Summary of Scott E. Page's The Model Thinker

Everest Media

Description

Please note: This is a companion version & not the original book.
Sample Book Insights:
#1 Models are formal structures represented in mathematics and diagrams that help us understand the world. They improve our ability to reason, explain, design, communicate, act, predict, and explore.
#2 A model is a simplification of the world that can be applied within it. It must be simple enough that within it we can apply logic. A model is a collection of models that accomplishes even more.
#3 We have access to unprecedented amounts of data, but we are not capable of understanding why certain things happen. Empirical findings may be misleading. Data on piece-rate work shows that the more people are paid per unit of output, the less they produce.
#4 Models are used to make sense of the firehose-like streams of data that cross our computer screens. Without models, people suffer from a laundry list of cognitive shortcomings: we overweight recent events, we assign probabilities based on reasonableness, and we ignore base rates.

Sujets

• Data Science

Informations

• Publié par: Everest Media LLC
• Date de parution: 28 mars 2022
• Nombre de lectures: 0
• EAN13: 9781669369295
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Insights on Scott E. Page's The Model Thinker
Contents Insights from Chapter 1 Insights from Chapter 2 Insights from Chapter 3 Insights from Chapter 4 Insights from Chapter 5 Insights from Chapter 6 Insights from Chapter 7 Insights from Chapter 8 Insights from Chapter 9 Insights from Chapter 10 Insights from Chapter 11 Insights from Chapter 12 Insights from Chapter 13 Insights from Chapter 14 Insights from Chapter 15 Insights from Chapter 16 Insights from Chapter 17 Insights from Chapter 18 Insights from Chapter 19 Insights from Chapter 20 Insights from Chapter 21 Insights from Chapter 22 Insights from Chapter 23 Insights from Chapter 24 Insights from Chapter 25 Insights from Chapter 26 Insights from Chapter 27 Insights from Chapter 28 Insights from Chapter 29
Insights from Chapter 1

#1

Models are formal structures represented in mathematics and diagrams that help us understand the world. They improve our ability to reason, explain, design, communicate, act, predict, and explore.

#2

A model is a simplification of the world that can be applied within it. It must be simple enough that within it we can apply logic. A model is a collection of models that accomplishes even more.

#3

We have access to unprecedented amounts of data, but we are not capable of understanding why certain things happen. Empirical findings may be misleading. Data on piece-rate work shows that the more people are paid per unit of output, the less they produce.

#4

Models are used to make sense of the firehose-like streams of data that cross our computer screens. Without models, people suffer from a laundry list of cognitive shortcomings: we overweight recent events, we assign probabilities based on reasonableness, and we ignore base rates.

#5

The traditional approach to modeling is one-to-one: one problem requires one model. Many-model thinking challenges that approach. It advocates trying many models.

#6

The models in this book are meant to be used together. They are designed to simplify and formalize, which allows them to create tractable spaces within which we can work through logic, generate hypotheses, and fit data.

#7

We need models to think coherently, but any single model with a few moving parts cannot make sense of high-dimensional, complex phenomena such as patterns in international trade policy, trends in the consumer products industry, or adaptive responses within the brain.

#8

The wisdom hierarchy starts with data, which are raw, uncoded events, experiences, and phenomena. At the bottom of the hierarchy lie data. Information names and partitions data into categories.

#9

Knowledge organizes information. It is often model form. Wisdom is the ability to identify and apply relevant knowledge. It requires many-model thinking.

#10

We can use models to make decisions, and we can use them to check the internal consistency of different accounts that may explain the financial collapse of 2008.

#11

The Cuban missile crisis was explained by three models. A rational-actor model, which showed that Kennedy had three possible actions: start a nuclear war, invade Cuba, or impose a blockade. He chose the blockade. The model neglects to add a stage in which the Soviets put the missiles in Cuba.

#12

Allison’s book shows the power of models alone and in dialogue. Each model clarifies our thinking. The rational-actor model identifies possible actions once the missiles have arrived, and it allows us to see the implications of those actions. The organizational model draws our attention to the fact that organizations, not individuals, carry out those actions.

#13

Many-model thinking is not about dividing the system into independent parts. It is about partially isolating the major causal threads and then exploring how they are interwoven. In doing so, we will find that the data produced by our economic, political, and social systems exhibits coherence.

