Must you build a Commercial District in every city? Are they really that OP?

[MyOtherName]

In earlier civs it was combatted with corruption. Coruption would prevent us to get too many cities (going too wide). Not only did people not like coruption, it also had the adverse effect of leading to ICS: the practice of building 100s of towns just to have a few scientists in each.

It wasn't the existence of corruption in Civ 3 that pushed ICS, but the fact corruption only affected trade and shields, and furthermore never eliminated the first shield.

Civ 5 did the best job so far at balancing these things, It did so well that we actually go to choose between tall and wide, hence this term was introduced.

My recollection is that Civ 5 was very ham-fisted about this -- there wasn't a tradeoff between tall and wide, but instead an artificial barrier that you're penalized for crossing unless you commit fully to doing gimmicky things.

Unsurprisingly, I think Civ 4 had it best.

On the one hand, neither the current value nor the future potential of your existing cities are reduced by placing new cities or by growing them, so Civ 4 doesn't have the failings of Civ 5/6 in that regard.

On the other hand, the costs of wide expansion are such that you can get meaningfully better returns on your investment by focusing on technology and developing core cities.

Right now its only the minimum city distance really that stops us from ICSing the way we did before. But the principle is very much the same as it was in civ3 where we would fill the map with towns 1 tile between them, running scientists in each.

The increasing settler/builder costs and inflation of district costs also work against ICS. The fact you get full yields from your citizens eliminates one of the motivators for ICS.

I think that the biggest push towards ICS is the rapid growth in how much food it takes to grow a large, especially in light of the fact that tile yields are so poor.

 

[WackenOpenAir]

True, if the minimum city distance were small, the increasing cost of settlers would play a very important role. Now we just conquer AIs and they dont leave that many gaps to settle anyway.

I cant comment on civ4. I have played 2,3,5,6. Of course now people are telling me 4 was the best

In my view, the push towards ICS in civ6 is simply the lack of benefits from a bigger city. But of course it is both. It is the tradeoff between what it costs and what it provides.

The costs:
-lots of food (which can actually be produced reasonably effectively with the farms adjacency bonus, which people now ignore)
-lots of production invested in amenity and housing
-The need to purchase tiles for the citizens to use because cultural border expansion does only a few tiles and then quickly becomes a non-factor even with monument. (the culture need per tile grows even much more exponential than the food need per citizen growth)

The benefits:
-0.3 culture per citizen which is rather insignificant
-0.7 science per citizen which has to compete against empty districts providing 10 science with help of CS making it also too insignificant.
-some food and production per citizen which are primarily needed to make the citizens grown in the first place.
-the ability to build more districts which we dont really need beyond 3 / size 7.

All together, we could estimate the price of a midgame citizen (which is when we would grow it beyond size 7) to be something like 30 production for the tile imporvement it needs to work, 20 production for the extra housing needed, 150 gold for the tile and lets say on average 150 food to grow it itself. Converted to shield/food resources that would make about 240 ish production/food. If the tile produces 5, -2 food eaten, that makes 3 surplus and thus an 80 turn time to repay for the investment. (ok, not counting that super significant total of 1 science/culture)

On the other hand, we spend 500 production on a little army, it captures half the world for us.

 

[Magil]

The games prior did concern themselves with it although it didnt have the same label.

Without getting into a very long block of text about it (which I could type), what the previous games of Civ did was try to make sure that each city had an ongoing cost of some sort. Corruption tried reducing yields for expansion. That didn't work. Civ IV's maintenance system neatly curtailed true ICS but still held true to the general rule that more is better than less. As it should be in a 4X empire-building game. Any time you see resources nearby and you don't want to rush to claim them before another player can starts signaling alarm bells in my head.

I'd like to point out: Vanilla Civ V brought ICS right back. It took them until BNW to fix it, and look what they had to do in order to do so: culture penalty, science penalty, National Wonder restrictions, happiness penalty, diplomatic penalties (beyond normal border friction, they actually included a specific negative modifier for simply having more cities than the average player), etc etc. To me, it was not very fun to be penalized in so many different ways for simple expansion.

