[NFP] [Discussion] Are rainforests too prevalent and/or scattered on Civ 6 maps?

[kaspergm]

I'm just curious as to what people's opinions are about how Civ 6 map maker generates rainforests (jungles). Personally I dislike the way rainforests seems to be scattered randomly and almost evenly over the entire half of the map or more that's closest to equator.

I know hexes on the map are supposed to represent large patches of land, and hence one could argue that one or two patches of rainforest represents a coherent stretch of forest, but I feel the map would be more satisfactory and more realistic if we had larger and more coherent clusters of rainforests isolated closer to equator, and then less rainforests scattered throughout the temperate region.

I know one can also argue that the current distribution of rainforests is at least to some extent due to deforestation over historic time, so I'm curious to know, what other members here feel about the way Civ 6 handles this.

 

[Myomoto]

I hate the prevalence of rainforests. It seems 50/50 if my start is in rainforest, and I think they should really be similar probability to a desert start.

 

[agonistes]

I’m fine with jungle, its the deserts that annoy me. Every freakin start I seem to be within 6 hexes of a desert.

 

[FinalDoomsday]

Yea, its pretty cool to find a huge jungle, desert, mountain range or vast steppe I'd love a map script that made that happen more.

 

[Boris Gudenuf]

Two things:

1. The current distribution of forest and rainforest makes no sense compared to actual biomes: without outside interference (read: Humans and some animals) there is a Forest Belt from the North Sea to the Ural Mountains right across northern Europe, and a Rain Forest belt across the tropics anywhere there was enough water.

2. The exceptions are where large animals like Elephants and their earlier cousin Mammoths eat trees to death, then move on, creating a sort of 'rotating meadow' effect in which trees die out and open up the landscape, then regrow after the pachydermic appetites move on. Other animals like large herds of smaller herbivores can have the same effect - it has been debated whether the North American Bison contributed to the extent of the Great Plains that they grazed on by literally pushing over small trees to keep new forest from encroaching on the prairie. Also, Humans remake the landscape with fire, which they've been doing since at least the Mesolithic - burning out forests to stimulate new growth that both produces more food (berry bushes, etc) and produces more browse to attract prey animals.

In other words, we need both more dynamic terrain and more consistent placement of forest and rainforest.

The reason for the current Random Acts of Treeness across the Civ VI map is that the game is locked into a system in which to make consistent progress in several areas (Gold, Food, Production, Science, etc) you need a variety of terrain and botanical forms. Consequently, being stuck in a flat plain entirely full of trees for a Start position gives you little or no Food and few or no adjacencies useful for your first Districts - you are, essentially, crippled without an extreme distribution of 'bonus' resources or Civ UAs to make up for the terrain.

Which means, to get a Playable realistic set of biomes for the game, the ability to extract Food, Production, Gold, Science, etc from the terrain has to be more diverse, perhaps tied to specific adoptable Civics or Technologies early in the game that a Civ can adopt to be competitive in a less diverse environment. Just as an example, the Natives of the Amazon manufactured garden plots of good soil (using natural fertilizers and charcoal fixative, among other things) to practice intensive agriculture in a rain forest which normally has notoriously poor soils for agriculture - a unique "Technology" that was not needed or developed by people living outside of the rainforest.

 

[iammaxhailme]

I think rainforests should be smaller, but more dense and have near-complete coverage over certain spots. Not in random small patches all over the equator.

 

[Datian]

I like that they are a bit scattered. Although it seems less realistic, it suits the game purpose - as I often play Brazil, I will always find a nice place for my holy sites

 

[kaspergm]

The reason for the current Random Acts of Treeness across the Civ VI map is that the game is locked into a system in which to make consistent progress in several areas (Gold, Food, Production, Science, etc) you need a variety of terrain and botanical forms. Consequently, being stuck in a flat plain entirely full of trees for a Start position gives you little or no Food and few or no adjacencies useful for your first Districts - you are, essentially, crippled without an extreme distribution of 'bonus' resources or Civ UAs to make up for the terrain.

