Discussion about Tile Yields

[Deussu]

I've heard a lot of fuss about how tile yields in Civ5 are notoriously bland and differ very little from each other. Since this is the creation & customization board, I figured this could be discussed at length. What would be a better alternative? What downsides would the modifications have?

At present the distinction is simple and easy to understand. Grassland & Flood Plains 2, Plains and Forest 1 & 1, Hills 2, Marsh & Tundra 1, snow/mountain/ice/desert for none.

Initially I thought "how about doubling all yields, production costs, and food consumption?" I'm unsure how that would work out, but I find it to be plausible. In addition forests would not make every terrain exactly the same, but just give -1 & +1. Hills would have to be exeption though, likely to give +1 & -1.

Moving on. I've read things about production being neglectful. Mines being the only worthwile improvements. Some mods increase the production output of mines by 1. What else is wrong? Do food outputs need tampering? Bonus resources, like deer or cow or sheep, do they need that extra something? Should these additional resources be a sort of 'mini-luxuries', producing maybe 1 when worked?

 

[Shiggs713]

I'd like to see coastal tiles improved. 1 1 is pretty bad, especially considering most people have lived near the coast for the entirety of human history.

I'd like to see it at least 2 1 , and the lighthouse still adds +1

 

[Deussu]

I'd like to see coastal tiles improved. 1 1 is pretty bad, especially considering most people have lived near the coast for the entirety of human history.

I'd like to see it at least 2 1 , and the lighthouse still adds +1

That would indeed be fair. They still wouldn't match riverside farms, but would definitely be more worhtwhile than they are now.

That's easy to edit in the XML files too. Must consider that in multiplayer games. Of course this would mean fish tiles would have to give a bonus of only 1 to stay in moderate balance.

 

[Pazyryk]

You could just tweak around the edges with +1 food for coast, +1 production mines, etc. That might be some improvement but I don't think it would fix the "blandness" problem that you note.

What I would do is boost resources and improvements across the board, and by a large amount (I mean +3's or +5's, not +1's). Leave unimproved/no-resource tiles low-yield and bland as they are. Then, raise food requirement to 3 or 4. If you did this, geography and city placement would be much more interesting.

 

[Spatzimaus]

There are a few fundamental problems with increasing yields across the board.

1> Citizens consume 2 food per tile, with specialists only using 1 (half) at a certain policy. If you double the food outputs, or just add an automatic +1 to certain terrain types, you drastically change the balance; currently, the only tiles that at least break even on food (allowing you to continue growing at the same rate) are grasslands or tiles with farms. Everything else basically slows down your cities.
You can try doubling those food consumptions as well, but that leads to other problems. You'd have to double the effects of Maritimes, of Granaries, and so on. And that's just for food; you'd have to double the effect of every yield for every resource as well, and the improvements that harvest them. (As it is those are too insignificant, as evidenced by the "build trading posts on the iron" strategies.) And don't forget yields for specialists.
But it'd still affect city growth rate, because the equation for growth is 15+8x+x^1.5, where x is (size-1), so while you can double the 15 term and the 8x term easily, you can't do the same for the exponential part.
And of course, you'd then have to go back and change the balance for research rates, production times, maintenance costs, culture thresholds, and so on.

2> Golden ages add +1 production to tiles that add production, +1 gold to tiles that add gold. Increasing tile yields would have a substantial impact on the balance of this, and it's not something you can edit easily.

3> It'd look ugly in the UI; once a tile generates 5 or more of something it just uses the large icon and puts a number on it. You'd be seeing that on nearly every improved tile if you doubled the yields.

 

[Pazyryk]

There are a few fundamental problems with increasing yields across the board.

1> Citizens consume 2 food per tile, with specialists only using 1 (half) at a certain policy.

OK, bad choice of words on my part. My "across the board" statement was for improved or resource tiles only (assuming you are referring to my post). I also indicated that food per citizen be increased. It's been done in mods in civ4 very successfully (Orbis for example).

The idea is to have most tiles NOT self supporting. Then you have resource tiles that are a real big deal. The benefits are: 1) More tile variety -> less tile blandness. 2) City placement and development that is more "tied to the land" (as a minor point, it even works as an anti-ICS mechanism, since cities without resources or fresh water would suffer from it). Do you have to re-balance everything? Well, probably a lot, yes. You could balance food by having the boost in resources/improvements offset the gimping of non-resource/improvement tiles (raising food/pop is effectively the same as gimping all tiles for food). Costs would have to be adjusted to offset having some tiles that are substantially better in s than others.

