New Game Theory Model Incorporates Variable Rewards

Scientists have developed a new mathematical model to study game theory that incorporates dynamically evolving strategies and randomly changing rewards. As reported by Ars Technica, this approach helps researchers better understand why people make specific decisions in real life.

Traditional games typically take place under static conditions where the reward for each outcome is constant. However, in the real world, the consequences and payoffs of strategic choices constantly change, which limited the practical application of static models.

Prisoner's dilemma and the new approach

One of the most famous examples in game theory is the "Prisoner's Dilemma". In this scenario, law enforcement officers interrogate two detainees separately. If both prisoners remain silent (co-operate with each other), they are sentenced for a less severe crime. If one confesses and betrays the other (defection), the confessor goes free, while the other receives the maximum sentence. If both confess, both receive a sentence of medium severity.

The game organiser can alter the rewards and risks for co-operation and defection to determine how the optimal strategy changes over several rounds. In traditional settings, where the balance between co-operation and defection is fixed, the game often ends with all players choosing defection, which ultimately leads to everyone losing.

By introducing the new mathematical model, researchers can now observe how players' behaviour changes when rewards are not constant, but vary randomly. This allows scientists to study game theory using data that is much closer to a real, changing environment.