Yeah, the new build has better happiness distribution. I’ve been comparing happiness distributions between versions for the same save, and this is the best yet:
The compass is also looking more reflective (though that technically was never a problem, just a result of it only reflecting two out of many axes):
It’s definitely harder to get to a winning position, but not even close to difficult even at 150% difficulty. It just takes some extra turns. The thing I notice is that the game tends towards an average approval either fairly well above 50% or fairly well below- typically I am only around 50% voting intention early on, and then it’s smooth sailing the rest of the game.
This is a test game demonstrating what I’m referring to.
Maybe part of the problem is balance between membership influences changing the number of people within a group vs changing the strength of their affiliation with that group.
I also think it might be an idea to correlate more voter groups with each other, at least when those correlations are globally consistent. Maybe that could be conditional on other factors, to capture when correlations aren’t globally consistent. I’m not sure. But voting blocs do exist irl and the current distributions don’t appear to reflect that.
Part of the issue is also that happiness in general is too easy to come by. It’s kind of impossible to piss off the poor without being extremist, for example. Environmentalists are easy to please with very little spending in money or political capital.
I think another element is that the relationships between sim values and policies remains far too simple for balanced results- sometimes in terms of numbers of inputs, sometimes inputs chosen, sometimes the way the inputs behave. Work on the UI will make that progressively less problematic in terms of accessibility to resolve. But it’s mostly a lot of small things which I think probably warrant changing- the kinds of stuff that tends to fly under the radar even when mentioned on the forum because they are, taken individually, not significant.