Probability, Confidence Intervals, and Clear Problem Definitions
Summary
This essay explains classical, frequency-based, and subjective interpretations of probability, then introduces the axiomatic probability space. It emphasizes that probability describes uncertainty relative to a defined model: a scenario must specify its possible outcomes and the assumptions that assign their probabilities. Examples involving dice, repeated coin tosses, a host revealing a door, and randomly drawn chords show how different assumptions can change the answer.
The discussion distinguishes the long-run coverage meaning of a confidence level from the interpretation of a particular interval after data have been observed. It also explains how repeated sampling, large-sample behavior, hypothesis tests, and sample size inform inference, while stressing that data do not prove an assumed probability true. The article is conceptual rather than a trading method, and its examples are pedagogical; it does not provide empirical market evidence or a practical estimation procedure for trading data.
Key ideas
- Probability can be interpreted through equally likely outcomes, observed relative frequencies, or informed subjective judgments.
- A probability model needs a sample space, a suitable set of events, and a probability assignment.
- A confidence level describes the long-run coverage of a procedure across repeated samples, not the probability that a fixed observed interval contains a fixed parameter.
- The same observed frequency can provide different evidence depending on the number of observations.
- Ambiguous assumptions about how outcomes are generated can lead to different probability answers.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.