Thats my plan for democracy 2, and it's going well.
The first big change was fractional; group memberships (I'm not just a motorist, I'm 63.445% a motorist). That's in and working.
The second change was allowing recursive linkage between effects, so you can get positive and negative feedback loops. That was hell, but it's in and working.
Today I added invisible nodes, like "Terrorism" that you can see being affected, and affecting things, but you can;'t actually query their current value.
So you know the new spy satellite is preventing terrorism, you just don't know if it's enough...
Also today, right now, I'm adding targetable cost-multiplier nodes.
In the old game, a policy like 'National Health service' has a minimum and a max cost, based upon implementation, and also optionally scaled by a single other object.
So you could have car tax income being scaled by the number of motorists (makes sense).
But now rather than a single object acting as a multiplier, the multiplier itself will be a node, capable of multiple inputs all with different equations.
So you could say that the relationship between national health costs and alcohol consumption (number of drinkers) would be 0+(0.12^x), and this could be just one of many inputs into determining how much the health service will cost.
This game will be absurdly complex. will it be balanced? hopefully? will it be fun? hopefully, but it sure as hell will be interesting to play. Hey, someone has to make an antidote to all these dumbed down games right?