Which means, to get a Playable realistic set of biomes for the game, the ability to extract Food, Production, Gold, Science, etc from the terrain has to be more diverse, perhaps tied to specific adoptable Civics or Technologies early in the game that a Civ can adopt to be competitive in a less diverse environment. Just as an example, the Natives of the Amazon manufactured garden plots of good soil (using natural fertilizers and charcoal fixative, among other things) to practice intensive agriculture in a rain forest which normally has notoriously poor soils for agriculture - a unique "Technology" that was not needed or developed by people living outside of the rainforest.

Yeah, I agree wholeheartedly with your analysis here. I think the "solution" to this is moving away from the idea of flat terrain yields and instead towards citizen yields based on specialization. You mention the native amazon people as an example, and indeed there are examples from almost any biome that humans have specialized in finding food as the first step of building a civilization - that goes for pretty much any terrain except desert (perhaps?) and ice.

There was a mod in Civ5 that moved a bit along the same line, although in a slightly different context. I don't know the name, but it gave an attrition penalty to units moving in difficult terrain (desert, tundra, jungle) except if the unit was trained in a city build on this terrain. I think this thought can be merged with what you suggest also: Instead of a flat terrain yield, citizens should be able to work a terrain for (the analogy of) 2 food if they have the knowledge of this terrain. You automatically gain the knowledge of a terrain if you settle your capital in it, additional terrains need to be "learned" in a city that inhabits the new terrain form - i.e. if you settle your second city in a new terrain, you'll need some time to learn the ways of surviving in this terrain.

Obviously, there should still be differences in the ways different terrains cater to improvements - you have to remove forests and jungles to build farms (which will then increase per-pop food yields, similar to Civ6). On the other hand, Lumbermills should be placed on forests and mines placed on hills or resources, which will allow you to work this terrains for production (but then yield no food).

This would also tie into another suggestion that was discussed recently, namely that instead of flat yields from buildings, instead they should only give yields pr. citizen working in the building. So instead of the flat 2 science yield from the library, the library should allow one citizen to work there for a yield of X science (which could be lower than the yield of each specialist working in the university and research lab). So basically everything moves away from "free yields" towards a strict "per worker" yield.

PS: The "region" mechanism that for instance Humankind will apply will work nicely in this context. Then each region can have its own biome - i.e. forest, rainforest, desert, etc.

 

[Boris Gudenuf]

Yeah, I agree wholeheartedly with your analysis here. I think the "solution" to this is moving away from the idea of flat terrain yields and instead towards citizen yields based on specialization. You mention the native amazon people as an example, and indeed there are examples from almost any biome that humans have specialized in finding food as the first step of building a civilization - that goes for pretty much any terrain except desert (perhaps?) and ice.

The earliest cities in the deserts of southern Chile did not have agriculture for food, but did grow a form of cotton, wove the fiber into nets and ropes, and traded them to cities on the coast that netted great quantities of fish from the Humboldt Current and traded the food back - the most convoluted 'solution' to food in the desert that I know of, but an indicator of the extremes to which a culture/Civ can go to maintain itself.

There was a mod in Civ5 that moved a bit along the same line, although in a slightly different context. I don't know the name, but it gave an attrition penalty to units moving in difficult terrain (desert, tundra, jungle) except if the unit was trained in a city build on this terrain. I think this thought can be merged with what you suggest also: Instead of a flat terrain yield, citizens should be able to work a terrain for (the analogy of) 2 food if they have the knowledge of this terrain. You automatically gain the knowledge of a terrain if you settle your capital in it, additional terrains need to be "learned" in a city that inhabits the new terrain form - i.e. if you settle your second city in a new terrain, you'll need some time to learn the ways of surviving in this terrain.

