City Development

[Thalassicus]

It's just a matter of perspective since it's mathematically identical concepts, but I look at things in their final benefit. Increasing the workshop from 1.25 to 1.5 does not double a city's production capability, so thinking of it as a 100% increase feels odd to me. In addition, even assuming a 3:2 / ratio the workshop is still not recommended until 15+ base production.

The way I look at it is... since workshops are only useful in vanilla if you have the specialist policies (renaissance) or a city of 20+ base production (typically industrial) there's no reason to have the building available in the early medieval. It seems the developers intended it to be useful at that time period, since they put the building there. Changing the production modifier is simpler than a complex combination of cost, maintenance, and high-tier building adjustments.

From the playtesting I've done the 20% increase felt good. I could try a 12% increase though if from your testing you feel it's a bit too high (1.25 -> 1.4). Please keep in mind there's a limit to decision-making capability from theorycrafting with numbers alone... it's important to play through at least a few full games to get a sense of how changes really have an effect.

On another note, the representation policy error should be fixed in the latest version. I'd made the change in the modbuddy project but forgot to build it when making the zip.

 

[Txurce]

The way I look at it is... since workshops are only useful in vanilla if you have the specialist policies (renaissance) or a city of 20+ base production (typically industrial) there's no reason to have the building available in the early medieval.

This thought came to mind while reading Tomice's last post. There is something wrong with vanilla workshops now, just as there was with the other buildings Thal buffed. What if the rationale is reversed, and the majorly buffed workshop has a higher maintenance fee that only wealthy civs or large medieval cities can afford, whereas the factory is a more efficient building capable of transforming an entire society? That could still keep the workshop a choice, as opposed to a no-brainer.

 

[Ahriman]

@Tomice
I agree that this definitely comes down to the philosophical question.

And I agree that AI performance is important, but I think the AI should be able to do an ok job of yield/boost:cost calculations (though if the benefit is too high, then it might focus on infrastructure too much and neglect military). I think whatever % yield value is chosen, it should be easy enough for players to figure out how/when to use it, I don't see that as a big problem.

But I think that building changes in general are about a much more important topic; the value of lots of cities vs few cities.
In the vanilla game, one of the biggest problems is that you are generally better off to spam lots and lots of cities, that more cities generally make you better off, because science and trade income comes from population, and because it takes much less food to get smaller cities than larger cities.
Happiness doesn't offset this because of the linear cost of city unhappiness, and the cost efficiency of the Colosseum.

So, how then to encourage fewer, larger, more-spaced out cities?
The best mechanic for doing this is through buildings that have multiplier yields. Buildings with a fixed cost and a % yield (or population-based yield) give a much bigger benefit in large cities than small cities, and if the construction/maintenance costs are high enough, then they aren't worth building in small cities.
Since its easier to get two size 6 cities than one size 12 city, that size 12 city needs to be *much* more efficient than two size 6 cities.

Buildings like the university, public school, workshop, factory, etc. are great ways of encouraging highly developed cities.
Ideally what we want I think is building that is worth building in the size 12 city, but is not worth (or barely worth) building in the size 6 city.
This way, there are some clear advantages to focusing on larger cities; you get good returns from the infrastructure buildings.

[Another possibility; if the goal of the workshop is to encourage building construction in larger cities that are not necessarily production oriented in terms of tile yields, then a +1 hammer per 2 population when working on buildings might be more effective than a +hammer% yield.]

* * *
On a similar but tangential note: one of the other really annoying things with ICS is how it combined with the tougher city defenses.
Is it possible to tweak the city strength calculation so that it depends slightly more on city size, and slightly less on tech?
This would be another way to tweak towards larger cities rather than smaller cities.
* * *
edit: Ok, I will try to test. 1.4 feels more reasonable.
My point on the % yields is that comparing the percentage increase over vanilla isn't the right way to go, we care about the total value of the building, not the value of the building relative to vanilla.
If you don't like thinking about 1.2->1.4 as a 100% increase, try just thinking of it as having a 40% production increase from the building.
Otherwise, you can get yourself in mental traps.
For example, if the vanilla building only gave a 5% bonus, then increasing to 1.4 would be a 33% increase over vanilla, but if the vanilla building gave a 10% bonus then increasing to 1.4 would be a 27% increase over vanilla. And yet the final value is the same.
A low vanilla value just means it was *really* underpowered, it doesn't necessarily change what the appropriate value should be.