#14

We live in a time where we have access to an abundance of information and data. The same technological advances that generate those data make economic, political, and social actors more agile and capable of responding to economic and political events in an instant.
Insights from Chapter 2

#1

Models are used to explain data, predict, design, and take actions. They can be used to explore ideas and possibilities, and they can communicate ideas and understandings.

#2

When constructing a model, we can choose to aim for realism and follow an embodiment approach, or we can take an analogy approach and abstract from reality.

#3

The term tractable refers to how amenable something is to analysis. In the past, analysis relied on mathematical or logical reasoning. Today, when we run up against the constraint of analytic tractability, we can turn to computation.

#4

Arrow’s theorem is an example of how logic can reveal impossibilities. It states that if individual preferences aggregate to form a collective preference, then no collective ordering necessarily exists.

#5

The paradox of popularity is that highly popular people have more friends. However, most people’s friends are also better-looking, kinder, richer, and smarter than they are.

#6

The power of conditionality is revealed when we contrast claims derived from models with narrative claims. Narrative claims may have empirical support, but they are still only assumptions. Models provide conditions under which those claims become true.

#7

Theorems and proverbs are two different things. While opposite proverbs abound, opposite theorems cannot. Within models, we make assumptions and prove theorems. Two theorems that disagree on the optimal action, make different predictions, or offer distinct explanations must make different assumptions.

#8

Models are logical explanations for empirical phenomena. They can explain point values and changes in their values. They can explain the current price of pork belly futures and why prices rose over the past six months.

#9

Models aid in design by providing frameworks within which we can contemplate the implications of choices. They are used by engineers to design supply chains, by computer scientists to design web protocols, and by social scientists to design institutions.

#10

Models improve communication. They require formal definitions of the relevant features and their relationships that can then be communicated with precision. When we formally define an abstract concept like political ideology using a reproducible methodology, those concepts take on some of the same features as physical qualities such as mass and acceleration.

#11

Models are used by governments, corporations, and nonprofits to guide actions. They are linked to data. The Federal Reserve gave $182 billion in financial assistance to bail out the multinational insurance company American International Group in 2008, based on models.

#12

Models that guide action are not limited to organizations. People can also use models to think through important actions in their personal lives. When deciding to purchase a home, take a new job, or return to graduate school, we can use models to guide our thinking.

#13

Models can predict individual events as well as general trends. They can explain and predict, but they differ in that they can explain without predicting. Models that explain but cannot predict are like bomb-sniffing dogs.

#14

We use models to explore intuitions and possibilities. We can apply a model in practice any of several ways. We may use it to explain, predict, and guide action.
Insights from Chapter 3

#1

The many-model approach requires us to learn many models, but we do not need to learn as many as we might think. We can apply any one model to many cases by reassigning names and identifiers and modifying assumptions.

#2

The Condorcet jury theorem is a model that explains the advantages of majority rule. It states that by constructing multiple models and using majority rule, we will be more accurate than if we used one of the constituent models.

#3

The Condorcet jury theorem states that a majority vote classifies correctly with higher probability than any single person. It applies to models that make numerical predictions or valuations.

#4

The diversity prediction theorem states that if we could construct a diverse set of moderately accurate predictive models, we could reduce our many-model error to near zero. However, our ability to construct many diverse models has limits.

#5

We can apply the Condorcet jury theorem to categorization models. These models provide micro-foundations for the Condorcet jury theorem. The key insight is that the number of relevant attributes constrains the number of distinct categorizations, and therefore the number of useful models.

#6

There exists a set of objects or states of the world, each defined by a set of attributes and a value. A categorization model, M, partitions these objects into a finite set of categories based on the object’s attributes and assigns valuations for each category.

#7

The theorems of many-model thinking validate the logic of many-model thinking. They do not, and cannot, construct the many models that meet their assumptions. In practice, we may find that we can construct only three or five good models.

#8

The four uses of models - to reason, explain, communicate, and explore - require simplification. To apply logic and explain phenomena, we must simplify. The more data we have, the more granular we should make our model.

#9

A categorization model would divide the households into categories and estimate a value for each category. A more granular model would create more categories. As we add more categories, we can explain more of the variation, but we can go too far.

#10

The one-to-many approach is new, and it allows you to master a modest number of flexible models and a