Coherence versus Correspondence
In decision science, the terms “coherence” and “correspondence” refer to two different criteria for evaluating the quality or correctness of, beliefs, judgments, or decisions.
“Coherence” relates to their internal consistency. A coherent judgment is one that logically aligns with a person’s other beliefs or with accepted rules of logic and probability theory. For instance, if one believes that A is more likely than B, and B is more likely than C, then to be coherent, one must also believe that A is more likely than C.
“Correspondence”, on the other hand, refers to the extent to which one’s beliefs, judgments, or decisions match with external reality. A judgment is said to correspond if it aligns with real-world outcomes or objective facts. For example, a belief that it will rain tomorrow corresponds with reality if it indeed rains the following day.
The distinction between coherence and correspondence traces back to the foundational work of philosopher and decision theorist, Daniel Ellsberg, in the 1950s. Ellsberg argued that while traditional decision theories emphasized coherence (e.g., BayesianBayes' theorem refers to a probabilistic framework for updating beliefs in light of new evidence. Initially derived from the work of the Reverend (and statistician... decision theory), real-world decision-making often demands judgments that correspond with reality. This distinction underscores the tension between internal logical consistency and real-world accuracy in judgment and decision-making research.
The distinction has been central to scholars critical of the sometimes-automatic ascription of the term “bias” to decision processes that rely on heuristics and that can be demonstrated to deviate from classical normative models of rational choiceLink to Wikipedia entry More. This line of criticism points out that in their justified effort to demonstrate people use heuristics, decision scientists have designed experiments where the outcomes of a heuristic conflict with normative models, with the unintended consequence of emphasizing the the error-proneness of heuristics. In real-world settings, heuristic often works quickly and effectively and attempts to apply normative models—such as by estimating probabilities or utility—may be fraught with error. For example, estimating probabilities, potential outcomes, and utilityIn decision science, "utility" refers to a quantitative measure representing an individual's preference or satisfaction with respect to a particular outcome. In this framework, utility... when deciding whether or not to have an affair, as would presumably be prescribed by subjective expected utility theory, may be fraught with a lack of long-term perspective or unconscious motivations, whereas simply following a respected authority’s moral rule that it is wrong to have an affair may result in higher expected utility (though of course history shows that infidelity is not so easily reigned in by either cost-benefit analysis or moral heuristics).
Ecological and Adaptive Rationality
The above critical view of the heuristics and biasesIn decision science, bias refers to systematic and predictable deviations from accepted models of rationality. In scientific language, "systematic" and "predictable" are not meant to... tradition—with its emphasis on the distinction between coherence and correspondence—has been a centerpiece of work by Gerd Gigerenzer and his colleagues, in criticizing both normative decision theorists in economics and behavioral decision theorists in psychology and behavioral economics. They use the more specialized terms ecological rationality and adaptive rationality to emphasize particular aspects of correspondence and to make the argument that it is a far more relevant standard for evaluating the quality of decisions than coherence.
Ecological Rationality
Ecological rationality refers to an assessment of how well a heuristic performs at achieving a desired outcome given particular a particular environment. In this approach, it is not meaningful to discuss whether or not a heuristic is ecologically rational without also describing when and where it is used. Evaluating ecological rationality is about examining how well heuristics work in different environments as well as which heuristics tend to actually be used in distinct real-world environments.
Adaptive Rationality
Adaptive rationality emphasizes the idea that how we evaluate the rationality of a judgment or decision should be based on whether or not the process is adaptive rather than whether or not it is consistent or otherwise coherent. Central to this view is the idea that decision processes are products of evolutionary processes and that the standard of success in evolution is success rather than truth or logic. With that in mind, the expectation that drives much of the research emphasizing adaptive rationality is that in real-world environments—rather than experimental contexts designed to show where heuristics go wrong—heuristics will tend to be ecologically rational; that is, the heuristics people use will tend to fit well with their environments. Much of the emphasis is on biological evolution and the idea that there may be a set of non-heuristic, cognitive building blocks that can be used to build a large toolbox of different heuristics for different contexts, much like amino acids are used as building blocks to create a nearly unlimited variety of proteins. While the role of individual and cultural adaptation in building and selecting heuristics tends to get less attention than biological adaptation, their language is careful not to exclude these other forms of learning and adaptation, and occasionally they explicitly acknowledge the importance of these other paths of adaptation. This concept challenges classical notions of rationality that often demand complex computations for optimal decisions, arguing instead for a model of bounded rationalityThe concept of bounded rationality refers to the idea that human rationality is constrained by (1) available information, (2) cognitive / information processing limitations, and... where simplifying heuristics can lead to effective, real-world decision-making under uncertainty and with limited information and limited time.
Relevance to Casino Gambling
This distinction is particularly important to gambling decision making, and arguably to almost all decision making in the wildDecisions from Experience vs Decisions from Description "Decisions from description" and "decisions from experience" are two paradigms in the experimental study of behavioral decision making...., and many of the Substack posts are—or will be—about this topic, emphasizing both (1) the adaptiveness of heuristics and (2) how maladaptive trying to apply normative models can be:
- Many gambling strategies (heuristics) come as close to maximizing expected utility as might be expected, given bounded rationality, despite false beliefs about outcome probabilities and expected valueExpected value or EV is the average return one can expect to get from a repeated decision, such as from a gamble, if it is...,
- On the other hand, gamblers who attempt to calculate outcome probabilities and expected value—as might be prescribed by normative decision theoryDecision scientists often distinguish between normative and descriptive models of decision making, and less often add the additional distinction between normative and prescriptive models. Normative... models—often end up using dangerously misguided gambling strategies.
Specific examples in blackjackBlackjack, also referred to as 21, is a popular casino card game where players attempt to have a hand value closer to 21 than the... More, rouletteRoulette is a popular casino table game originating from France. Players place bets on either a single number, various groupings of numbers, the colors red... More, and slot machinesSlot machines, also known as a fruit machines (e.g., in the United Kingdom), pokies (e.g., in Australia), one-armed bandits, or simply slots, are games of... More will be discussed in the Substack.