This is the story of my game of Civilization IV Adventure Forty-three from Realms Beyond Civilization.
This idea somehow came to me when playing the Final Frontier mod and thinking about the food mechanics. City growth comes so slowly in that game, in part because 4 food is the max planet yield, there are no 6-food resources (or any resource tiles at all) as in standard Civ. I found myself wishing that the 6-commerce yellow planets could be 6-food planets instead, and then started thinking about how a standard game of Civ would play out with such.
The tile yields actually had to rotate in the opposite direction, with commerce not actually becoming food. Commerce yields blow up drastically in the later game, 100 commerce for a city is not uncommon (40 from trade routes alone), and I think everybody will be happier not managing size-50 cities. Therefore it's hammers -> food, and the rest of the cycle follows from that, hammers -> food -> commerce -> hammers.
Also, I simply rotated commerce, and let raw direct cash and beaker production function as normal. Direct beakers will be something of a safety net since I suspect food alone isn't enough to run a research economy. So specialists and the religious economy (shrine, Sankore, Spiral Minaret) work normally, and I strongly anticipate that that will be the best way to handle this game.
That plays well into the Spiritual trait choice. OK, religion doesn't theme well with aliens, but I'm always about the mechanics. Spiritual is key to allow free experiments with civics (Caste System workshops seem great) but can also fit with a religious approach. The Apostolic Palace giving +2 food per religious building is also an interesting interaction. And Mysticism is a good starting tech, allowing a clinch to an early religion if the player desires, by working a 3 food (beaker) tile. It also allows the player to sink excess production into Stonehenge whenever other build options run dry.
And I decided we must start with Fishing. Need to have commerce tiles available at the start, or the first worker will take forever. And a coastal seafood start makes for interesting options. When is a work boat good to build? Fish is a 0-2-5 tile - bizarro tundra hill silver - is that good to work? I actually don't know, so let's throw it to the players.
Mysticism + Fishing leads to Isabella as our leader. Expansive fits well too, the worker boost will help make up for scarce commerce, granaries will be needed to bootstrap slow-growing cities, and health helps specialists and will be a major help after the industrial era boosts food (wow, what is the Ironworks going to do!)
The map is small to reduce the micromanagement workload. Micromanagement is certainly required, since the governor doesn't understand the tile yield switch of course, so it will auto-assign cities into hammer starvation. And I suspect that the best economy will be specialist oriented, mostly ignoring food tiles (farming a grassland makes for a terrible 0-0-3 tile, a specialist is strictly better) which also means more MM.
Balance is not a primary concern here. I'm aware that stuff will be quite broken. Bureaucracy provides +50% food! The Great Lighthouse provides +2 or more production per city by trade routes! Golden Ages cause explosive growth! Plain old citizen and engineer specialists produce food! Wait, that's not broken at all, in fact it's a complete zero, the engineer produces exactly what he consumes and the citizen is a net loss.
Yes, there's the flip side. Food does not come in such great quantities (though I seeded the map with food resources) and research will be slow. Yes, big production and slow research add up to a medieval/renaissance military victory. (If you can pay for the army!) Those who want to experience the mod and then exit can do so, for which Spain fits well with an appropriate UU and UB. And those who want a deeper experience can choose a later win condition.
The point is not so much balance. The point is to experience Civ as a completely new game with just a small tweak. And I want to invoke mind-bending plays like workshopping flood plains (sacrilege!), grassland is junk and plains are better (ow my brain!), converting hammers to hammers with the whip (wait, what?), settling in hills for growth (*bend*), and settling near water for production (*snap*).
Forest chopping I changed to add to the food box instead. Other methods of hurrying production I really had to leave as-is. Accurately translating the whip's conversion of food-to-hammers would mean somehow creating a commerce-to-food conversion, but that really isn't possible, cities don't have a "commerce box" to whip out of. Same for cash rushing. So I decided to treat whipping as "converting population to buildings" rather than "food to hammers", while forest chopping literally creates hammers. That's the difference and I hope it was intuitive.
There's not a whole lot to say about the map generation. I rolled Fractals until I found this one, with some neighbors nearby and some good variety of city sites. I sprinkled in a bunch of hills and a few resources, including the ivory which I think makes for an interesting tile. I tweaked the capital site for more hills and the flood plains (just to demonstrate that it's not a great tile here), although the furs resource spawned there naturally (a very good tile, almost like normal gems.) I removed stone; if you want the obvious Pyramids for specialist economy, build them yourself. Finally, I've dinged Sullla for being predictable on this, but I kept copper away so as not to make for a turn 50 axe rush, and put both iron and horses in several places nearby so as not to cripple the player (they're also food growth resources here.)
I let the game roll random AIs, which happened to come out with a nearly perfect mix. Elizabeth as a soft target for a player that wants to try out limited war. Montezuma as a stiffer challenge for bigger conquest. Hannibal as a pretty good research rival, and Bismarck as a rival for wonders. The AI starts I left alone, with the only edit being to swap the original spots of Elizabeth and Montezuma. I don't think that needs elaboration. :)