Again, it depends. While decreasing uncertainty can increase the sense that one's action are rational it may also provoke rashness. Alternatively, while the existence of uncertainty may provoke action against an adversary to defeat the possible threat, it is equally likely to produce policy paralysis. The role of the individual actor again. That said, you will probably get better policy the more you can reduce uncertainty with solid data and analysis.

Don't be particularly concerned about being forced into a realist mode. Allison's work (mentioned previously) demonstrates that the best predicting of his 3 models is the classical rational actor. The other 2 (org theory and the 'political' [stressing individual background, biases, and preference] only fill holes in the rational actor explanation/prediction. The key to its use is the rigor with which the model is applied - and I don't mean quantification necessarily.

Cheers

JohnT