 

[alpaca]

My opinion can be found in more detail in my mod thread. I think that resources need to be more diverse and more powerful, this makes city placement a lot more important. Coast tiles could do with a minor buff, I set them to 1 2. I also added a tech yield improvement to both mines and trading posts.

Boosting resources does upset balance but in my opinion in a good way. Production was just too slow before and with higher yields it becomes better (directly via improved strategic resources or indirectly via higher food surplus). Gold amount is a problem, and I haven't fully addressed that yet but I did add a city maintenance cost and will increase unit upkeep. In addition to that, I think both CS bribes and RA can do with a bit higher costs anyways, so that will be taken care of.

Another thing that I think is a problem is the river power, especially during golden ages. I added +1 gold to roads, which makes the river boost during GAs less important but, of course, also inflates the total gold supply even more. There's always a use for gold between CS bribes, RAs and purchasing buildings or units, though.

 

[Deussu]

A good point about all those consequences, Spatzimaus. Guess increasing consumptions and production costs wouldn't be an answer to any problem. Increasing coast tile yields by 1, however, sounds acceptable.

I've been pondering about these bonus resources, and how useless they appear to be. They aren't special in any way. I'm sure people have been doing a lot mods making them better they feel is right. Just like I would like to make them (cows, sheep, deer etc.) mini-luxuries, which would grant +1. The effect wouldn't be drastic, but it'd be nice nevertheless.

 

[Spatzimaus]

OK, bad choice of words on my part. My "across the board" statement was for improved or resource tiles only (assuming you are referring to my post).

I was actually referring to Deussu's original post, where he talked about his initial idea of just doubling yields and tweaking from there. The full statement I made obviously applies to that, while components of it still apply to some of the less extreme ideas people gave (like adding +1 food to all Coast tiles).

As to your idea, I'd point to my Alpha Centauri mod; I took heavy advantage of the Improvement_TechYield table to have each improvement's bonus increase at various near-future techs. And yes, it had a substantial impact in the ways you've described: resource tiles become more important (to the point where in the end, they were the only land tiles actually worked, while everyone else went Specialist because of a variety of Statue of Liberty-style wonders), and the second-order dependence actually mattered (i.e., Grasslands became substantially different from Plains, not in the base yield, but because each had a different distribution of resources).

The idea is to have most tiles NOT self supporting.

I'd agree with this, but there's one fundamental problem. If you make it to where only one terrain type (Grassland) or one improvement type (Farm) can ever make a tile self-sustaining, then the "flexibility" becomes a myth, because you'd need a minimum fraction of the tiles to be in this category but there'd be a practical maximum as well. It's like the sliders in Civ4; in theory they'd allow for a wide range of combinations, but in practice, they didn't mean anything; you'd set the gold slider to the absolute minimum needed to break even financially, you'd set the culture to 0 unless you absolutely needed it (which you almost never did), and that was the end of it. It was more a feedback mechanism than anything else, telling you when you'd expanded too far.

If you really wanted to rework tile yields and improvement yields, then, I'd suggest having no "pure" yields; have Farms give +2 food and +1 gold, Trading Posts give +2 gold and +1 production, Mines giving +2 production and +1 research, and maybe a new "Research Lab" improvement that gives +2 research and +1 food. That sort of thing. Or apply it to the terrain types; instead of grassland being +2F and plains being +1F+1P, have grassland be +1F+1G and just boost farms a bit more.
(Again, I'd point to my mod: the improvement yields often were boosted this way, so that Farms gained +1 research instead of even more food, while the Pasture had its food increased and so on. It really improved the experience, in my not-so-humble opinion.)

 

[Zyxpsilon]

Interesting points made by everyone...
Lemme just add this tiny spreadsheet for a quick overview of what i believe could work given an extensive rebalancing system where Units/Buildings/Wonders are totally dependent on steady (and highly variable) resources supply & a strict concept where relative depletion of many & empire-wide stockpiling of the essential are factored in.

 

[Pazyryk]

I'd agree with this, but there's one fundamental problem. If you make it to where only one terrain type (Grassland) or one improvement type (Farm) can ever make a tile self-sustaining, then the "flexibility" becomes a myth, because you'd need a minimum fraction of the tiles to be in this category but there'd be a practical maximum as well. It's like the sliders in Civ4; in theory they'd allow for a wide range of combinations, but in practice, they didn't mean anything; you'd set the gold slider to the absolute minimum needed to break even financially, you'd set the culture to 0 unless you absolutely needed it (which you almost never did), and that was the end of it. It was more a feedback mechanism than anything else, telling you when you'd expanded too far.

