- Why model?
- Outline
- Online Course Format
- Section Structure
- Reason #1: Intelligent citizen of the world
- Reason #2: Clearer thinker
- Reason #3: Understand and use data
- Reason #4: Decide, strategize, and design
Outline: topics are independent
- Collective action
- Wisdom of crowds
- Tipping points
- Markov processes
- Game theory (Colonel Blotto)
- Spatial voting
- Lyapunov Models
- Decision theory
- Linear models
- Learning models
- Economic growth
- Videos - 8-15 minutes, and questions
- Readings - linked on Wiki
- Assignments
- Quizzes
- Discussion Forum
- The model - assumptions, results, applications
- Technical details - measures, proofs (easy, medium, hard), practice problems
- Fertility - where does the model work?
Lecture 2- Intelligent Citizens of the World
Models are simplifications, and abstractions. They are wrong, but are useful.
"essentially, all models are wrong, but some are useful" ~ George E. P. Box ~
Grate books movement includes hundreds of books with thousands of ideas.
One of the favorite idea is "Tie yourself to mass."
Models gives us conditions of the phenomena we try to understand and explain.
Models tie us to the mass of logic.
- Economics: A description of service, utility function for an agent to maximize their payoff.
- Biology: Models of the brain, genes, and species.
- Sociology: How does your identity affect your actions and behaviors.
- Political science: Understand the voting behaviors.
- Linguistics: Verbs, nouns, and structure of languages.
- Law: Network model of how they appoint supreme court judges
- Game theory: Strategic behaviors
- Models: Accuracy of models (Calibration) vs. Precision of models (Discriminatory)
- Prediction: Based on Philip E. Tetlock's book*, thinking styles such as hedgehogs (people use single models) is worse than foxes (people use lots of models), and foxes are worse than formal models (people use years of data)
- Smart people like Bruce** use models, blend with their experiences. They don't just follow models.
Models developed in one domain may be applied to other domains.
- Markov process used to predict dynamic processes such as disease spread. It can also used to predict who wrote a book.
Sometimes we can predict correctly only after we constructing a model.
Model help us understand what's going on.
- Tulip mania in 1637.
- Home price crashes.
- People with lots of formal models predict the best, and make sense of the world.
- The only people who are actually better than the random choices at predicting what will happen are people with multiple models.
* Philip E. Tetlock, Expert Political Judgment: How Good Is It? How Can We Know?, 2006.
Expert Political Judgment: How Good Is It? How Can We Know? (Chapter 1)
** Bruce Bueno de Mesquita
"The New Nostradamus", Good Magazine Issue 007.
Lecture 3- Thinking More Clearly
Models an help us become clear thinkers.
Clear Thinker
1. Name the Parts: only the relevant parts.
Example: How people choose restaurant? Prices, money, time, and preferences.
2. Identify Relationships:
Example: Extensive foreign game in game theory.
3. Work Through Logic:
Example: Make a ring 1 meter larger than the radius of Earth.
4. Inductively Explore:
Example: A post prevent people jam in the exit.
5. Understand Class of Outcome:
Four possible outcomes of a model- equilibrium, cycle, random, and complex.
Example: Price of the oil.
6. Identify Logical Boundaries:
Example: Models help up find the conditions of the opposite proverbs.
7. Communication:
Example: Use likability and policy space to explain how people vote certain candidates.
