ABSTRACT
A novel agent competitive learning model-computing techniques for analytics economics is developed with description computing logic models. Recalling a basis for model discovery and prediction planning from the first authors preceding years, new multiplayer game models are developed. The development is a basis for planning for predictive analytic support systems design. The techniques are forwarding a descriptive game logic where model compatibility is characterized on von Neumann, Morgenstern, Kuhn (VMK) game descriptions model embedding and game goal satisfiability. The techniques apply to both zero-sum and arbitrary games. The new encoding, with a VMK game function situation, where agent sequence actions, are have embedded measures. The import is that game tree modeling on VMK encodings are based on computable partition functions on generic game model diagrams. Newer payoff criteria on game trees and game topologies are obtained. Epistemic accessibility is addressed on game model diagrams payoff computations. Furthermore, criteria are presented to reach a sound and complete game logic based on VMK and hints on applications to compute Nash equilibrium criteria and a precise mathematical basis is stated.