Obviously, there should still be differences in the ways different terrains cater to improvements - you have to remove forests and jungles to build farms (which will then increase per-pop food yields, similar to Civ6). On the other hand, Lumbermills should be placed on forests and mines placed on hills or resources, which will allow you to work this terrains for production (but then yield no food).

This would also tie into another suggestion that was discussed recently, namely that instead of flat yields from buildings, instead they should only give yields pr. citizen working in the building. So instead of the flat 2 science yield from the library, the library should allow one citizen to work there for a yield of X science (which could be lower than the yield of each specialist working in the university and research lab). So basically everything moves away from "free yields" towards a strict "per worker" yield.

PS: The "region" mechanism that for instance Humankind will apply will work nicely in this context. Then each region can have its own biome - i.e. forest, rainforest, desert, etc.

IMHO, the Specialists need to return to the game in force, as a mechanism for 'working' Buildings and Improvements.
In fact, I've come around to the Endless Legend/Humankind system in which every tile adjacent to a city tile is automatically 'worked' regardless of the city population (which makes building up an early city much faster), but I'd add that every point of Population also generates a Specialist (and maybe more - training facilities, Guilds, schools, etc. would produce more Specialists per Population) which can be applied to Buildings. Whether a building is next to a mountain. marsh. coast, hill or hollow doesn't matter: how many skilled people are using that building to produce Food/Gold/Production etc is the critical point. Couple this with Units costing Specialists, and the real cost of a large army - or of mobilizing a large army for a war - would come home to roost =: every Specialist in the army is one less working the lathe, potter's wheel or millstone back home to produce the weapons, goods and food required to keep the whole Civ going.

And another thing to make the 'realistic' forest/rainforest extent work is that it must be possible to remove the trees or otherwise modify those tiles much earlier than now in the game. In fact, as mentioned, people were burning out forests long before Start of Game, and rainforest/jungle areas were being extensively modified with Ancient technologies (another Fun Fact: a large percentage of the trees and plants in the Amazon Basin were planted: the distribution is not natural, and it is now recognized that many of the concentrations of some trees and other plants are there because people planted them there because they were useful, and did it without metal tools).

 

[Naokaukodem]

In gameplay term I know that when I take Brazil, there's never enough rainforest. Except in one recent game : I could manage to create 3 cities (capital included) that could have a 6 science adjacency bonus from start. (rainfall : wet) With natural philosophy civic and scientific city-states (I was at 6 envoys with two of them, last were conquered), the adjacency bonuses of 3 of my science districts were the number of 22 !

The sad part is that Japan was ahead of me quite a long time, and it was on Settler difficulty... I can't imagine if it was on Deity. (my two favorites difficulty settings )

Overall I dislike the map generator, you often start near the poles or other crappy locations. Not talking about sometimes awful mountains ranges... and it doesn't make you feel you start in an isolated biome like Aztecs or Pygmées... the map must be too small I guess.

I should try huge map, rainfall : wet, number of civs : 6/8 with Brazil.

 

[8housesofelixir]

The earliest cities in the deserts of southern Chile did not have agriculture for food, but did grow a form of cotton, wove the fiber into nets and ropes, and traded them to cities on the coast that netted great quantities of fish from the Humboldt Current and traded the food back - the most convoluted 'solution' to food in the desert that I know of, but an indicator of the extremes to which a culture/Civ can go to maintain itself.

Many earliest cities of human history - the 10000-year-old cities in the Levant and Fertile Crescent - followed the same route: These cities' emergence and development was not a result of local food surplus, but a result of acting as a trading/industrial center, and they would trade their products for food. So basically, a highly specialized city can still survive via being in the trading network IRL.

However, the current Civ 6 game mechanics don't really favor city specialization - the most "specialized" situations we have are simply coastal city=gold, mountainside city=science/faith, hill city=production, etc. And a city specialized in only food/production/science or culture yield generation will develop very, very slowly unless you invested in Magus (for chopping districts, something I personally consider as unrealistic) and/or internal trades.

