[GS] Shall we inherit number-of-city Science and Culture Penalty to Civ6?


May 25, 2017
In Civ5 and CivBE we have that, 5% more cost on techs and civics for every city you have.

This reduces the power of domination by penalitizing the benefits from having lots of cities.

Shall we inherit this? In fact I don't know why they give up this feature. They really shall inherit this from Civ5 and CivBE.

An implementation:

We start at 100% cost, and eureka/inspiration provides 50% cost of techs/civics, when you settle your first city all of your costs become 105%, then follow by (100+5n)% cost if you have n cities. However, eureka/inspiration still and always provides 50% of the base cost.

So if you have 20 cities under this rule your cost of techs/civics will be 200%, however eureka/inspiration still only provide 50% so you need to research the following 150% yourself.

This may help reduce the power of domination and make small empire compatible.
I don't see how this would make the game more fun. I enjoy playing with a small and/or peaceful game but I think it should play the same as large and/or domination games. There are places where cost benefit could be adjusted so that expanding always beneficial. It's always worth having more campus because they have such little maintenance cost.

Maybe you could make a mod that would do this? If it's popular enough it might get adopted by Firaxis.
You would expect many things in this life to scale correctly and would think that the more pop you have to more science you get in equal proportion. However the reality is there is a lot of complex things involved and typically within one civ I would expect less value to come with greater scaling. I think in game in a SP game does it alter the game for the better? Or do people prefer the OP ICSto the turn escalating science dampening.
I am honestly not sure.
So the biggest cost to Wide Empires is Settler and Builder and District cost. I remember at one point to buy a builder or settler it cost 3k gold or 5k gold.

At one point I could either build the Eiffel Tower for 20 turns or a settler for 24.

I did build 20 cities that game and 4 joined via culture pull.

I will say - There should be some extra cost. I notice that as a human vs AI scenario - going multiple cities is an easy way to out-scale the AI. Civ5 also rewarded Tall really well - like the Science and Culture scaling was insane with Pop. Now the buildings just give a base yield so it makes more sense to build more campuses with more adjacency bonuses than to build taller cities.

Also I wish taller cities resisted loyalty more. You can grow a city to be very populus but it can flip. It makes more sense to build 3 smaller cities so they stay loyal.
Also it goes without saying - multiplayer if you were to build a wide empire you better defend that wide empire. So far I haven't seen the AI mount dual invasions in a combined manner yet. Emergencies are getting there though.
It makes more sense to build 3 smaller cities so they stay loyal.
Nope. A pop10 city has more loyalty than a pop 4 city supported by 2 pop3 cities.

A pop10 city has 10 points of citizen loyalty pressure
A pop4 has 4 + 2*3*60% = 7 points of loyalty pressure

You are probably thinking of when taking cities using other cities to support it but that is because you have issues growing the pop of the taken city.
Just read your loyalty guide
needs a bit of an update but the key points are sound. You can just take a builder or 2 and chop in the pop.
This leads me on to the Eleanor twins.
Forward settling tends to face off other ickle cities. Chop in masses of pop and watch them flop. And any other cities within 10 that may have gone free before you arrived will instead decide they prefer you. Seems to work a treat
I would rather see a solution that target domination directly instead of punishing wide empires, something more aligned with how the alliance system works but with a meaningful effect on science/culture, in a way it punishes the player for being a warmonger instead of punishing for having a large empire, encouraging players to use all tools available to avoid grievances, instead of encouraging them to go tall. I would like to see a system that really force you to stop and think before you declare a war and conquer cities, so you can't just steamroll eliminating everyone, surprise war after surprise war. If you go out conquering without much consideration, the tendency should be for you to get weaker and struggle to keep your Empire together, giving opportunity for other Civs to strike back if you get too greedy. War weariness try to do that but the effect isn't always felt and can be ignored more often than not.

