[C3C] AI "Build Often" building categories

[md4]

Apologies if this has already been discussed: I did spend quite a bit of time searching old threads and pinned guides to no avail.

How does one align the "Build Often" categories of [Growth, Production, Happiness, Science, Wealth, Trade, Explore, Culture] to the actual improvements themselves? There's not a clear 1-1 mapping between the Characteristics (e.g., Militaristic, Religious, Industrious, etc.), so is it just inferred from whatever the Improvement generates (Culture, Production), or options that are selected (such as Increases Luxury Trade, +50% Research Output, etc.)?

Also, what would be the difference between the categories Wealth and Trade? Wouldn't they almost completely overlap?

 

[tjs282]

Pretty sure the AI Build-often flags map primarily to the various Wonder/ Improvement properties such as "Adds Culture", "Makes Citizens happy" and so on.

Also, what would be the difference between the categories Wealth and Trade? Wouldn't they almost completely overlap?

In this context, I believe "Wealth" is simply the build-option which converts shields to gold at a 4:1 ratio (and 2:1 after learning Economics).

Not sure what "Trade" means, but I'd assume it applies to improvements that have the "+50%-Tax" flag, such as Markets, etc. -- or possibly specifically Harbours/ Airports, which "Allow Trading [of resources]".

 

[md4]

Harbours and Airports are the only improvements I'm aware of in vanilla Civ3 that enable trade routes, so I'd be pretty confident that "Trade" buildings would include these. Possibly Marketplace too, since it has the "Increases Luxury Trade" attribute. Wealth may refer to shield conversion to gold, but what about buildings like Bank & Stock Exchange?

I can make (semi) educated guesses, I was hoping somebody had proof of some description.

 

[md4]

For that matter, what would "Explore" cover in "Build Often"? Explorer units? Anything with the Expansionist characteristic?

 

[Predator145]

Growth: Granary, Aqueduct, Hospital (improvements that makes cities grow faster or allow a new size)
Production: Factory, Powerplants (Shield production bonus multiplier)
Happiness: Improvements that give happy faces.
Science: Improvements that give a beaker output multiplier
Wealth: Improvements that convert shields into gold income (Wealth in stock game)
Trade: Improvements that give a gold income multiplier (market, bank, stock exchange), improvements that connect trade routes (harbor, airport)
Culture: Improvements that give produce culture per turn
Explore: Units that are flagged with the Explore AI strat flag.
Military: Improvements that makes the city train veterans, allows upgrades, improvements that increase a city's population/improvement defense value as well as that of defending units, improvements that have anti air and coastal zone of control value.

These categories overlap. If you want a civ to max prioritize building libraries you'd tick : Science and Culture because libraries have both the science output multiplier and a per turn cultural output.

Unfortunately there is no Military category to encourage the AI to build barracks/harbors/walls. You can encourage it to build offensive and defensive strategy units. So C3X perfume is the only way.

 

[md4]

Thank-you for the extremely helpful answer. I was surprised that Wealth only covers the shield-to-gold project. Seems kind of useless, but probably means that I can uncheck that from any Civs which have it, and use it for something better.

Unfortunately there is no Military category to encourage the AI to build barracks/harbors/walls. You can encourage it to build offensive and defensive strategy units. So C3X perfume is the only way.

You can't do it with Flavors?

 

[Predator145]

Thank-you for the extremely helpful answer. I was surprised that Wealth only covers the shield-to-gold project. Seems kind of useless, but probably means that I can uncheck that from any Civs which have it, and use it for something better.

You can't do it with Flavors?

Yeah, I forgot about flavors. That works too. But not as much as perfuming. But the good thing is that you can have different civs prioritizing different things with it.

Still, even with flavors, things like barracks, walls and harbors aren't built in time. But perfuming is tricky. Too much and you'll ruin their expansion phase. From what I've seen, the AI can afford to have highly perfumed walls. They are stuck waiting for their pop to catch up to the settler production anyway.