I'm not sure I understand what you're getting at here. I personally hate the "flexibility" given by the slider in civ≤4. Partly for the reason you state: that it is false flexibility because you tend to max it out. However, I really hate the slider for a different reason: I don't want that flexibility. I don't want gold and research to be completely interchangeable (and thus equivalent). Killing the slider gives the potential for differentiation between high gold and high science producing cities (although Civ5 does not realize that potential for other reasons).

With reference to the topic at hand (tile yields): Do you mean by "flexibility" that all city sites should be viable? This is pretty much the case in Civ5. This is a product of two things: 1) yield differences between tiles are small and 2) cities have a wide 3-radius reach to get the tiles they need (especially high-food tiles that are more generally prioritized by the border expansion). I happen to like the flexibility given by the wide reach of cities, so I wouldn't want to mess with #2. If tiles differences were more accentuated (especially resource vs. non-resource), then you would indeed have some "bad" city locations and this would be a loss of flexibility. I would be happy with that. (Yes, I know it would lead to less balanced starting locations. But that is always a danger if location is made to matter more. That's something I'd be happy to live with.)

 

[Spatzimaus]

With reference to the topic at hand (tile yields): Do you mean by "flexibility" that all city sites should be viable?

No. I mean that the logic for making a city site viable follows a very simple logic:
1> Make some tiles that produce more food (3+) than the worker consumes (2).
2> Then, use the surplus to support a number of tiles that don't produce enough food and/or specialists.

The problem is that, currently, the Farm and Lighthouse-boosted Lake tiles are the only way to do #1 unless the tile has a resource that produces extra food (some of which are harvested with the Farm), so the ratio of food producing to non-food producing, once set, requires X tiles to be farmed. Then we add a third component:
3> There's little benefit to having large cities. Once you've worked all of the resource tiles, you're usually okay if you just stopped there, because the increased output of the next tile just won't be worth the effort.
(Population increases trade route income and research, obviously. But unless you're using a mod that prevents ICS in other ways, we all know how well you can get by without large-population cities.)

So what happens is that there becomes an optimal number of farms, depending on the local terrain types, and that number tends to be fairly low. It's like the Civ4 sliders, in that the system isn't actually as flexible as it at first appears.

Now, what I was trying to say is, forget about modifying terrain yields for the moment. Instead, tweak the IMPROVEMENT yields. For instance, imagine the following array of improvements that can now be built on any tile (even without a resource)
Farm: +2 food
Pasture: +1 food, +1 gold
Plantation: +1 food, +1 research
Watermill: +1 food, +1 production
Mine: +2 production
Quarry: +1 production, +1 gold
Windmill: +1 production, +1 research
Trading Post: +2 gold
Workshop: +1 gold, +1 research
Laboratory: +2 research
(This doesn't mean that each needs to start at +2 or +1/1. You can do it like the Farm, where it starts at +1 and gets boosted at certain techs.)

So that's the full array of improvements for four yields. Adding culture could add even more, but let's not go there yet. For this discussion, only the first four matter.

The point was this: you could still use Farms to make tiles be self-supporting, but there'd now be other ways to do this as well. So there'd now be a few possible strategies; maybe you farm the grassland tiles to make a few high-food tiles and then go for mines and trading posts on other tiles to balance out, or maybe you use the Pasture/Plantation/Watermill to make every tile capable of breaking even and just work whatever you need to on any given turn.
THAT is flexibility. And more importantly, it's flexibility the AI can handle better than the current setup. Right now, if you tell a city to emphasize production, it'll do so at all costs, even if that means going negative on food. The AI just doesn't understand the idea of adding "...but make sure I'm neutral on food" to the statement. But if you could use these half-and-half improvements to make more tiles be food-neutral, then the AI is free to mix and match as needed.

What'd also help? A Farmer specialist. Call it what you want (Baker/Butcher/whatever), but a specialist that adds 3+ food to a city would allow you to find other ways to have a city break even on food, without needing a specific number of farms.

