**10. Conclusions**

In this work, we demonstrate the estimation of a new learning model based on an existing decision theory, Case-based Decision Theory. This form of decision-making under uncertainty when applied to game theoretic experiments performs well when compared to two other leading learning models: Reinforcement Learning and self-tuning Experience Weighted Attraction. An important feature of Case-Based Learning is the ability to systematically incorporate information that is available to subjects into choice decisions. Real people condition their behavior on their observations of their environments, and the case-based approach incorporates this in a natural way.

The parameters of Case-Based Learning indicate a relationship between recency and the type of game played. Constant sum games exhibit a smaller recency effect than non-constant sum games. This indicates that subjects weight experience with a opponents differently, depending on the type of game played. Significant attention is given to the average rates of play, and changes to those average rates of play, through the inclusion of moving averages in the definition of the 'Problem'.

Further work in applying CBL to other decision making environments is important in understanding its limits and sensible parameterizations of information vectors, hypothetical references, and how deliberate thoughtful decisions are affected by institutions that include information. CBL could also be used and adapted to predict behavior across different types of games, or more generally changes to decision-making environments, to understand the influence of previous game play on decisions in new environments and how subjects encode information across games. A natural extension of this type of investigation are repeated prisoner dilemma games or repeated coordination games, which allow more complex equilibria, but also allow for a greater freedom to explore the primitives of learning across environments.

**Author Contributions:** T.G. performed the formal analysis; T.G. and A.D.P. contributed equally to all other parts of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