I don't know if Civ VI has it right currently. Probably not. But my preferred approach is this: expanding should have a cost, and the cost should get higher the more you expand, with a soft-cap somewhere in there to prevent it from scaling too much. But the city should be able to pay for itself with some investment. I do not support any artificial benefits for staying small, it should only mean you don't have to pay the expansion costs. Of course, in doing so, you lose out on potential mid to long-term gains. Which is bad. But may be necessary at times. Building is important, I love to build up and develop cities. I simply like to have lots of cities to build in. Not a half-dozen.

 

[ShinigamiKenji]

... And I think we derailed the thread again. Maybe I should put something like that in my signature (and finally get around having one). Going back to the issue of building or not CH:

After reading all the discussion, I believe it's fair enough to say you'll build CHs in your core cities. They'll develop enough to have both the pop and the production to support it, and early enough to reap the benefits. The discussion lies, then, in whether they're worth it in your satellite cities.

After running mental experiments (no time to apply the theory yet), I'd say the need for another one is tied to when you expect the city to reach 4 pop, if at all. 4 pop means you can get your victory district and a CH/Harbour. Let's run some calculations.

I think it's fair to say that the additional value of a CH outside your core cities lies mostly in another Trader. If you need gold, you can build Markets and Banks in your core cities, which are probably far less expensive than another district. You can also run policies for some time to help with immediate needs. If you've built Markets and Banks in all core cities, then probably you won't need much more money, or you're overdoing something, be it units or buildings. GPP are covered by your core CHs and their buildings too, and can always be patronized. If you absolutely need Adam Smith or maximize gold for Big Ben, then by all means build CH, but you may want to re-evaluate your strategy and redirect your efforts at somehing else (such as projects for GPP). The extra cog for trade routes is nice, but not many trade routes are going to a satellite city anyway.

Having cleared that, let's crunch some numbers. A Trader, on average, brings 2 food, 3 production and 3 gold (assuming half the game with Caranvansaries, and half with Triangular Trade) every turn. Now, how much does a city need to reach 4 pop with a CH and a Trader, and how much production do those cost?

To reach 4 pop, a city needs 102 food (from the Formula Thread). The production cost is more hazy, because CH cost depends on when you placed it (Locked it right after Currency or not? After how many techs?). I'll say you locked its price after AH, Archery, MIning, Bronze Working, Pottery, Irrigation, Writing and Currency; those are all needed early game for defense, worker improvements, jungle chopping and prerequisites. According to the formula thread (with a slight mistake corrected), after 8 techs, districts cost 118 production, so let's roll with that number for now. Traders cost 40 production, according to Wikia. So in total it's 158 production.

So, in how many turns does it take for a new trade route to repay itself? It's roughly 51 turns for food, and 53 turns for production. Gold is a nice plus, too.

But this calculation leaves out the opportunity costs for other districts. How important is an early Campus, Theatre District, Holy Site, or Entertainment Complex + Colosseum? You might have advanced the tech tree quite a bit more if you locked a Campus instead of a CH right after Writing. You may get a better government much earlier if you built a Monument before a CH.

All in all, it essentially boils down to when do you plan to have them completed. Built before turn 80, they're a good investment. After 140 turns, not so much. It can be hazy between turns 81 and 139, because it's unclear if it's better to have more science/culture for an edge, or another trade route to develop more infrastructure. Those are better dealt case-by-case.

 

[WackenOpenAir]

Without getting into a very long block of text about it (which I could type), what the previous games of Civ did was try to make sure that each city had an ongoing cost of some sort. Corruption tried reducing yields for expansion. That didn't work. Civ IV's maintenance system neatly curtailed true ICS but still held true to the general rule that more is better than less. As it should be in a 4X empire-building game. Any time you see resources nearby and you don't want to rush to claim them before another player can starts signaling alarm bells in my head.

I'd like to point out: Vanilla Civ V brought ICS right back. It took them until BNW to fix it, and look what they had to do in order to do so: culture penalty, science penalty, National Wonder restrictions, happiness penalty, diplomatic penalties (beyond normal border friction, they actually included a specific negative modifier for simply having more cities than the average player), etc etc. To me, it was not very fun to be penalized in so many different ways for simple expansion.

