Your call, I suppose...
This would also allow us to charge less for each expansion pack and/ or make them larger, without making the higher-level backers feel like they're not getting their money's worth (so if we only have 2 or 3 Kickstarter-related expansion packs, and we charge $4 or $5 each, it's not a case of "hey, I paid $20 extra and all I got was this lousy $15 worth of extra content").
For the "Basic Game"... Argh! The "trial version" will be free, but require a code to unlock additional content. It sounded right, right up until you made me reread it. Nor will the game ever run on an old flip phone - "smart Phone" and "Cell Phone" seemed reasonable interchangeable.
The smaller form-factor devices do need "fairly" simple graphics, but it's a relative term. It will be the same source images on every device (which means it won't be pixel-perfect), and the graphics will have to be simple enough that they are still recognizable. As far as UI, it will be different layouts, but - again - the same source images, just with different locations and scale. The framework I have uses layout definitions, and will load static images and play sprite sheets in designated locations, according to those layout definitions. The style will be decided by vote and argument (in the classical sense), but they're going to be at least reasonably hi-res images.
Consoles are actually fairly simple, since I'm using Xamarin/ MonoGame, and it allows you to write code against a "generic" controller (the deciding factor on whether I can port it to a console will actually be whether they allow an "indie" game to go out to an FTP server and download game files).
And the money... Everybody hates the way EA handled things, but only slightly less well hated is when you can't play a game anymore because the activation server isn't there. So, just like taking it on faith that there will be follow-on expansion packs, I'm taking it on faith that most people will be honest. And the terms are "be honest"... Plus, my current plan is that the key generation/ validation will be tied to your email address, to help keep the honest people honest.
MOO3? Really? Have you not read all the other comments? Or are you just trying to start a flame war?
Under the hood, however, it all comes down to fuzzy logic and decision trees really simple, old-school decision trees, similar to a choose-your-own-adventure book.
Each computer player will have a set of numbers that indicate, on a sliding scale, how they feel about every other player, and these numbers will change over the course of the game: how much they (and their subjects) like the other; how much they trust the other player; and even whether they respect the other player, or look down on them with contempt. There will also be relative statistical numbers, for stuff like tech level, civilization size, etc.
For every type of interaction, and for every race, there will be dozens of possible things that can be said at each step in the conversation – that list will be filtered down to a few items based on how the speaker “feels” about the other player at that point in time (and also based on previous selections during the conversation). And the choices that are made during the conversation will, in turn, affect how each side feels about the other (by modifying those underlying numbers).
For the AI players, they will be given the same choices as the live player, and each available choice will be assigned a percentage chance, based on the player’s feelings and the relative standing of each civilization, and a simple (virtual) dice roll will determine their choice.
The reason why worry kills more people than work is that more people worry than work.