 

[Seek]

This thought came to mind while reading Tomice's last post. There is something wrong with vanilla workshops now, just as there was with the other buildings Thal buffed. What if the rationale is reversed, and the majorly buffed workshop has a higher maintenance fee that only wealthy civs or large medieval cities can afford, whereas the factory is a more efficient building capable of transforming an entire society? That could still keep the workshop a choice, as opposed to a no-brainer.

That's a very good point.

 

[Ahriman]

Worth considering.
Maybe a large bonus, like 40%, but a higher maintenance cost?
But this makes me worry about the AI performance; will the AI build too many and destroy itself with maintenance costs?
Will the human player be able to build the structure, use it to build the structures they want, and then sell it off to avoid the maintenance cost, while the AI is unable to sell?
I don't think the AI ever sells buildings?

 

[Txurce]

On a similar but tangential note: one of the other really annoying things with ICS is how it combined with the tougher city defenses.
Is it possible to tweak the city strength calculation so that it depends slightly more on city size, and slightly less on tech?
This would be another way to tweak towards larger cities rather than smaller cities.

I've only encountered the annoying aspect of this once, but it drove me nuts. It seemed as if classical-age Alexander was waging war on me not just with units but with settlers, marching them up to my borders with a hoplite, then settling. Each one was a pretty tough nut to crack, and I didn't have the numbers to take them all on. Again, I've only encountered this once. But I haven't forgotten!

 

[Txurce]

Worth considering.
Maybe a large bonus, like 40%, but a higher maintenance cost?
But this makes me worry about the AI performance; will the AI build too many and destroy itself with maintenance costs?
Will the human player be able to build the structure, use it to build the structures they want, and then sell it off to avoid the maintenance cost, while the AI is unable to sell?
I don't think the AI ever sells buildings?

Yes, something like this.

I don't think we should worry about the advantages of selling off a building vs whether the AI does it. This is a sophisticated strategy that most players won't use, either.

 

[Ahriman]

An interesting test might be: create a building that costs 10 hammers, gives +100% hammer output but has a maintenance cost of 40 gold per turn.

Does the AI ever build it?

If so, then the AI isn't doing a good job of understanding maintenance costs.

Though for small tweaks, we're probably still ok.

 

[Tomice]

... But I think that building changes in general are about a much more important topic; the value of lots of cities vs few cities.
...

An important point and a good post overall!

Another possibility; if the goal of the workshop is to encourage building construction in larger cities that are not necessarily production oriented in terms of tile yields, then a +1 hammer per 2 population when working on buildings might be more effective than a +hammer% yield.

* * *
On a similar but tangential note: one of the other really annoying things with ICS is how it combined with the tougher city defenses.
Is it possible to tweak the city strength calculation so that it depends slightly more on city size, and slightly less on tech?
This would be another way to tweak towards larger cities rather than smaller cities.

The second point is very true IMO, size 12 cities should be MUCH tougher than size 3 villages.

The first point sounds very interesting, but has a very risky aspect: There are some ways in Civ5 to get production into cities where the terrain doesn't provide them, especially from policies (commerce and order tree), but your suggestion would make food-heavy, production-poor locations MUCH stronger in the early game. Science and commerce cities don't produce much else than buildings, and those two types represent maybe 75% of my cities.

Maybe this change would be beneficial, maybe we should change some of the other production-boosting buildings to this type, too (military?). Maybe this would improve the value of farms compared to mines/lumber mills/trading posts. Admittetedly, the blurry definition of production (both representing materials and manpower!) bugs me big time since years.
But this change would fundamentally alter the gameplay!

 

[Ahriman]

But this change would fundamentally alter the gameplay!

Oh I agree, I'm not sure at all that it would be the right thing to do, just throwing it out there as a possibility to consider.
Alternatively; rejig the factory using 1 hammer per X population mechanism. That might be even more effective and realistic; factories associate with big urbanization, which represents something of a liberation from land-based yields with raw materials that can be imported from a long way away, whereas early game industry is much more dependent on local resources.

Still, big changes like that would be very risky indeed, but they are a great solution to the Civ4 problem, where successive % modifiers have declining impacts because they all just multiply the base yield. That 5th +25% modifier only increases yields by 1/8 (2.0 to 2.25). I really like how they did this with the library.

