Decision theory

Sometimes we must choose between many things.

We can think about what might happen next. This helps us pick the best way.

Some people use math to help them decide. They look at what they might gain. They also look at what they might lose. This can be hard when we do not know the future. We might even think a win is "due" when it is not.

Learning to choose well is a big part of life.

Sometimes we must choose between many different paths.

Decision theory is a way to study these choices. It uses math to help us pick the best path. This is hard when we do not know what will happen next. This is called uncertainty.

Long ago, thinkers like Blaise Pascal studied chance. In the 1700s, Daniel Bernoulli looked at risk. He thought about how much a person values a prize. This is called expected utility. It means looking at what you might gain or lose.

People often make mistakes when they choose. They might use a shortcut called a heuristic. One mistake is the gambler's fallacy. This is thinking a win is "due" because of what happened before. But each turn is still a new chance.

Some people also study how we choose over time. You might choose a small prize now or a big prize later.

Military planners even use these ideas to study how groups act. They use math to model how people behave in big groups.

Every day, we face choices that involve uncertainty. We might choose between different snacks or decide how to spend our time. Decision theory is a way to study these choices using math. It looks at how a person might act to make the best possible choice. This field uses probability and economics to model smart behavior. It is different from science that just describes how people act. Instead, it is prescriptive, which means it suggests how people should act to be rational.

To make a smart choice under uncertainty, you can use expected value. First, you look at all the possible things that could happen. Next, you find the value of each outcome, whether it is positive or negative. Then, you look at the probability, or the chance, of each thing happening. You multiply the value by the chance to find the expected value. This gives you an average expectation for the outcome. A rational person should pick the path with the highest total expected value.

The history of these ideas goes back hundreds of years. In the 17th century, Blaise Pascal and Pierre de Fermat developed probability theory. This helped people understand risk and uncertainty. In 1738, Daniel Bernoulli wrote a paper about measuring risk. He used a Dutch merchant to show that people care about more than just money. He introduced the idea of expected utility to explain this. Later, in the 1940s, John von Neumann and Oskar Morgenstern helped create modern Game Theory.

There are many different branches of this study. Normative decision theory looks for the perfect, ideal decision. This often uses tools called decision support systems to help people. Descriptive decision theory tries to describe how people actually behave. Some people also study intertemporal choice, which is picking between rewards now or later. For example, you could spend money on a holiday today or save it for a pension later. This involves looking at things like interest rates and how much you trust the future.

Even with math, humans often make mistakes called heuristics. Heuristics are mental shortcuts that help us decide quickly. However, these shortcuts can lead to errors like the gambler's fallacy. This is the mistaken belief that a random event is affected by what happened before. For instance, if a coin lands on tails many times, someone might think heads is "due." In reality, the chance stays the same every time. Scholars like Daniel Kahneman and Amos Tversky studied these patterns to understand real human behavior better.

Decision theory is an interdisciplinary field that combines probability, economics, and analytic philosophy. It uses mathematical models to understand how individuals behave when facing uncertainty. Rather than just describing how people act, decision theory is often prescriptive. This means it aims to identify the optimal decisions for a rational agent. A rational agent is an ideal decision maker who calculates outcomes with perfect accuracy. By studying these models, social scientists can better understand human behavior in complex fields. These include sociology, criminology, moral philosophy, and political science.

At the heart of this subject is the concept of choice under uncertainty. To make a rational choice, one must follow a specific mathematical procedure. First, a person identifies all possible actions and their potential outcomes. Each outcome is assigned a value, which can be positive or negative. Next, the person determines the probability of each outcome occurring. The rational step is to multiply each value by its probability to find the expected value. This represents the average expectation for that specific action. A person should choose the action that results in the highest total expected value.

The history of these ideas began in the 17th century with probability theory. Mathematicians Blaise Pascal and Pierre de Fermat developed early frameworks for understanding risk. In 1738, Daniel Bernoulli published a paper that changed how we view risk. He used the St. Petersburg paradox to show that expected financial value is not enough. He introduced the concept of expected utility to better measure risk. Bernoulli used the example of a Dutch merchant deciding whether to insure cargo. This showed that the personal value of money matters more than the amount itself.

Modern decision theory was further shaped in the 1940s by John von Neumann and Oskar Morgenstern. They formalized expected utility theory and helped establish the foundations of Game Theory. Following World War II, the field expanded into economics through scholars like Milton Friedman. They applied these mathematical models to consumer choice and market behavior. Later, Abraham Wald's 1939 paper helped synthesize these concepts with statistical theory. Wald linked decision problems to hypothesis testing and parameter estimation. This revival included new concepts like loss functions and minimax procedures.

There are several distinct branches within the field. Normative decision theory focuses on finding the best possible decisions through ideal logic. This branch leads to decision analysis, which creates software known as decision support systems. In contrast, descriptive decision theory attempts to describe how people actually behave. It looks for consistent rules in human actions, such as procedural or axiomatic frameworks. Some researchers also study intertemporal choice, which involves decisions made over different stages of time. This includes choosing between immediate rewards or future benefits, like a holiday today versus a pension later. The outcome depends on interest rates, inflation, and life expectancy.

Despite these mathematical models, human behavior often deviates from rational predictions. Scholars like Maurice Allais and Daniel Ellsberg found systematic departures from expected utility. This led to the development of prospect theory by Daniel Kahneman and Amos Tversky. They discovered that humans use heuristics, which are mental shortcuts for quick decisions. While fast, these shortcuts can lead to cognitive biases. One example is the gambler's fallacy, where people believe a random event is influenced by previous results. For instance, someone might think a coin is "due" to land on heads after many tails.

Decision theory also explores the complexities of social and organizational environments. In social decisions, one must account for how other people will respond to an action. This is often studied through mathematical methods in fields like socio-cognitive engineering. Military planners use these principles during extensive simulations to predict actor behavior.

Additionally, some decisions are difficult because of bounded rationality. This occurs when individuals have limited time or intelligence to process complex information. The way options are framed can also change a person's choice, a phenomenon known as the distinction bias. Understanding these layers helps bridge the gap between mathematical ideals and real-world complexity.