B. A validation of the structure of the model from the best verified models from outside economics (e.g. cognitive science);
C. Verification of models against hard data, using not only on point-by-point prediction but also the qualitively correct behaviour of its output;
D. Explicit statements of the expected conditions of application of models;
E. The use of computational tools from AI to implement a simulation approach based on an explicit agent to agent correspondance, avoiding artificial 'stand-ins';
F. The use of declarative and logic-based formalisms and tools to allow the formal capturing of qualatitive and emergent processes;
G. The inclusion of the content of communication and agreements between interacting agents (other than purely numerical ones such as price);
H. The regular re-evaluation of models against totally new data;
I. The production of possible scenarios to enrich the discourse of executives and politicians rather than the pretence of forecasting;