 

[Ahriman]

I had another thought on city defenses.

In vanilla-pre-patch, cities were too weak, and too easy to capture, so the patch bumped up their strength, among other things.

Post-patch they bumped up the city strength, it feels like its good that cities are harder/slower to take, but that they deal out a ton of damage, particularly under an ICS framework where its hard to attack without being in range from 2 cities.
This is problematic, because its not very fun for the human, but it also means that the AI can't really threaten the human, because its too easy for the human to use the powerful city bombardment to destroy the armies the AI sends at you, which tend not to be large enough to take the city.

So, maybe the right solution is to leave the higher defense values (though maybe tweak the pop/tech dependence) but to reduce the ranged attack multiplier? From memory, the ranged attack strength of a city is something like 40% or 60% of its combat value; what if we find that parameter and tweak it down slightly?
This will make cities just as resilient as they were before, but less wtfpwnage in terms of their ability to destroy the invading army by themselves.

 

[Seek]

@Ahriman - I like your first idea better (make city strength be more weighted toward population), but a slight decrease for attack strength may be worth looking into as well. In my experience the massive increase in city's healing rates make the greatest difference, but I'm not sure how to tweak this - maybe limit healing by era (3 in ancient/medieval, 4 in renaissance, 5 after?). - Or better yet, number of techs since there are already so many bonuses reliant on era.

Definite +1 for factory getting hammers per pop! Another thought; what about making Production modifiers multiplicative rather than additive? It would provide an indirect buff for the Workshop and military production buildings, and in conjunction with the % boost for the Factory removed, may work well. (Obviously some serious testing would be required to ensure balance.)

BTW, good to have you here Ahriman, followed the Core Balance Changes thread and find your ideas consistently insightful.

 

[Ahriman]

The city defense ideas aren't mutually exclusive.

Healing is already effectively limited by era, since it depends heavily on the strength provided by the defensive buildings in the city, it is heavily tied to walls, castle and military base structures.

I'd oppose making % yield changes multiplicative rather than additive, even if that were possible.
That would be very confusing for the player to calculate, and even more confusing
This in my opinion is one of the big advantages of the whole library yield-per-pop mechanic; it does make modifiers multiplicative, but does so in an easy to interpret manner, and makes it clear which modifiers add which which.

For example:
Size 12 city, with library produces 18 science. With university and no library, produces 18 science. With 2 buildings that both just gave +50% science, would produce 24.
But with university + library, it will produce 27, because the library increases base yields, which happens before % modifiers.

What would be even nicer about making factories give hammers per pop, is that this would then stack multiplicatively with the railroad bonus. It'd be great to see synergies between railroad and factories, where you got the largest boost from making sure that your cities with factories were the ones that had railroad connection.

Thanks for the kind words. I'm a big fan of what Thal is trying to do here, and I like doing design work (check out Dune Wars, we just released a new version) but I am incompetent on coding.

 

[Seek]

The city defense ideas aren't mutually exclusive.

Healing is already effectively limited by era, since it depends heavily on the strength provided by the defensive buildings in the city, it is heavily tied to walls, castle and military base structures.

OK, I just ran a quick test in world builder: Medieval start, Emperor handicaps for both sides (don't know if that makes a difference). I gave the enemy (Liz) a palace, walls and castle with a treb in the city. I had three trebs with one promotion and three swords. Liz's city with 12 pop had a defense value of 37 and healed 5pt! By turn 3, one of my trebs was dead, making it impossible to damage the city more than 5 in a turn until I got the siege promotion, and then only by about 1 hp per turn. The city's damage was on average 1 against both my trebs and swords.

Since in my experience with their capitol being threatened the AI will build defensive buildings, this makes for a very difficult situation for the player let alone another AI. Thus I would argue that the city healing is somewhat OP now (or alternatively that siege is somewhat underpowered). I do realize that they aren't mutually exclusive of course, it just seems to me that city offense (which is 50% according to Alpaca) is less of an issue.

I'd oppose making % yield changes multiplicative rather than additive, even if that were possible.
That would be very confusing for the player to calculate, and even more confusing
This in my opinion is one of the big advantages of the whole library yield-per-pop mechanic; it does make modifiers multiplicative, but does so in an easy to interpret manner, and makes it clear which modifiers add which which.

