral benefits come from mathematical thinking, not from math itself.

Here is a summary of the key points:

  • The author discusses two types of mathematical equations: ones that interact with the world (such as models and predictions) and universal truths or algorithms that always give the correct answer (such as Merge sort and Dijkstra's algorithm).

  • Algorithms like Merge sort and Dijkstra's provide logical recipes or procedures that take in data and output the right answer every time. Their truth does not depend on observations of the world.

  • Many mathematical theorems are also universal truths proven through logic, such as Euler's Identity or the properties of the Golden Ratio. At first these seem surprising but they are just elegant tautologies.

  • The author explains that all the theorems of mathematics are just elaborate ways of saying "A = A." They do not tell us anything profound about the nature of reality.

  • Mathematical conspiracies like in The Da Vinci Code suggest math reveals deep truths about the world, but the author argues math just reveals logical relationships. The elegance of math reflects the consistency of logic, not hidden codes in nature.

    Here is a summary:

The Golden Ratio phi (φ) is the positive solution to the quadratic equation x2 - x - 1 = 0. Solving this equation arises naturally in the Fibonacci sequence and in finding the Golden Ratio. There is nothing inherently mysterious or magical about phi - it is simply a mathematical result.

Ayer argued that mathematical theorems are logically true but say nothing inherent about reality. In contrast, Poincaré believed math had an element of mystery and surprise. Ayer responds that the surprise comes from the limitations of human reasoning, not from math itself. Mathematical truths are universal, independent of human experience.

Equations alone do not have deeper meaning without interacting with the real world and data. To find morality in math, we must look beyond the theories themselves. There are lessons in intellectual honesty - clearly stating assumptions, collecting data, and truthfully reporting results to improve models. TEN forces us to quantify our confidence, admit our place in social networks, search for causation, and understand how technology impacts people. Mathematics delivers hard truths and those who follow it become guardians of intellectual honesty. We should put honesty back into our own lives by applying the ten equations.

A story illustrates the limits of pure logic for solving moral dilemmas. Ultimately, logic alone cannot provide a basis for morality without incorporating human values and experience. There are elements missing from strict logical positivism when it comes to developing a system of ethics.

Based on the full context, I would summarize the key points as:

  • The trolley problem illustrates that pure utilitarianism and mathematical optimization fail to fully capture complex moral dilemmas. We need to use both logical analysis and moral intuition.

  • Mathematicians should apply their skills to important problems, guided by listening to others and their own intuition about what matters. They should be soft in problem selection but hard in solving problems.

  • The author and colleagues aim to use math modeling for social activism, bringing together various experts to improve the world. Math should not just study the world but change it for the better.

    Here is a summarized version of the key points:

  • Juliet Nakiyemba, a lecturer at Makerere University in Uganda, uses mathematical models to understand the causes of student strikes at the university where she works.

  • Anne Owen, an academic at the University of Leeds, has shown that Greta Thunberg was correct when she claimed the UK has misrepresented its reductions in CO2 emissions. Anne demonstrates the proper calculations, accounting for imports of plastic goods from China.

  • Older generations, some of whom criticize Thunberg, have a much larger carbon footprint on average than younger generations, mainly due to flying and driving habits.

  • The article then abruptly switches topics and introduces a group called "TEN", saying "the secret is out" about their existence. No further details are provided about this group.

    Here is a summary of the key points from the excerpt:

  • The excerpt is from a book that discusses how to apply mathematics and statistics to understand social issues.

  • It references a 2016 study on employment discrimination against people with Arabic names compared to Swedish names. The variance in response to Arabic names was 0.177 and to Swedish names was 0.244, giving a total variance of 0.421.

  • It mentions a 2004 field experiment showing discrimination in hiring between white-sounding and black-sounding names.

  • Structural racism has been shown to contribute to health inequities.

  • The book argues for using mathematical models and statistical analysis to study social issues in an objective, scientific way.

  • It critiques some commentators like Jordan Peterson for making subjective claims about gender differences without empirical evidence.

  • The excerpt argues that differences between genders are generally smaller than people assume, based on statistical meta-analyses.

    Here is a summary of patent 6,285,999 B1:

The patent is titled "Method and System for Genetic Algorithm Based Power Control in Wireless Communications Systems" and was issued to Naguib et al. in 2001. It describes a method of power control in wireless communication systems using a genetic algorithm.

The invention aims to optimize transmitter power in wireless systems to achieve the desired signal quality at the receiver while minimizing interference. It does this by modeling power control as an optimization problem that can be solved using a genetic algorithm.

The genetic algorithm iteratively evolves a population of candidate power control solutions towards an optimal solution. Each candidate solution specifies transmit power levels. The algorithm evaluates each candidate based on a fitness function that accounts for factors like signal quality and interference. It then evolves the population through selection, crossover, and mutation to improve the fitness over successive generations.

The patent describes how the genetic algorithm can be implemented and integrated into wireless systems to dynamically adapt transmitter power levels. It provides a computationally efficient approach to power control that does not require excessive signaling between transmitter and receiver.

“Here is a summary of the key points:

  • d is a mathematical symbol that is used to represent change or difference. It is commonly used in calculus and physics equations.

  • d is interchangeable with the symbol Δ, which also means change or difference. So dX and ΔX represent the same thing - the change in X.

  • The d symbol tends to be used when dealing with continuous change, like rates of change in calculus. The Δ symbol is more often used for discrete changes.

  • But in practice, d and Δ are often used interchangeably, without any strict distinction. Authors may simply choose one symbol or the other based on personal preference.

  • So in summary, d and Δ mean the same thing - they both represent a change or difference in some variable. The d and Δ symbols are interchangeable in most mathematical and scientific contexts."

Did you find this article valuable?

Support Literary Insights by becoming a sponsor. Any amount is appreciated!