I don't know if Civ VI has it right currently. Probably not. But my preferred approach is this: expanding should have a cost, and the cost should get higher the more you expand, with a soft-cap somewhere in there to prevent it from scaling too much. But the city should be able to pay for itself with some investment. I do not support any artificial benefits for staying small, it should only mean you don't have to pay the expansion costs. Of course, in doing so, you lose out on potential mid to long-term gains. Which is bad. But may be necessary at times. Building is important, I love to build up and develop cities. I simply like to have lots of cities to build in. Not a half-dozen.

I dont see any disagreement here really (btw i never played civ5 vanilla, and i now kind of feel like should not be playing civ6 vanilla)
Indeed more is better is how it is supposed to be, with some diminishing soft cap, i agree there.

In the suggestions i linked, i did not suggest limiting growth, in fact i did not even suggest much of a limiting to expansion. Just to make conquest seriously more difficult so it cant be done so massively and so early.
Meanwhile growing your cities should be rewarding because as it is now, expansion is not penalized, but growing your city taller is made pretty much pointless.

 

[WackenOpenAir]

To reach 4 pop, a city needs 102 food (from the Formula Thread). The production cost is more hazy, because CH cost depends on when you placed it (Locked it right after Currency or not? After how many techs?). I'll say you locked its price after AH, Archery, MIning, Bronze Working, Pottery, Irrigation, Writing and Currency; those are all needed early game for defense, worker improvements, jungle chopping and prerequisites. According to the formula thread (with a slight mistake corrected), after 8 techs, districts cost 118 production, so let's roll with that number for now. Traders cost 40 production, according to Wikia. So in total it's 158 production.

So, in how many turns does it take for a new trade route to repay itself? It's roughly 51 turns for food, and 53 turns for production. Gold is a nice plus, too.

But this calculation leaves out the opportunity costs for other districts. How important is an early Campus, Theatre District, Holy Site, or Entertainment Complex + Colosseum? You might have advanced the tech tree quite a bit more if you locked a Campus instead of a CH right after Writing. You may get a better government much earlier if you built a Monument before a CH.

All in all, it essentially boils down to when do you plan to have them completed. Built before turn 80, they're a good investment. After 140 turns, not so much. It can be hazy between turns 81 and 139, because it's unclear if it's better to have more science/culture for an edge, or another trade route to develop more infrastructure. Those are better dealt case-by-case.

I dont really see where you got 51 turns from. after saying the trader gives 5food/prod and costs 158 total. Did you calculate the price of the city as well ? (you shouldnt, it can do more things than build the CD and trader, the town is an investment in itself which i consider as follows: A settler costs 80+x production and around 30ish food depending on the size of town that builds it. A freshly settled town provides around 6 production/food so repay time is around 20turns for the first settler, more for later ones. However the town also gives you the opportunity to make some really magnificent investments in extra citizens as the first costs you i believe 15 food and probably adds 3 (4 from tile - 2 food + 1 scienc/cult, simply valuing the latter about equal for now) output to your town, thus having a repay time of 5. The second, third and fourth will also have a good repay time, after that, the value of growing citizens starts to decline into unfavorable numbers.)
However there are so many factors gong on that makes it difficult. Ill give my view on the cost/reward of the trader:

I dont see much missed opportunity cost. Growing to size 4 happens very rapidly so when you unlock the commercial districts, most of your cities can have locked in both the science and commercial districts no problem. Later when you get the industrials, you want them to be size 7.

If you build them early, your trade routes probably only give 4 food/production, later on it grows higher. That makes it kind of hard to put a number on the repay time of them. For just the trader cost we can say around 10 turns, which is magnificent. (i usually considered 20 turns repay time the benchmark in civ though in civ6 things generally seem to repay themselves a little less fast)
If we add the commercial district, that can well provide 18 gold while being empty. 18 gold is worth about 6 production, so with a cost of 120ish, that would be about 20 turns repay time. Which is pretty good too. This does however also not happen overnight, you need 10-12 envoys for that, and i cant really put a price on those.