 

[Hesha]

I don't see a problem with the regular tiles either. They are okay, but the resources are indeed very weak compared to Civ4. A pigs grassland tile produced 6 food in Civ4, when pastured. I think the best thing to do would really be to improve the bonus ressources when connected. That would also make up for the pain in the butt fact, that you cannot build trade posts on food resources and are instead forced to build a farm for only 1 food extra! Bonus ressources should offer up to four points of food, production and gold, where cows might have +2 food, +2 production and wheat have +3(maybe 4, since there are no piggedies) food, etc. what do you think?

 

[Pazyryk]

Lots of interesting stuff above. Here is another thought I've had. When you start mucking around with tile yields, you have to consider other things and one of these is the "basket" size for city growth (I mean the amount of food needed for next pop growth in city). One complaint folks have for Civ5 is that city growth in big cities is too slow, making that 1 pop too expensive relative to the benefit it will provide. So 1 pop might cost 20 total food in small city but 200 in a large city -- although 1 pop can be worth more in a big city, it is never 10x better. How is this related to tile yield? I think there are two conceptual models for city growth where tile yield and basket size (or really basket growth) need to be considered together:

1. Slow and steady. Most tiles are self-supporting or better. Basket size increases with an exponential function. The basket size really has to increase a lot because food intake increases a lot as population goes up. If you experiment around and try Civ5 without ICS or rapid expansion (which I know is non-optimal, but you can do it), you might notice that it works very much like this. It's quite easy to have 1 pop growth per 8 or even 6 turns throughout the game, having city sizes of 35 - 40 by turn 270 (i.e., 1900s) and without really bumping the happy limit at all. Again, I'm describing the situation if you don't ICS. (If you do ICS, you probably will have won long before 1900 anyway.) This is just my opinion, but I think this is the way Civ5 is "meant to work."

2. Fast change that is limited. Here each city is limited by some factor such as happy or health caps from Civ4, or perhaps food in a marginal or a very big city. For most cities (with good food tiles), you take an alternating strategy of: (a) work high food tile to grow city very quickly, (b) hit limit, so shift tiles worked, (c) raise limit in city by building to raise limit, (d) go back to (a). City size can "turn on a dime," so to speak, changing 1 pop in a few turns. You can have great food tiles (like pig in Civ4). Although the basket may grow, it doesn't do so nearly as fast as Civ5. It doesn't need to because it is not the limiting factor. The main limit here are things that affect the cap, which are luxury resources and tech level (which determines what limit-raising buildings you have). Or you may reach a sort of max carrying capacity for the area based on total food intake, either sooner (for a marginal city spot) or later (i.e., when your city is very big).

The main point here is that city size is limited by:

• in #1, time
• in #2, other things (geography, tech level, etc.)

If you go with a growth model like #1, then you have to have a fast growing basket and you can't have stand-out super food tiles. Otherwise, you have a runaway situation where more pop -> more food intake -> ever faster growth.

However, if you go with #2, then you can have more tile differentiation. You can reduce the basket growth (or perhaps do away with it? I don't know, but would be interesting to try).

I'd like to see some modding that goes more in the #2 direction. That would involve changes in both tile yield and basket growth.

In any case folks should think more in terms of ultimate "desired" effect. This may differ between us: Spatzimaus' likes "flexibility"; for me, geographical/city diversity is a bigger priority (although the concepts are related). Little tweaks like upping food resource yield, in isolation, are unlikely to do what you want.

 

[Hesha]

I think the dumbest thing in Civ (ALL versions, except COL) is that there is no "global food". That is SO ridiculous...

And I don't think you get any problems with balance if you improve some tiles' yields by one... it will only get the game closer to Civ4 again!

 

[alpaca]

Actually, I think you can have the slow&steady growth even if you boost some tiles like bonus resources. It will increase the growth speed for the city but I don't see that as a bad thing because to grow quickly in vanilla you really have to forsake all other things and the only thing that works is a grassland river.

I actually prefer this to the civ4 (local cap) version.

 

[Pazyryk]

I actually prefer this to the civ4 (local cap) version.

I do to, at least for the happy limit. I hate local happy limit in Civ4 because it leads to an empire where all cities are nearly the same size (no city diversity). However, I would be happier if local geography had a bigger effect (because this increases city diversity, at least for good map scripts like Tectonics and PerfectWorld3).

If you push up tile yields only, then you would (to state the obvious) speed growth and bump into the global happy limits a little more. Tiles would be less bland and that is a good thing. However, with the large exponential basket growth, city growth would still feel about the same: cities are always growing slowly but never reach the full size supported by the land/tech level. (Unlike Civ4 where you had spurts of growth each time you bump up the limit.) That's fine if you like it. I think it is what the designers wanted. The only "objective" complaint I have (if there is such a thing) is that this system requires a situation where 1 pop costs an enormous amount of food in a big city. I don't like that much.

