Probability
Probability expresses uncertainty about outcomes. Jones emphasizes checking the assumptions behind estimates, combining probabilities with the appropriate rules, and communicating numerical judgments without implying more precision than the evidence supports.
THE IDEA
Probability and estimation
Estimating is what we do when we run out of data, Jones writes. Probability gives estimates a numerical language, and its rules constrain how those numbers can be combined. Choosing relevant data and judging whether a model fits the problem still require care; a correct calculation can rest on poor assumptions.
Probability can be estimated from a model, observed frequencies, experience, or a combination of them. These are sources of information rather than a strict division between problems with data and problems without it.
For mutually exclusive events, the probability of an “or” outcome is the sum of their probabilities. If the events can overlap, that overlap must be subtracted to avoid counting it twice. For an “and” outcome, multiply the probability of the first event by the probability of the second conditional on the first. Multiplying the two unconditional probabilities is appropriate when the events are independent.
Equal probabilities are justified by a suitable symmetry assumption, not simply by an absence of knowledge. When evidence does not support a precise estimate, state the uncertainty and examine how different plausible estimates would affect the decision.
In a probability tree, the branches leaving any node should describe mutually exclusive and collectively exhaustive alternatives: they cannot occur together, and together they cover the possibilities. Their conditional probabilities must sum to 1.0. Events at successive levels of the tree need not be mutually exclusive; a path represents events occurring together.
Words such as likely and unlikely can mean different things to different readers. Numerical percentages make a probability judgment more explicit, but excessive detail can suggest precision the analysis does not support. Explain the basis and limits of the estimate.
Related concept: utility
The extraction also introduces chapter 14, on utility. Utility represents the value an outcome has to the person making the choice. Someone may value a car for convenience, independence, or the journeys it makes possible. A career choice may involve salary, satisfaction, working hours, and prospects for advancement.
Separating probabilities from utilities helps clarify a decision: how likely is each outcome, and how much does each outcome matter? People may value the same outcome differently. Identifying those values is useful without assuming that every actual choice perfectly maximizes them.
When to use it
Estimating the likelihood of project success based on historical data and subjective judgment.
Making investment decisions by calculating expected returns and considering risk probabilities.
Analyzing personal career choices by identifying the utilities (benefits) of each option and comparing them.
Communicating uncertainty in reports by using numerical probabilities carefully to avoid false precision.
FROM THEORY TO EVERYDAY LIFE
In practice
Project risk assessment
A project manager uses a probability tree to estimate the likelihood of completing a project on time. She assigns probabilities to events like supplier delays, technical issues, and budget cuts, then calculates the overall probability of success. This helps her communicate risks to stakeholders.
Career choice evaluation
When deciding whether to stay in a current profession or switch to a new one, an individual weighs the utilities of each option: salary, job satisfaction, work-life balance, and growth opportunities. Despite uncertainty about future outcomes, they choose the path offering the highest expected utility.
A MOMENT TO REFLECT
Put it to work.
How can you use probability trees to make better decisions in your professional life?
Think of a recent decision where you estimated probabilities. What biases might have influenced your estimates?
How does the concept of utility explain why different people make different choices in the same situation?
CONTINUE THE CONVERSATION
Source texts
- The Thinkers Toolkit by Morgan D Jones