 

[Magil]

For me, a lot of the reason why I instinctively build a commercial hub in each city is a desire to have a city "pay for itself". I almost always try to arrange that a newly-settled city receives at least one trade route (preferably two) to help it get off the ground quickly. The extra food and production makes a huge difference in the development of a fledgling settlement. Thus, in order to "pay" for those trade routes, it needs at minimum a Commercial Hub, and preferably a Harbor as well.

Now, does a city really need such things? I suppose it matters most to keep the grand scope of your eventual victory in mind, and what contributions you expect out of this new city towards it. Like I said before, I like to build. I'll not deny that to some degree it can be even a bit "roleplay"-like that I like the idea that a city is "off to a good start", even if it's not going to contribute fully towards victory (as an aside, this is one reason I disliked V's penalties, because they scaled quite steeply as % and could actively hinder your victory if you founded a city "too late"). But ultimately I think this is a legitimate concern throughout the game in terms of best use of resources, assuming you're playing a peaceful game, it's likely that founding and supporting a new city can have better effects within a few dozen turns than simply running district projects over and over, because it doesn't require much effort on the part of your other cities other than a few trade routes.

I haven't done any number-crunching, and I am certain there is better room for optimization here. It's hard to judge when the game isn't terribly challenging even on the higher difficulty settings, partially because the AI does not pursue victory in an intelligent manner (to be fair, this is not really a new thing in Civ VI). So I suppose the conclusion I'd make is this: it seems to me that continuing to invest in commercial hubs is a solid choice so long as you intend to keep expanding.

 

[ShinigamiKenji]

I dont really see where you got 51 turns from. after saying the trader gives 5food/prod and costs 158 total. Did you calculate the price of the city as well ? (you shouldnt, it can do more things than build the CD and trader, the town is an investment in itself which i consider as follows: A settler costs 80+x production and around 30ish food depending on the size of town that builds it. A freshly settled town provides around 6 production/food so repay time is around 20turns for the first settler, more for later ones. However the town also gives you the opportunity to make some really magnificent investments in extra citizens as the first costs you i believe 15 food and probably adds 3 (4 from tile - 2 food + 1 scienc/cult, simply valuing the latter about equal for now) output to your town, thus having a repay time of 5. The second, third and fourth will also have a good repay time, after that, the value of growing citizens starts to decline into unfavorable numbers.)
However there are so many factors gong on that makes it difficult. Ill give my view on the cost/reward of the trader:

I dont see much missed opportunity cost. Growing to size 4 happens very rapidly so when you unlock the commercial districts, most of your cities can have locked in both the science and commercial districts no problem. Later when you get the industrials, you want them to be size 7.

If you build them early, your trade routes probably only give 4 food/production, later on it grows higher. That makes it kind of hard to put a number on the repay time of them. For just the trader cost we can say around 10 turns, which is magnificent. (i usually considered 20 turns repay time the benchmark in civ though in civ6 things generally seem to repay themselves a little less fast)
If we add the commercial district, that can well provide 18 gold while being empty. 18 gold is worth about 6 production, so with a cost of 120ish, that would be about 20 turns repay time. Which is pretty good too. This does however also not happen overnight, you need 10-12 envoys for that, and i cant really put a price on those.

It's 2 food and 3 prod on average for me (if you send it to a core city, which will probably have a CH, a victory district, and maybe an Encampment or IZ). You need 102 food for a city to grow from size 1 to size 4 to support your victory district as well as a CH, so you need 51 turns for the trade route to repay the food. Also, you need 158 production to build a CH + Trader, if you lock down district costs, so the trade route repays the cogs invested in 53 turns.

I didn't try to quantify the impact of gold, because units, buildings, tiles and GP have different costs, and you can't buy districts. But of course if you account for that the breakeven should occur earlier. You can expect to have 6 envoys in a Mercantile CS, but don't count on many more. I've seen games with only 1 Mercantile CS at all, and it takes some time before you can get 6 envoys on many of them. Especially with many others offering direct bonuses to your victory path and good suzerain bonuses.

