I thought about modelling sensitivity, and how close to "reality" it could be. And came with up this.

There are various rooms spread around the continent. Let's call them modelling rooms. They all work independently. They share a large percentage of common root knowledge from published literature, maybe going back a generation or two. But each room also has their own original research and knowledge to add value. Each room is full of the top rated experts in their field.

They all have the same thing to model. A competition between countries. Each country models their country in competition with the others. They are allowed to model strategies and see how these flow through iterations. Eventually deciding on the optimal one. They also state an expected length of the game and the game resources used.

And the strategies and results all look reasonable. And the modeled winner from each room, the game length, expended resources, all look reasonable. Clear winner, within acceptable time frames and game resources.

So everyone is happy let's start the game for real.

Reality? Trench warfare starts.

The greatest modelling minds.