This week our goal is to learn the tools and methods for improving our individual prediction accuracy, and to practice those skills. We cannot hope to navigate complex decisions unless we first understand how to correctly analyze an event, and to have reasonably well-calibrated expectations about the future.
Video Lectures [30 minutes]
Watch these three lectures during class time. Write down any questions and post them in your cohort channel so that the lecturer can address them asynchronously.
- Thinking in Buckets briefly explains the heuristic of breaking any prediction down into categories for efficiency.
- Watch this video by 3Blue1Brown on a better version of Bayes Rule and how to use odds ratios to more efficiently update your prior.
- Probabilities and Distinctions discusses the rules of constructing probability trees and using them to solve confusing problems.
Calibration Game [20 minutes]
Individually play the Metaculus Calibration Game for ten minutes. Then play for a further ten minutes collectively, sharing your thoughts and perspectives to try to come to a group consensus.
Metaculus [20 minutes]
Make an account on Metaculus and make one prediction every day for the next week. You are encouraged to collaborate with your cohort on these predictions; at the end of this course, the cohort with the best calibration will be victorious. For now, each group member should peruse Metaculus quietly, select a question, and make a prediction. Allocate 10 minutes for this process. Then come back to the cohort and take turns discussing the question you chose, what your prediction was, and your reasoning for it.
Personal Decision [20 minutes]
Think of a decision that you’re facing which is not straightforward, and which you wouldn’t mind sharing and discussing with your cohort.
For this exercise, simply break the space of possible outcomes of the decision down into a small number of bins that make sense, according to the "Thinking in Buckets" video instructions. Discuss your reasoning for how you allocated the buckets with your cohort.