The broken promise of game theory

The broken promise of game theory

11 Nov, 2025 at 10:07 | Posted in Economics | Leave a comment

Game theory is, like mainstream economics in general, model-oriented. There are many reasons for this – the history of the discipline, ideals borrowed from the natural sciences (especially physics), the search for universality (explaining as much as possible with as little as possible), rigour, precision, and so on. Most mainstream economists and game theorists seek to explain social phenomena, structures, and patterns based on the assumption that agents act in an optimising (rational) way to satisfy given, stable, and well-defined goals.

When building their models, modern mainstream economists ground them on a set of core assumptions (describing agents as ‘rational’ actors) and a set of auxiliary assumptions. Based on these, they attempt to explain and predict both individual and social phenomena.

The model is typically seen as a kind of thought-experimental ‘as if’ benchmark device. Its purpose is to enable a rigorous, mathematically tractable illustration of social interaction in an ideal-type model world, and to allow a comparison of that ‘ideal’ with reality. The ‘interpreted’ model is supposed to supply the analytical and explanatory power, enabling us to detect and understand mechanisms and tendencies in real-world economies.

However, for the models to be relevant, we must also be able to argue that their precision and rigour hold when applied to real-world situations. Often, they do not. When addressing real economies, the idealisations and abstractions necessary for the deductivist machinery to work simply do not hold. If the real world is fuzzy, vague, and indeterminate, then why should our models be built on a desire to describe it as precise and predictable? Being told that a model is rigorous and amenable to ‘successive approximations’ of reality is of little avail, especially when its core, law-like (nomological) assumptions are highly questionable and extremely difficult to test.

Many mainstream economists still maintain that game theory is useful, can be applied to real life, and yields important results. That, however, is a rather unsubstantiated view. Strictly speaking, what game theory does is nothing more than investigate the logic of behaviour among non-existent robot-imitations of humans. Knowing how these ‘rational fools’ play games does not help us to decide and act when interacting with real people. Knowing some game theory may actually make us behave in a way that hurts both ourselves and others. Decision-making and social interaction are always embedded in socio-cultural contexts. By failing to account for this, game theory remains an analytical cul-de-sac, unable to produce useful and relevant explanations.

Over-emphasising the reach of instrumental rationality and abstracting away from the influence of many known important factors reduces the analysis to a pure thought experiment without any substantial connection to reality. Limiting theoretical economic analysis in this way – by not incorporating both motivational and institutional factors when trying to explain human behaviour – makes economics insensitive to social facts.

Game theorists extensively exploit ‘rational choice’ assumptions in their explanations. This is probably also the reason why game theory has been unable to accommodate well-known anomalies within its theoretical framework. That should hardly come as a surprise. With its axiomatic view of individuals’ tastes, beliefs, and preferences, game theory cannot accommodate much real-life behaviour. It is hard to find compelling arguments for continuing down its barren paths, since individuals obviously do not comply with, nor are guided by, game theory.

Apart from a few notable exceptions, it is difficult to find truly successful applications of game theory. Why? Largely because the boundary conditions of game-theoretical models are false and baseless from a real-world perspective. And, perhaps even more importantly, since they are not even close to being good approximations of real life, game theory lacks predictive power. This should come as no surprise. As long as game theory sticks to its ‘rational choice’ foundations, there is little hope for improvement.

Applications of game theory have on the whole resulted in massive predictive failures. People simply do not act according to the theory. They do not know or possess the assumed probabilities, utilities, beliefs, or information required to calculate the different (‘subgame’, ‘trembling hand perfect’) Nash equilibria. They may be reasonable and use their cognitive faculties as well as they can, but they are obviously not the perfect, costless, hyper-rational, expected-utility maximisers that game theory posits. And fortunately so. Being ‘reasonable’ allows them to avoid the made-up ‘rationality’ traps that game theory would have ensnared them in, had they tried to act as consistent players in a game-theoretical sense.

Game theorists can, of course, marginally modify their toolbox and fiddle with the auxiliary assumptions to get whatever outcome they want. However, as long as the core ‘rational choice’ assumptions are left intact, it seems a pointless effort to hamper an already excessive deductive-axiomatic formalism. If you do believe in the real-world relevance of game-theoretical ‘science fiction’ assumptions—such as expected utility, ‘common knowledge’, ‘backward induction’, and correct and consistent beliefs—then adding concepts like ‘framing’, ‘cognitive bias’, and different heuristics does not ‘solve’ any fundamental problem. If we want to construct a theory that can provide us with genuine explanations of individual cognition, decisions, and social interaction, we have to look for something else entirely.