We can do the calculation under another lens: given that you'll build a CH as soon as you place a city, how many turns does it take to be able to build your victory district?

We might average every single tile as a 2f-2p tile. That way we will have a reasonably productive city with a reasonable growth. It takes 8 turns to reach 2 pop, 12 turns to reach 3 pop, and 17 turns to reach pop 4. It's 37 turns so far. Production wise, it's an accumulated 74 production so far. Assuming then you switch to production focus (working instead 2 1f-3p tiles), you need 5 more turns to finish the CH, for a grand total of 42 turns.

You can hasten it using trade routes, but then you'd need to divert traders to that city for at least 22 turns. That's 44 food and 66 cogs less to the rest of your empire in the meantime.

Can the benefit (including gold) of a new mid-game trade route compensate for all these factors? Can the early science output of a Campus, or culture/tourism output of a Theatre District, of faith from a Holy Site, outweigh the cost/benefit ratio of CH? Those are the questions you need to answer before placing a CH.

 

[MyOtherName]

Only 75 (ish) food to grow from size 1 to size 4; you're plugging in the wrong value (off by 1 error). I'm not entirely sure it's fair to count this cost anyways, since you were probably planning on growing the city in the first place; small city population points are extremely cheap.

By the way, if you're big on trade routes, you should earmark one or two core cities to be your trade targets, and grow them large and choose districts based on the trade yields. (e.g. one cog city, one food city)

 

[ShinigamiKenji]

Only 75 (ish) food to grow from size 1 to size 4; you're plugging in the wrong value (off by 1 error). I'm not entirely sure it's fair to count this cost anyways, since you were probably planning on growing the city in the first place.

By the way, if you're big on trade routes, you should earmark one or two core cities to be your trade targets, and grow them large and choose districts based on the trade yields. (e.g. one cog city, one food city)

I got this from the Formula Thread, which is quite accurate at the beginning of the game in my experience. I also don't like placing districts solely based on trade route yields, for two reasons: it's probably a sub-optimal location (Harbours don't mix well with Factory AoE, for example), and that city will eventually become too far away for the trade route be completed (and thus freeing the Trader) in a reasonable time.

The point of all of this is: When, if ever, is it good to stop building CH in favor of other things? Is it better to use those trade routes to build yet another one for more rewards later on, or is it better to begin investing heavily in your victory path? I've made two rough calculations: the Trader breakeven, and time until next district. What should we include in our assumptions, and how (like gold impact)?

 

[MyOtherName]

I got this from the Formula Thread, which is quite accurate at the beginning of the game in my experience.

Formula thread says 73 food.

 

[greygamer]

So the way I am looking at it now.

You need some CD's at the start and the more the merrier. It just quick starts your game and the necessity is stronger as you go toward deity.
At lower levels the need becomes less faster but also there is not so much competition for other districts unless truing to be as efficient as possible.
As time moves on there is still need but from what I see some of the need is not always a sensible use.
The game is not challenging enough to really nail down its requirement and everyone has quite different views

We will have to wait for the G&K moment, that changed Civ V significantly, just go back and play vanilla Civ V and you will see what I mean.

 

[WackenOpenAir]

Yes, but as i said, thats not part of the cost of the CD. Its a seperate investment, a very good one too.
Since you usually can choose between tiles producing more food or production, i value food and pruduction as the same in all calculations. I value gold 1/3 production. For now i value science / culture about equal to production because i dont yet know what else to do and it feels more or less correct.
Hence, according to my calcustimations repay times look more or less like this:

Settler: 20 + 3n turns.
Citizens: 5 - 8 - 11 - 15 - 18 - 22 - 26 - 30 (assuming you have tiles with 4 total production/food to work)
Early worker: 17 + 1n

CD with free envoys*: 20
CD without evoys: ~120
Trader: 10
CD without envoys but combined with trader: 25
CD with free envoys and combined with trader: 14

In civ i take 20 turn repay time / 5% ROI as a benchmark average.