Another thing that they could target is military strength, adding a mechanic that force you to choose between having a large army formed by weaker units that eventually might get too weak and a smaller army with better units. I'm not sure how that would work but it make sense for me that the larger your army gets, the more you will struggle to keep your units trained, well equipped and modernized. Some mechanics try to achieve that, like unit upkeep and late game strategic resource cost per turn, but none have a meaningful effect, to really force the player to hold back or risk seeing his powerful army turn into an expensive joke.

What Civ VI could use is to expand the concept of loyalty and make a larger Empire that was established through war harder and harder to keep together. Conquering fast and without consideration should make cracks show up all over your Empire, eventually splitting it up with Civil wars and all kinds of crisis, making domination more about knowing how to put out fires on your Empire than just steamrolling everything. Eventually you should need to stop advancing and look inward or risk falling apart while other Civs take back what is theirs and more. Instead of conquest always being something that will benefit you, since the negatives effects aren't as meaningful as the positives ones, the game should have something that create a "ulcer" here and there, like Spain was a ulcer for Napoleon. Sometimes you should just regret conquering this or that city, fighting on this or that war.
I'm by no means an expert so I will just give my opinion.

I do desperately want tall to be more competitive. Frankly, I get tired of feeling the need to continuously expand. I think there should be a balance though. I found Civ 5 was too harsh on wide with its final iteration but I find Civ6 is VERY far in the opposite direction.

I think there are a number of reasonable options.

They could make campus science yield scale better with population generally and possibly lower the inherent science yield of the campus.

They could also increase the cost of districts even more, if you have many of them (yes, I know this exists but I'm suggesting they raise the production cost even more).

The final idea, perhaps the most controversial, would be implementing some kind of system similar to Civ4 where maintenance costs increased the further you were from the capital. Perhaps, without a road connection loyalty is lower and maintenance on buildings is higher.

I'm also of the opinion that culture should be more deeply tied to loyalty and empire management. I propose that culture establishes a civs "area" or "sphere" of influence. The simplest way to do this would be that each tile once again has a % culture belonging to various civs. A tile with majority your culture will have higher loyalty and lower maintenance costs than one with majority another culture. The cultural spread would be an entirely passive process but you could perhaps add in policy cards to help flip tiles cultural identity faster.

How would this play into making tall more viable? A civ with lower culture would incur increased maintenance costs and lower loyalty for cities outside both their loyalty and "cultural" spheres and the along with my other change of roads effecting yields and loyalty would greatly prohibit mass expansion and make players reconsider the time investment required to go really wide vs tall.
They could make campus science yield scale better with population generally and possibly lower the inherent science yield of the campus.

They could buff science/culture from population, then make it so having more cities decrease the amount of science/culture you get per population, in a way that a citizen in a tall Empire is more valuable than a citizen in a wide Empire. Then they could add a policy that buff this science/culture further, similar to Rationalism but buffing science/culture from population instead of buildings, also giving the bonus only to cities that have X population. They could add multiple tier 3 buildings to all districts, like the Industrial zone power plants, with one of these buildings giving yields based on that city population. The campus and theater square tier 3 buildings would specifically add a multiplier to the science/culture from population, making it more valuable for a tall Empire. This way a wide Empire can build more campuses and theater squares while a tall Empire get more science/culture per citizen.
I like the idea of scaling up science and civic costs for wider empires.

It makes sense for a huge unfocused empire to be a little less efficient than a focused medium sized empire. The only downside is whether the AI can manage the increased costs.

A lot of this boils down to empire management. There are no downsides to wide empires and no extra challenges to the player to maintain it.
No, because it is the worst mechanizm in the history. The game about expansion should not penalise expansion. What can be inherited, is slider and maintanance model from civ4.
Civ5 was too harsh when it came to the # of cities penalty. That being said, I think a lite version or some other mechanic that serves a similar purpose is a good idea.

On the other hand, I think that so many systems might need to be revamped or tweaked that it would just be too much tinkering for a game that has been out for 3 years.