This is mostly just a thought experiment for me. I can test out ideas here and if they are not thoroughly destroyed by analysis and critique (which is very thorough, did I say?) then test in mod. I've tried some but not all "balance" mods, and while I like a lot of individual ideas many of them keep the same basic "slow and steady" growth of Civ5. Or else they break the game (e.g., this happens if you reduce basket multiplier and exponential a lot without other drastic changes). I'm trying to figure out a system where 1 pop costs the same in food whether in a small city or big city (i.e., multiplier and exponential = 1) and it is not broken. Perhaps impossible. If it is possible, such a system would have to have some kind of limit on growth other than the basket size, which pretty much limits me to happiness or total food yield (without adding new mechanisms). I never really enjoy hitting happy limit in either civ4 (local happy) or civ5 (global). On the other hand, I was always happy in Civ4 when I maxed out a city to its food limit (size 25 or 30 or whatever, but 0 or +1 on net food). This lead me to think that I would like total food yield (which is strongly modified by tech, of course) to be the prime determinant of city size.

So here is the system I'm considering. I'm not trying to recreate Civ4. My objective is that cities are limited primarily by geography and tech level, but that they reach this limit quickly. (And secondarily, that I don't have to fuss with the happy limit too much.)

• Drastically reduce (or even remove) food basket multiplier and exponential (a bigger basket early but much much smaller later)
• Raise food/pop to 3 or 4 (which, by itself, is exactly equivalent to gimping all tiles for food).
• Leave most base (unimproved/no-resource) tiles with current low and bland yields (although I would give water tiles +1 food).
• Boost all resource tiles. Say +3 (food/production/gold or a mix) instead of +1
• Farms are only +1 but have 3 or 4 tech boosts (so they may reach +5 but only near end game; there are many techs like Refrigeration where this could happen) Food/tech boosts for most pastures/resource and some plantation/resource combos too.
• Possibly additional food boosts from tiles or buildings spaced throughout the tech tree.

I believe this would give you a system where you had both really big and really small cities (based on geography) and that size would surge at specific techs. The surge would be short (due to non-growing basket) and you would quickly reach "carrying capacity" again based on new tech level (cities would, as they do anyway, prioritize food and then overflow onto mines and such). Grassland and plains would be less different (they only differ by 1 when a tile might be 5 or more food). The city size at any time would be pretty much determined by local geography plus current tech level (no need to mention improvements since that is a given).

 

[JohnnyW]

I'm following this thread closely as I am currently debating these ideas for my own mod. I was actually leaning towards making tiles yield base 3 instead of 2 and increasing food requirement per pop to 3 as well, similar to what Pazyryk described.

With this system adding +1 yield to a tile only grants a +33% bonus instead of a +50% bonus, which is where I think a lot of weird issues come into play in the early game, especially when you hit CS and you get +100% bonus from riverside tiles (less when you consider the free gold....but I'm probably removing that in my mod).

Basically, this system (or one with an even larger multiplier to base yields/food reqs) would allow more gradients to bonuses, which I consider a good thing up to a certain point. I feel like you could get away with a base of 4, but 5 would be pushing it IMO.

 

[Deussu]

I'm following this thread closely as I am currently debating these ideas for my own mod. I was actually leaning towards making tiles yield base 3 instead of 2 and increasing food requirement per pop to 3 as well, similar to what Pazyryk described.

With this system adding +1 yield to a tile only grants a +33% bonus instead of a +50% bonus, which is where I think a lot of weird issues come into play in the early game, especially when you hit CS and you get +100% bonus from riverside tiles (less when you consider the free gold....but I'm probably removing that in my mod).

Basically, this system (or one with an even larger multiplier to base yields/food reqs) would allow more gradients to bonuses, which I consider a good thing up to a certain point. I feel like you could get away with a base of 4, but 5 would be pushing it IMO.

At the moment increasing tile yields by 1 would make Golden Age a bit less fantastic. To still keep golden age a formidably advanced period, something should be done to enhance it.

 

[JohnnyW]

At the moment increasing tile yields by 1 would make Golden Age a bit less fantastic. To still keep golden age a formidably advanced period, something should be done to enhance it.

I believe you can mod that in XML.