With the introduction of the moon/mars wonders, I note that some of them provide bonuses in the form of -% infra cost and/or upkeep. Since the effectiveness of -% bonuses supposedly gets diminishing returns the more of them you have, I am curious about how bonuses are aggregated together for stacking purposes.
For reference, sample infra reduction -% values:
Resources - Construction set (Lumber, Marble, Iron, Aluminum): -29% total
Improvements - 5 Factories: 5x-8% = -40% total
Wonders - Interstate System: -8%
Government - Some government types: -5%
The total isnt simply adding all of these up; if it were, the bonus here would be on the order of
-29% - 40% - 8% - 5% = 82% total reduction = 1 - 0.82 = 0.18 base infra cost
If something like this is how bonuses are aggregated, then the full result would be something like:
1 * (resources) * (improvements) * (wonders) * (governments), or in this case 1 * (1 - .29) * (1 - .4) * (1 - .08) * (1 - 0.05) = 0.372 base infra cost
In short, the more 'groups' there are for lumping bonuses together, the less effective any single -% bonus is; i.e. from the above example the -5% from government isnt actually -5% from the base, but rather -5% * -29% * -40% * -8% = -1.95% from the base.
A worse scenario would be each individual wonder/bonus resource/etc being counted separately; i.e. 1 * (interstate) * (moon/mars base) * (resources) * (construction BR) * (steel BR) * (government) * (events) ... etc
So my question if anyone can tell me how much incremental -% boosts (such as those provided by the expensive moon/mars wonders and some of their bonus resources) are reduced in effectiveness due to how many divisions there are in stacking these bonuses together. I understand wonders are not meant to be a huge boost, but am interested to hear details.
Thoughts?
-Preston