That being said, a "quick fix" could be implemented. One quick fix I've always thought about would be reducing the % amount of eurekas and inspirations depending on how many cities you have. This would obvious mean you progress a tad slower, but it would also be a trade-off since wider empires can simply get more eurekas because they can diversify and then specialize more cities. If you only have two cities and no neighbors, it makes little sense to waste a district slot on an encampment -- which means no military training inspiration. Or, if you've just three cities, it might be difficult to get two universities up in time to get the eureka for printing press. Wide empires will encounter this problem a lot less often so they're getting a double bonus -- more base science and culture and more eurekas and inspirations.

China would also get an indirect buff because they could expand to -10% size (however many cities that would be) and only lose their bonus -- not be penalized.
Another solution would be to have some sort of production cost scaling for buildings and/or revamp the district cost system to take into account number of cities. Or to give a production increase number of the same district beyond the second.

There are lots of "solutions." I use air quotes because I'm not really sure this is a problem to begin with. Would making playing tall more viable be good? I think it would. Is penalizing playing wide the way to do it? I'm not so sure about that one.

Maybe a better way of looking at it would be to give tall empires some sort bonus depending on what era it is and how many cities they have? If they're below the cap, they get a +% bonus to yields?
This reduces the power of domination by penalizing the benefits from having lots of cities.
The per city penalty was a very blunt instrument that creates a lot of anti-fun, mainly because it severely curtailed how you should play.

But it is important to look at the differences between civ5 and civ6 and see what exactly is the problematic snowballing we want to curtail.
Yes, in civ5, they had a science penalty added. The game always had a social policy penalty though. Isn't that interesting - what did they identify before release that prompted them to add it?
Well, initially, culture was a pure function of number of cities + wonders. There was no per pop culture, only monument/temple/theater/broadcast building line. So they were definitely going after city spam there.
But city spam did nothing for your science inherently, which was entirely built on per pop effects. But since city spam allowed a player to grow more pops (because of how civ growth works) and the gold to pay for all the extra instances of buildings (this is why BNW came in and heavily curtailed gold) the follow on effect was that they'd produce a lot of science.
In civ6, city spam is only useful as a proxy for spamming campuses. You can have 10 cities to my 5 and if I have 5 campuses and you have 2, you're going to lose.

City spam =/= science spam. It is naive to target cities like this when in civ5, pops+gold were the underlying issue (they solved the gold problem in BNW) and in civ6, campus buildings+CS bonuses are the issue. (When i talk about civ6 science, I am implicitly also referring to culture/theater squares, its just easier to say science.)

The big problem with per city flat penalties is you end up with one of two regimes. Given the balance of the rest of the game, either you:
create a peak point beyond which you start losing if you keep expanding (Penalty is high, like civ5.) Players hate this effect because they can see things getting worse.
create a peak point that is so far out that it doesn't matter and you've only slowed down the game pace (penalty is very low, eg a 0.5% penalty would mean you'd need like, 50 cities or something, so practically there might as well not be one.)
Keep in mind that the per city penalty did nothing to bring science spam in line, at all. The return on science was still linear, it just mattered if it came from 4 cities or 10. How often did people drop everything to national college rush (aka science spam!!) in G+K vs BNW? Exactly.

Note: the next section is general education and you can skip to my science spam proposal
Spoiler :

It's mostly a mental effect but we have to keep in mind that the game is played for fun and we have to consider options that basically will trick the brain. Players like when more yield is monotonically beneficial - that is, more input means more output, although it may not always be 1:1. One mathematical function that works really really well in this sort of case is arc-tangent function, or inverse tangent. (You may recall it from a trigonometry class.) But the Arctan function has an amazing property of being almost linear near zero, and then leveling off towards infinity:

(ATAN is what the function is called in excel.)
It works for positive and negative numbers, and by scaling the function we can make that value it asymptotically gets closer to whatever we want; by scaling the input we can extend that linear region to whatever we want too.
EG. f(x) = a*arctan(b*x+c)+d will let you move it to wherever you want.
"Sostratus," you say, "that math mumbo jumbo is cool but how does this relate to science spam?"
Here's an example in a quick excel chart:

Under civ5's linear penalty we can see there's a peak point on the orange line. The blue line is just input-output 1:1, which is civ6. But the grey line is a scaled arctangent penalty. Notice how it's very close to Blue line at the beginning, then slowly starts to taper off. Imagine if input in this case was a research rate like % of a tech per turn, and output was the adjusted value. With an arctan penalty, we can guarantee that more science in means faster research, always, but we can also set some safety in the system by keeping players from getting too out of control. In this case the grey line will approach but never hit ~12.
Implementing that would require inserting some math and people might not like feeling like they are directly penalized once they realize it even exists. Okay, fair.

