I just started looking at The Place of Probability in Science, ed. by Ellery Eells and James H. Fetzer (Springer, 2010). There’s a potentially interesting paper “Learning to Network” by Brian Skyrms and Robin Pemantle that uses the principles of game theory to theorize about the evolution of interaction networks. But I got sidetracked by the following passage:
“In the simplest belief learning model, Cournot dynamics, an individual assumes that others will do what they did last time and performs the act that has the highest payoff on that assumption. More sophisticated individuals might form their beliefs more carefully, by applying inductive reasoning to some or all of the available evidence. Less confident individuals might hedge their bet on Cournot dynamics with some probabilistic version of the rule. Strategically minded individuals might predict the effect of their current choice on future choices of the other agents involved, and factor this into their decision.” (pp. 278-79)
I trust that most of my readers can analogize from this paragraph to the formulation of various trading strategies and don’t need me to connect the dots. We don’t have to buy into the notion that game theory provides an accurate model of how trading decisions should be made. Just take the passage at face value and act as if you never heard of John Nash, game theory, or a Nash equilibrium. And then, as an exercise, bracket your knowledge of trading strategies as well. From this hypothetical tabula rasa develop reasonable guidelines for each learning/trading model (the simple, sophisticated, less confident, and strategic). In some cases you’ll undoubtedly be reinventing the wheel. But who knows? Perhaps you’ll come up with a new idea or two.