Growing a city up to size 7 (actually used the formula now) falls roughly within this, after that it starts becoming what i consider a poor investment. That is of course assuming you have housing and amenities available for "free". pretty much meaning you have built farms or camps, which in turn only makes sense if you use those because you want to use other high production tiles with <2 food. If you have to add the ~20 production from granary cost (20 per housing + 20 for the 1 food), then your repay time is about 6-7 turns more for your citizens and thus, the 5th is the last decent investment. Well, we all knew by experience that it is about that size that we feel growing our cities becomes too much of a hastle, but i like to see it in numbers.

For some citizens might not seem like an investment but something that happens automatically, but for me it is a rough indication of where i prefer production tiles over food tiles. Of course other factors go into that decision as well, but it is clear that for a tiny city, i will prefer food while for a bigger one, i will prefer production as long as i have any decent investments to make. For a size 1 or 2 city, the value of growth is so big that choosing 1 food over 2 production is not insane. For a size 7 city i will pretty much always prefer production over food. Size 3-6 is where it depends on available tiles (If there are no more 4yield tiles available, the value of the next citizen is much less), available things to build and the wish to get 3 districts or not.

The early worker is a decent investment. Later workers might be a little more expensive, but they will have 5 actions and the tile improvements might be worth 2 production/food. That will make them totally awesome investments, potentially with repay times below 10 turns. (not even counting the forest cuts that allow the worker to pay for itself right after production)

Only the first settler seems a decent investment, the civic that makes them cheaper allows for a few more settlers to be a decent investment just by themselves. Of course the fact that the extra city allows you to make other awesome investments (citizens up to 7, tile improvements, commercial districts) makes that you can build lots of them before they really become bad.

Commercial districts by themselves are terrible. With either a trade OR an the effect envoy effect of CSes, it becomes decent. Only with both, it becomes really very good.

When do you build a CS and when not ? First of all, you build one when you have no better investment to make. When you have better investments to make, you do those first. So that may very well depend on when you expect to have the envoys and CS bonusses going. When you have those, the CD is one of the best possible investments. Workers right after getting feudalism are probably an even better one though and therefore can have priority. As long as you do not have the CS bonusses (and also probably your trade routes might not yet be worth 4 food/prod), you may want to prioritize settlers, workers, or units. (cant really calc the repay time of units, but we all know they are an awesome investment in civ6)
Again, im sure much of this is what most players were already doing. I am aware that all this number stuff may seem useless

Beyond that, you stop making them when it is so late in the game that it is better to focus on your victory condition. If the repay time is 14 turns, that is NOT 15 turns before the end of the game but quite a lot earlier. Cant really put calculations on this though, choosing the right moment to swich from growth to victory condition is the key to early victory dates. If you can read GOTM8 threads you can see i made the wrong choise there

*You do get the envoys for free, but still they do have a price of course as using the first ones for commercial district boosting means you are not (yet) enhancing your science districts or industrial ones.

 

[Victoria]

I don't know if Civ VI has it right currently

Agreed, they do have the limits in there that sort of work but I feel the have got ICS right accidentally. Too many trade routes amd too many cities slow the game down so it becomes unpalatable IMO

We will have to wait for the G&K moment, that changed Civ V significantly, just go back and play vanilla Civ V and you will see what I mean.

I went vanilla to BNW but know what you mean. I have my fingers amd toes crossed

 

[Quillan]

Whacken & ShinigamiKenji, i like that you're trying to quantify. Right or wrong exploring these things analytically leads to better decisions and more pew pew. Unfortunately most people are still just saying it works and i like gold so its better, which makes this discussion more about patting themselves on the back for their good decision making rather than analyzing what makes better sense mathematically or strategically.

ie (someone, sry lost quote)- The ability to add more trade routes, and what they give, makes it a priority for me to build CH's in every city! -

I still dont understand why most dont want to acknowledge that x food and y prod and z gold are replicatable by other means so that, for example, growing a city by a couple of more spots is literally the same as a cd + trader (depending on hexes of course, and excluding the road, gpp). ie, building a new city, say on the coast, can easily produce more gold/prod/food than your investment in a cd/trader (prod hex + crab for example). In addition, the opportunity cost of that district is, for example, a campus, which will push you further and faster down the tech tree and allow you to outpace them technologically which snowballs in value. Or earlier culture = earlier polyphi, etc. (Shinigamikenji is considering this, which for those confused about opportunity cost, this is it. The hard part is quantifying various benefits and costs when there aren't discrete payoffs.) Personally I value culture and science more than other people seem to as they're harder to get and provide powerful asymmetric benefits, just like adding a GG to an attack.

