Game Theory For Decision Support Part 2: Advanced Decision Games and Graphs
In the last post, I used a very simple game to introduce some of the concepts of advanced games. I talked about Game Theory (GT) for Decision Support Systems (DSS) and loosely defined a few terms while telling the story of the Prisoner's Dilemma:
Heuristics
Bias
Dominant Strategy
Rational and Irrational Actors
Strategies (GT Definition)
Nash Equilibrium
Decision Space
Competitive and Collaborative Strategies
Zero-Sum Games
Destabilizing Mutations
I concluded by introducing the disadvantages of incomplete information and the experimental process DSSs introduce into the game.
Main Points of This Post:
Real-world games involve multiple players and happen over time, increasing the complexity of the game.
Games can be represented using graphs.
Graphs allow us to present games to decision-makers so they can provide expert information.
Simple games and graphs are easy to simulate, and the DSS can run with low levels of human intervention.
Frequent complex games provide enough data to create quality DSSs and quickly improve based on expert feedback. Over time, the level of human intervention decreases.
Infrequent complex games have irreducible complexity. DSSs can improved decision quality but will never operate autonomously.
Using all three approaches, we can build high-quality DSSs.