But barracks I still can't figure out what to do. A moderate perfume of 40-100 means they'd build it at an OK timing during their late expansion phase or afterwards. But once that stage is over and they're fighting wars, they'd need to build it ASAP. Because once the AI is at war, it will vastly prioritize offensive units. A perfume value of 100 will be useless. You'd need something like 10 000. I hate seeing the AI make regular knights from their size 12 cities.

The solution could be another heavily perfumed barracks-like improvement that appears with a late ancient or early medieval tech. Or you could have ancient barracks go obsolete in the middle ages to be replaced with a new type of heavily perfumed barracks. Or you could have an AI exclusive barracks that maybe gives 1 happy face on top of being a barracks to justify the AI building both the regular and AI exclusive version. This inelegant solution is what I'm using right now. The AI could use more happiness improvements anyway.

To have an an AI exclusive improvement use C3X to make it require another improvement to be built. I call this improvement "Flag AI". "Flag AI" is not searchable. and can only be generated in all cities via a heavily perfumed ( 100000 value) 0 shield small wonder called "AI SW" which requires the palace. The AI will build this SW ASAP. The human player will just have to pay attention not to build it in their capital. This way I minimize seeing anything AI exclusive.

 

[Fergei]

This is very interesting and I am grateful for the thread and posts. I had never noticed the 'build often' applied to categories like growth (even though it is as clear as day!). I also had not used flavours. It helps explain my sense that the different AIs are quite distinct, on average, beyond simply their traits. For example, I buffed Rome by giving it additional traits, but its behaviour remains pretty much the same and I struggle to get them to actually pounce when they have legionnaries. Similarly, I switched trait combinations between some Civs and behavioural changes were subtler than I expected.

My changes will be minimal. I like the AI so much I'd be concerned I'd mess up something I already enjoy. I also would not want the non-militaristic AI to produce too many barracks (I think that would nerf the militarism trait a bit given the AI can't use armies). But the following are the Civs that routinely underperform with me, so I will try to give them a little assistance and help them be competitive in the ancient era.

India - (flavour to increase uptake of walls)
Ottomans - (flavour to increase uptake of barracks)
America - (flavour to increase uptake of walls)
Mongols - (flavour to increase uptake of barracks)
Rome - (flavour to increase uptake of barracks)
Aztecs - (flavour to increase uptake of walls & barracks)

 

[Fergei]

Instant results that are to my liking, thanks again for this thread. I also now understand why China is so good late on in my games, they focus on science despite not having that trait. They always mystified me.

 

[C0rTeZ48]

Growth: Granary, Aqueduct, Hospital (improvements that makes cities grow faster or allow a new size)
Production: Factory, Powerplants (Shield production bonus multiplier)
Happiness: Improvements that give happy faces.
Science: Improvements that give a beaker output multiplier
Wealth: Improvements that convert shields into gold income (Wealth in stock game)
Trade: Improvements that give a gold income multiplier (market, bank, stock exchange), improvements that connect trade routes (harbor, airport)
Culture: Improvements that give produce culture per turn
Explore: Units that are flagged with the Explore AI strat flag.
Military: Improvements that makes the city train veterans, allows upgrades, improvements that increase a city's population/improvement defense value as well as that of defending units, improvements that have anti air and coastal zone of control value.

These categories overlap. If you want a civ to max prioritize building libraries you'd tick : Science and Culture because libraries have both the science output multiplier and a per turn cultural output.

Unfortunately there is no Military category to encourage the AI to build barracks/harbors/walls. You can encourage it to build offensive and defensive strategy units. So C3X perfume is the only way.

I think I've read multiple times on CivFanatics that checking more than 2 'Build Often' markers will end up not having any effect, since the AI won't behave differently as a result (building more of that flag type). Or was that only for units?

What I intend to ask is: could one afford both one building 'Build Often' flag, as well as one unit 'Build Often' flag? Or will even that end up making the AI not prioritize certain units/buildings over other units/buildings?

Also, I'm mostly interested in the higher difficulties, where the AI will already build pretty much everything in all their cities at a fast pace. So I guess only one 'Build Often' unit would end up being most effective? Am interested in what others have to say on this. I definitely think the AI should prioritize workers more often.