I cede the point about using traders to jump starts new cities but of course that's still possible since no one is saying NO CD NO traders, just are they optimal FIRST and in every city. Again, by mid game a lot of my cities have CD, so its not about building them at all.

Anyways, I get that 95% of people dont want to consider that something else may be better than what they're doing (cognitive dissonance is a b1tch, even in anonymous online gaming), but the pursuit of the thought will let those other 5% become better gamers, regardless of whether or not a CD is optimal, so its all good.

The problem with CDs is that it's so rarely a *bad* idea to build one - even if you don't exactly need one, you'll generally always find good use of the extra trade route.

Yup, this is totally fair. Still doesn't make it optimal though, which is what we're aiming towards.

So the way I am looking at it now.

You need some CD's at the start and the more the merrier. It just quick starts your game and the necessity is stronger as you go toward deity.

I disagree that they're best at the start. I think they're best once infrastructure and built cities are in place as you start to ramp up your military and need to pay for it. I argue that you're missing out on too much science, culture, and GP by doing this.

 

[ShinigamiKenji]

Thanks for clarifying my point, @Quillan. I made those posts around 1am, so they might be a little confusing

I think we might draw a rule of thumb after all this. We can simplify it as a function of how many cities you have, how early they're founded, and how many you still plan to found. Running some hypothetical scenarios:

• If you have a decent number of cities (say, 18) and plan to found only a couple more, it might be better to skip CH in favor of your victory district, because you're likely to have enough traders to jumpstart those cities already and support your empire.
• If you only have 7 cities, but don't plan on founding much more, then you might want a CH, because you have only a handful of them to boost production (wonders, spaceship parts etc) and gold.
• If you conquer a city around turn 200, you want to skip CH regardless of number of cities. The city isn't likely to even finish a CH + victory district, let alone repay the CH investment.

As a rule of thumb, maybe we can agree that a CH in a satellite city that's finished before turn 100 is worth it, because it repays itself in time and there's enough time to reap the benefits. If that's not the case, then a CH's still worth it if the trade route per expected final number of ciies ratio is less than 90%, or you have less than 10 trade routes. Any CH past turn 150 won't contribute as much, and should only be chosen if you built everything needed for victory at that point.

Those numbers seem fine for a fast victory. Well, unless you're someone like @whacker, who can consistently get record-breaking fast victories according to his reports in the Science Victory and GotM threads (and whose opinion in this thread would be greatly appreciated).

What would you think? Do you think there's much more to add?

 

[Chibisuke]

I just did a bit more experimentation. As I mentioned earlier, I used to not focus on commercial districts and instead did primarily campus/industrial and then did just enough commercial to keep me in the black money wise. Well I played 2 games (one as Germany one as Kongo) on deity and in both games I built maybe 3 campuses at the start and then switched to building commercial in pretty much every city. Both games I am performing more strongly than usual. This is primarily due to 2 things:

(1) I can purchase granaries rather than hard build them. This allows me to grow faster while still building district stuff and settlers.

(2) I can upgrade my military to swordsmen/crossbows faster, which allows me to conquer more quickly, and thus force the AI to give more money sooner in peace deals. Then I could use that money to accelerate other things. So yea, overall I think buidling as many CD as early as possible is a strategy change that I will be adopting.

 

[Victoria]

Great feedback @Chibisuke thanks for testing the opposite and finding it works well. This is an interesting way to go

I disagree that they're best at the start. I think they're best once infrastructure and built cities are in place as you start to ramp up your military and need to pay for it. I argue that you're missing out on too much science, culture, and GP by doing this.

