摘要: The expected utility framework developed in War Trap is revised to correct several deficiencies. Risk-taking orientations are now treated as an integral part of the model by introducing concavity or convexity into functions. zero-sum properties theory largely eliminated, and tendency toward interpersonal comparisons removed. Several earlier results replicated with new model, annual capabilities data. New propositions deduced that identify important limitations on conflict initiation, relationships resulting from differences perceptions tested. Support found for contention version theory, which original a special case, powerful tool integrating many extant hypotheses about explaining substantial portion some threats escalate violence warfare others be resolved peacefully.