For example:
Size 12 city, with library produces 18 science. With university and no library, produces 18 science. With 2 buildings that both just gave +50% science, would produce 24.
But with university + library, it will produce 27, because the library increases base yields, which happens before % modifiers.

What would be even nicer about making factories give hammers per pop, is that this would then stack multiplicatively with the railroad bonus. It'd be great to see synergies between railroad and factories, where you got the largest boost from making sure that your cities with factories were the ones that had railroad connection.

That's a good point, I hadn't thought of it like that.

 

[Ahriman]

Here is the formula for city defense and healing, AFAIK:

http://forums.civfanatics.com/showpost.php?p=10029376&postcount=14

It depends on city defenses and whether or not the city is a capital.

 

[Tomice]

@ Seek:

Siege units ARE buggy as of v10141, check out this thread:
http://forums.civfanatics.com/showpost.php?p=10036001&postcount=1

@ Ahriman:

I'm a big fan of what Thal is trying to do here, and I like doing design work (check out Dune Wars, we just released a new version) but I am incompetent on coding.

Exactly the same is true for me, I am only able to do minor XML changes, but I can't do much else, I'm from a non-computer profession



About the topic:
It's irrelevant if we put the "production from population" ability to workshops or factories. We have to deal with the new mechanic and its possible problems anyway.
I'm all for discussing this further, but we should be careful that city placement doesn't become irrelevant. Not that I think pulling magical "production" out of every hill or turning forest yields into tanks is a good concept we definitively have to keep

 

[Seek]

It depends on city defenses and whether or not the city is a capital.

And it's hardcoded. Oh well.

 

[Ahriman]

Siege units ARE buggy as of v10141, check out this thread:

I'm not convinced it was a bug, I think it may have been a deliberate nerf. Siege was very powerful.
One option is to keep siege as good but not insane, but to remove its iron requirement.
Another is to buff its bonus vs cities.
I'd certainly like to see an artillery nerf vs units in the open.

I'm all for discussing this further, but we should be careful that city placement doesn't become irrelevant.

Agreed. Its a fine line to walk. We do want major production cities to need to be using mines and lumbermills, so that different terrain favors a production city vs a

I suspect the best thing for the Workshop is tuning on the % bonus. I'd say ~35% feels roughly right to me, along with the engineer slot.

 

[Seek]

I'm not convinced it was a bug, I think it may have been a deliberate nerf. Siege was very powerful.
One option is to keep siege as good but not insane, but to remove its iron requirement.
Another is to buff its bonus vs cities.
I'd certainly like to see an artillery nerf vs units in the open.

This and this and this.


@Tomice: I don't think it's irrelevant which building to put a per pop bonus in, it makes a huge difference how soon it can be gotten over the course of the game. Also, your concerns about city placement would have less merit if the factory got it, as your core cities are already built: In most games I rarely get >10 coal, so I'm only building the factory in my most populous core cities, which not incidentally have the strongest production, anyway.

 

[Ahriman]

[By units in the open, I didn't just mean open terrain, I meant units, as opposed to cities. Arty needs a ranged attack nerf. It should be a support weapon, not for blowing away infantry units.

I think what Tomice was saying is that you have to think carefully about whether +hammers per pop is a good idea at all before thinking about which building should get it.
In terms of which building, I think the factory makes far more sense, both from a thematic point of view and a coding point of view (its probably easy to code +hammer per population, its probably harder and more confusing to the user to code +hammer per population-while-constructing-buildings).

But I think Tomice's core point is; most yield boosters are designed to encourage specialization. Gold cities have lots of gold, rivers, and trading posts.
Science cities have lots of food and population.
Production cities have lots of hammers, and use lots of hill and forest and plains tiles.
I think his point is; do we really want a city with large population via lots of food to still have large production? Because doing so reduces the value of mine yiles and boosts food and population supporting things (including Tradition tree, Maritime city states, granary, etc.). And these balance mods already massively boost city growth relative to vanilla, through food boosts (slaughterhouse, harbor, fresh-water-farms) and through growth boosts (aqueduct).

I think a Factory might be late-enough to do this, and a good way of representing industrialization.
But is it fixing a problem that doesn't really exist? The factory seems pretty well balanced as is.