As a result, the map generation need to provide a variety of terrain features - I fully agree with @Boris Gudenuf's point here - to help the city to grow. Only have one type of terrain/feature near your city, esp. the starting city, will hurt you a lot.

Unless the civs can have more diverse abilities to survive in extreme terrains, or can have more specialized cities/more developed trade routes to help the cities to support each other*, we would still encounter these two-tile rainforests on the map for balance reasons.


*I know that both internal and external trade routes can be very helpful for city growth, but it usually depends on the player to send out trade routes, because the AI is really bad at sending trade routes to players. It seems to me that, although AI love trading with you (see the World Congress trading resolution, if you are playing peacefully and focus on developing your cities' districts, AI will always vote you as their best trading partner for the +4 gold benefit), their trade route destination is always your border city (nearest to them) or your capital, and they tend to build only a few traders. IRL cities can count on their neighboring cities' traders to help them, but in the game the player cannot really benefit from AI traders, making the "global trading network" very unlikely.

 

[Naokaukodem]

I did a couple tries in Huge... that's exactly how you describe it ! So it's not a question of map size...

 

[hhhhhh]

I think one likely reason is that dev set a percentage for rainforests, say 15% but then they can only appear near the equator so it feels really dense, like 50% of the tiles around the equator are rainforests. Meanwhile historically it's kinda accurate for forest (not necessarily rainforest) to be very dense, see for example here. That woods and rainforests being scattered on map is only true post the industrial revolution.

 

[Boris Gudenuf]

I think one likely reason is that dev set a percentage for rainforests, say 15% but then they can only appear near the equator so it feels really dense, like 50% of the tiles around the equator are rainforests. Meanwhile historically it's kinda accurate for forest (not necessarily rainforest) to be very dense, see for example here. That woods and rainforests being scattered on map is only true post the industrial revolution.

The majority of us live in, near, or surrounded by urban or semi-urban areas, and for most of us, our ancestors going back at least one or more generations did the same. Consequently, we have forgotten what the land looked like before humans extensively modified it.
I recently reread an account of the "Corps of Discovery" - the Lewis & Clark Expedition, in which every member kept a diary or journal - probably the most eye-witnessed journey in world history. These men, from a society that was less than 10% 'urban' in any form itself (Post-Colonial North America) were absolutely stunned at the sheer mass and profusion of wildlife in the country they traveled through: herds of herbivores from antelope to deer to bison, flocks of birds that covered the skies, so many Grizzly Bears that they got tired of having to shoot them away from their camp every day - and the bear was the Apex Predator, and so should have been the rarest of all animals. It was still in such profusion as to be a constant companion!
Add to that the Roman accounts of traveling north of the Alps in Classical times where the forests were so huge they actually frightened travelers from the Mediterranean (where most of the forests had been thinned out or wiped out over the centuries of building and burning) by their sheer bulk, and there is a picture of wide expanses of plains or forests that is simply missing entirely from the game.

But, as stated earlier, to make a 'natural' extent of the forest/rainforest/jungle biome work in the game, it has to be possible to develop a viable city in which every tile within the city radius starts out as forest/rainforest/jungle - or plains/grasslands, for that matter. Right now, that cannot be done: the cities would be missing critical adjacency bonuses, and could only develop unevenly: lacking either Food, Production, or Gold to develop while fighting off the incessant barbarian incursions . . .

 

[kaspergm]

As a result, the map generation need to provide a variety of terrain features - I fully agree with @Boris Gudenuf's point here - to help the city to grow. Only have one type of terrain/feature near your city, esp. the starting city, will hurt you a lot.

Unless the civs can have more diverse abilities to survive in extreme terrains, or can have more specialized cities/more developed trade routes to help the cities to support each other*, we would still encounter these two-tile rainforests on the map for balance reasons.

