It's part of a larger approach to making you make actual decisions about settling.
...
Personally it's boring if all tiles are the same.
While I agree on both counts I don't believe
"avoid water like the plague" makes settlement decisions particularly hard.
I hope we can all agree on the following points.
- Production is the strongest yield in the game, by far.
- Food is second, partially due to necessity and partially due to it being the mechanic to actually grow your population.
- Every other yield is far down the line in priority. A nice bonus, perhaps, but not a reason to actually work a tile on its own.
- The ability to place districts, both due to the bonuses they provide through adjacency alone as well as the buildings unlocked is important to the Civ6 experience.
- The ability to improve tiles is quite strong. Late-game farms provide massive amounts of food as well as increased housing capacity. Mines add 3 production on top of the base yields, as well providing adjacency bonuses to Industrial Zones. And so on.
- Desert tiles are undesirable due to their low base yields, though they can still be used for district placement and can offer significant bonuses with the right pantheon beliefs or wonders.
- Tundra tiles are undesirable for much the same reasons. Without any resources they can't be improved and even Russia can't sustain any city on tundra tiles alone.
- Rainforest tiles usually have decent base yields 2F/1P or 2F/2P and usually offer a good mix of bonus and luxury resources. Base rainforest still suffer from not being possible to improve but in the worst case it can still be chopped for bonuses while allowing the underlying tile to be improved, or buffed through wonders or civilization-specific features (Kongo's Mbanza district or Brazil's adjacency bonuses for example).
Now for water tiles.
- The base water yield is arguably worse than tundra but better than desert. 1F/1G is roughly equal to 1F but water can't have hills or forest, which means tundra gets the nod due to production bonuses.
- Base water tiles cannot hold any districts aside from the harbor, which is good but very low priority.
- Base water tiles cannot house any improvements.
- There are no adjacency bonuses for base water tiles, nor any pantheon beliefs or civilization-specific bonuses to water tiles. (God of the Sea applies on to sea resources.)
- The water-based wonders are all very lackluster.
- Since ocean access is actively detrimental to city development and there's no need for actual ships to explore across the ocean you have no secondary reasons for wanting ocean access in any city.
I find that the land-based analogy to base water tiles are flood plains, or arguably marsh. However, these tiles have base yields of 3F and can house improvements (once you clear the marsh) and districts or wonders (marsh only, unless you're playing as Egypt). You're generally happy to work an early marsh for the food but come mid-game it's frequently better to clear for the significant farm bonuses or to house a district or wonder.
You'd not be particularly happy to settle a city where a significant part of the immediate tiles were flat desert or tundra, mountains or snow and yet this is exactly the same as settling a coastal city. Except unlike these other tiles mentioned you're getting absolutely zero bonuses from water tiles, and there's no bonus, pantheon belief or improvement that will help you.
As such I find that at a bare minimum water tiles should be 2F/1G and be possible to buff to 3F/1G, my proposal would be with an economic policy but it's quite academic overall.
Since even self-sustainable food yields wouldn't be enough however, because a city without decent production is a very sad sight indeed, I'd also like to see some kind of early non-food sea resources added to the mix.
Even then I'd expect little reason to settle coastal, rather it being less painful if you have to. Buffing the water wonders, introducing more of them, and giving reasonable bonuses from water tiles in other ways (adjacency, pantheon, whatever) and providing map scripts that actually encourage water exploration and navy-building are still very much required.
I always played Small Continents in Civ5 despite finding water lackluster because it, at the very least, forced me to explore the oceans for trade and relations. And I frequently found Small Continents to produce too large landmasses.
In Civ6 I see no reason not to play pangea maps, which I usually hate, because it's just not possible to utilize water in any reasonable way and the default continents map seem to only generate two continents which both span the entire north-south divide at maximum (Huge) size anyway.