Science Spam Proposal starts here
But. We can use the principle of arctangent (linear in the middle, adjustments towards the edges) a lot more cleanly by just using what's already in the game: adjusting tech costs based on world era. It's not science itself we want to control, it's when players get ahead of other players. All we really need to do is take what in there now (currently, any tech more advanced than the world era is 20% more expensive) and just expand it- in both directions. Most eras are 2 tiers of techs wide. Let's just say we formally class techs into tiers. Each tier outside of the current era gets, say, a stacking 25% cost factor, X=1.25*N. Techs ahead get a cost adjustment of BaseCost*X. Techs behind the current era get an adjustment BaseCost/X. This way, The effective rate of research stays monotonic (more science still means you always research faster; you will never overtake someone with more science output than you have) but it becomes less and less efficient to keep investing in science the more ahead you are, and more and more efficient the more behind you are.
In this example, players who produce X science will be at the current world era. Players who produce 1.5X science will remain one era ahead, while producing 0.66X science will remain one era behind. In this way, civs can make a comeback, but only up to the current world era. I suspect the 25% is actually too low, but it would take a few games to suss out that. This one change right here will not solve science spam, but it will leave it limping. It's also AI friendly.

Then we have this issue:
There are no downsides to wide empires and no extra challenges to the player to maintain it
Where civ6's district system is so ICS friendly that large empires win on every level a little too easily. Domination is a proxy for the fastest way to get large empire. I agree that there's an element of empire management missing where a large, sparawling empire should feel a little like they are fighting themselves. I suggest two simple changes:
1) Cities lose the one free amenity they get. This right here is a subtle effect that in civ5 would be like if you gained 3:c5happy: per city instead of 3:c5angry:. It's super strong and doing away with it would be great for balance. In exchange, we could bump up the amenities from luxuries to, say, 5 cities instead of 4.
2) Gold income gets shifted to be more scarce. Without sweeping changes to districts which I could write pages on, simply toning down CH gold a little and re-balancing maintenance costs of districts, units, and buildings so that you really need to watch your income would be a great change in empire management. Go ahead and try to spam nothing but campuses - just watch out, because you'll have to pay for all that somehow. I would suggest trading some CH gold income towards +trade route bonuses on the buildings themselves, to really focus trade routes as a thing.
These are both changes that are little more than table updates.

To be more extensive I'd want to bring back city connections and change loyalty a bit, but I want to stay on topic RE:science spam.
I guess Amenities were supposed to do that in a way.
I am of oppinion that wide empire in Civ6 is not unbalanced, if expanding peacefully.

There are more then enough effects in the game that limit this expansion, although if you do play that style you do want to limit city growth to only focus on 4+ yeld tiles at first, at least until settlers become prohibitively expensive making tall good again.

But truth to be told this style only makes sense on big maps with lots of space to expand.

On the other hand, what is really unbalanced is military expansion. Currently getting proper units is cheapest way to get more culture, science, districts and any wonders you missed, then going tall or even peaceful wide, and on top of that you do not only get stronger in everything but also destroy competition in the process.

Also those increasing costs of district benefit warmongers the most, since unit costs are fixed, and most of the core army will be upgrades of older units that were originally produced cheaper, not building new expensive units from scratch, so conquering those districts is optimal strategy.

Also, the way how difficulty works, the higher the difficulty the more rewarding is conquest, since opponents will build the more in their cities.
Last edited:
Top Bottom