Yeah, I think that is the dogma we need to move away from, as also argued by Boris Gudenuf above: That the terrain type is the be-all and end-all of what resources your cities have access to. We need something more dynamic in the way that citizens interact with the terrain based on city and civilization specialization - and not in the "Brazil always loves rainforest and everybody else don't" way of Civ6, but rather in terms of adaption to your initial conditions.

 

[8housesofelixir]

Yeah, I think that is the dogma we need to move away from, as also argued by Boris Gudenuf above: That the terrain type is the be-all and end-all of what resources your cities have access to. We need something more dynamic in the way that citizens interact with the terrain based on city and civilization specialization - and not in the "Brazil always loves rainforest and everybody else don't" way of Civ6, but rather in terms of adaption to your initial conditions.

In terms of lacking of resources in certain terrains, currently, the generation rule of bonus resources is designed for compensating the lacking of certain yields in the particular terrains. For instance, stones usually generate on grassland tiles, because grassland is lacking production; fishs and crabs for making sea tiles worth of working; and deers often clustered near tundra. As for the topic of thread, rainforest, we have bananas for compensating the lack of food/gold.

On the other hand, bonus resources still suffer from RNG and weird distribution, and the idea behind them still falls into the trope of "having a variety of features everywhere". Ultimately, we may need a more dynamic/specialized way of generating tile yields for cities in different terrains.

Some types of terrains are surely not ideal for developing a city - for example, deep in the desert or rainforest - but with adequate infrastructure one should be able to support a city in extreme terrains (with a high investment cost, as they are really not ideal).

 

[OzzyKP]

I just started my first ever Brazil game and I can’t find rainforest anywhere.

 

[Shadowstrike]

Rainforests are just too high-yield IMO. With two food, it's just as good as an undeveloped grassland tile, and when you add a lumber mill to it, you get super powerful rainforest cities. That's kind of contrary to real life, where rainforests are some of the least populated areas in the world, and are regularly cut down for cultivation.

 

[Boris Gudenuf]

Rainforests are just too high-yield IMO. With two food, it's just as good as an undeveloped grassland tile, and when you add a lumber mill to it, you get super powerful rainforest cities. That's kind of contrary to real life, where rainforests are some of the least populated areas in the world, and are regularly cut down for cultivation.

Historically (and archeologically) the Rainforest could be very productive, but it took a lot of Hard Work - an 'unimproved' Rainforest was, as you note, practically worthless except to hide in.
The Indonesian Rainforest, though, provided relatively high-value hardwood lumber for ship building and aromatic woods (sandalwood, as an example) which were a lucrative trade item with Song China.
The Amazon rainforest was intensely cultivated, possibly including the planting of useful types of plants and trees, but the amount of labor involved beggars the imagination - they literally manufactured soil in which to plant their garden/farms.

So, there's nothing wrong with having Rainforest be very productive tiles - once Improved in various ways. Raw, they provide neither a good concentrated food source nor a Gold/Production benefit compared to more conventional terrain. The problem, as I've stated before, in game terms is that the techniques, technologies, Civics, etc required to Improve the rainforest should be available very early for those cities and Civs living in the rainforest: The Indonesian/Southeast Asian teak-built 'junque' ships were plying back and forth between the islands and the Asian mainland by the Classical Era, and the Amazon rainforest was being heavily exploited without metal tools, so effectively in Ancient Era Technology level.

 

[Shadowstrike]

Part of what makes rainforests so powerful in Civ though is that they produce 2 food. That makes it sustainable to grow cities on rainforests, without supplementary tiles to balance out the food production. We don't historically see giant cities in rainforests - Indonesia's population, for instance, is concentrated on Java, which has great volcanic soil for cultivation, and not in the jungle-heavy parts of Sumatra or Borneo. What might help with this would be to give certain civs with rainforest preferences (i.e. Kongo, Maya, Brazil, Indonesia) an ability to produce 2 food on rainforests, but not for other civs. That kind of forces you to choose between cutting down the rainforest to grow food, or keeping them for their production.