Too much science/culture?

So growth of these tends to not be linear. I suspect a slower start but with the ability to ramp up the gold may get you at the steeper part of the curve faster. Anyone able to prove or disprove.. maybe with a pretty graph? I'll have a think but I know things are so complex its not easy.

In case people are unaware As England up until this threat I would typically have as many harbors and CZ as I could have and they seem to work. My issue was, If I have 10 cities and 20 Trade routes and good money making cards I start going crazy with what I spend it on.

 

[WackenOpenAir]

So growth of these tends to not be linear. I suspect a slower start but with the ability to ramp up the gold may get you at the steeper part of the curve faster. Anyone able to prove or disprove.. maybe with a pretty graph? I'll have a think but I know things are so complex its not easy.

Indeed, that is the point. But i also think too complex to try and turn it into a simple graph.
There is so much that goes into this.

What are all the factors that get you on that steep curve ?
A-Obtaining many cities with campusses in each of those.
B-Envoys with scientific CSes.
C-Getting buildings in all those campusses.
D-Great persons who provide envoys (fugger, astor) or affect your campus buildings (hypathia, einstein)
First of all, since size 4 is easilly reached, i think we can pretty much disregard the growing cost for districts. You can place them both asap and then build them later when you decide to.

I have done a lot of analysing on this just now but its too incoherent for publication.
Some thoughts:

When you get to make this choise, you dont yet have the envoys to get much gold or science out of the bare districts. The commercial district really is about the trader at this point. The campus provides little science, but the great people points matter too.

Traders can greatly help your poor production towns get the science buildings faster. Using the traders on these weaker cities will provide you a lot more turns on those buildings than they help a strong production town. Therefore building CD+trader first in your strong production towns is an obvious choise.

As you get further in the game, you at some point reach a point where most of your cities already have both districts as well as all available buildings for the science districts. You want to divert trade routes to stop benefitting those an instead benefit the newest/weakest towns. Getting a commercial district in these newest towns is no longer going to get you any campusses and libraries faster. You are now also getting into more envoys. These cities should get campus right away.

Somewhere in between these extremes is going to be the turning point.... (yep, still nothing new...)

While you first build commercial districts, i think you should be spending envoys in scientific CSes first because this will help you get to that steep curve faster.

 

[UWHabs]

A lot also depends on which city states are around, did you get to them first, etc... I remember I had a game (that i quit because I wasn't interested), but I met 3 scientific city states within the first ~10 turns, and was first to each, and I was researching at an insane rate. I mean, that's probably better science production than building a campus. Similarly, if you can get some free culture from an early city-state, or happen upon a relic early on, those can make a huge difference.

I think the fact that policy cards are set up to maximize certain strategies also pushes you in certain directions. For example, if you build a commerce district everywhere, and are running the bonus policies for both trade routes and double CD yields, that's a huge bonus. But if you don't, and you build districts evenly, then none of those cards are overly valuable (in those cases, you'd probably be better to run the cards that give bonus amenities, housing, builder charges, etc...). Plus, as mentioned, the fastest way to get new cities up and running is to either run a trader from them or buy all the base buildings (or even rushing a builder to chop or at least lay down some mines), and it means that if you're not running with a cash surplus, it's harder to establish new cities. And then as well, on the higher levels, it can actually be a huge fight for great people. For example, in my Japan test game I ran where I built no commerce districts, I was doing well with getting engineers, as I had IZ in every city. But other than them, I think I only got maybe 2 scientists the whole game. I basically build roughly evenly of all the other districts, so that I had 3-4 each of harbors, theatres, campuses, holy sites, etc... and because of that, I kept getting out-classes for great people. But because I build the industrial zones in every city, I basically could pick and choose whichever of them I wanted, since I had enough points that by the time I got one, the cost of the next one jumps up enough that I could catch up by the time it came near.

Essentially, there's a lot going in the game that encourages you to really focus heavily on one district type. And since almost every other district happens to be very focused on one task, it just makes sense to focus on the one that can be used to aid every victory type.