Also, does the 'Never Build' really makes the AI truly never build that unit/building? Meaning that in that case, checking multiple flags will actually make the AI behave in never producing those units/buildings? Wonder if anyone has experimented with this extensively.

Still, even with flavors, things like barracks, walls and harbors aren't built in time. But perfuming is tricky. Too much and you'll ruin their expansion phase. From what I've seen, the AI can afford to have highly perfumed walls. They are stuck waiting for their pop to catch up to the settler production anyway.

But barracks I still can't figure out what to do. A moderate perfume of 40-100 means they'd build it at an OK timing during their late expansion phase or afterwards. But once that stage is over and they're fighting wars, they'd need to build it ASAP. Because once the AI is at war, it will vastly prioritize offensive units. A perfume value of 100 will be useless. You'd need something like 10 000. I hate seeing the AI make regular knights from their size 12 cities.

I am astonished that this has been your experience with AI. In stock Conquests games, even in the Ancient Era pretty much all AI units are already veterans, rather than regulars. But then again, I also only play Demigod+. Perhaps you play often on lower difficulties? Or are you playing particular mods where you've added tons of new buildings? Perhaps that may be part of the problem that causes this issue for you.

 

[Predator145]

I think I've read multiple times on CivFanatics that checking more than 2 'Build Often' markers will end up not having any effect, since the AI won't behave differently as a result (building more of that flag type). Or was that only for units?

What I intend to ask is: could one afford both one building 'Build Often' flag, as well as one unit 'Build Often' flag? Or will even that end up making the AI not prioritize certain units/buildings over other units/buildings?

Also, I'm mostly interested in the higher difficulties, where the AI will already build pretty much everything in all their cities at a fast pace. So I guess only one 'Build Often' unit would end up being most effective? Am interested in what others have to say on this. I definitely think the AI should prioritize workers more often.

Also, does the 'Never Build' really makes the AI truly never build that unit/building? Meaning that in that case, checking multiple flags will actually make the AI behave in never producing those units/buildings? Wonder if anyone has experimented with this extensively.

I am astonished that this has been your experience with AI. In stock Conquests games, even in the Ancient Era pretty much all AI units are already veterans, rather than regulars. But then again, I also only play Demigod+. Perhaps you play often on lower difficulties? Or are you playing particular mods where you've added tons of new buildings? Perhaps that may be part of the problem that causes this issue for you.

I play on Deity pangaea with whatever start is given and no common exploitations. And no, the AI doesn't even build barracks until they're done expanding with the regular settings. I've played enough debug on both stock game and mods to know that. We all have seen plenty of standards knights being built on Deity. Once war starts, they will prioritize units first and will never pause just to make the rax. Maybe the AI raxes up fast enough for your taste, but not mine.

Stock game has up to 5 build often flags (Byz). And Alexmann who worked with Firaxis has confirmed many years ago that 6 is the max effective number. His source is the most credible.

Things like Growth and Workers don't need the editor. C3X allows you take care of that easily. The editor is only there for subtle nudging. So you can just tick Happiness and be done with it if you don't feel like ticking multiple boxes.

I've played extensive test games on having up to 5 boxes ticked and the AI build the things I want them to build just fine.

 

[C0rTeZ48]

I play on Deity pangaea with whatever start is given and no common exploitations. And no, the AI doesn't even build barracks until they're done expanding with the regular settings. I've played enough debug on both stock game and mods to know that. We all have seen plenty of standards knights being built on Deity. Once war starts, they will prioritize units first and will never pause just to make the rax. Maybe the AI raxes up fast enough for your taste, but not mine.

I just checked a rather recent Sid game and it does seem the AI had quite a lot of regular units in the Ancient Era, which is more than I remembered, However, in the Middle Ages pretty much every single unit did seem to be veteran. Also for another Deity game I played, first being on a Large map, the Deity on a Standard one. These were both unmodded games.

 

[Predator145]

I just checked a rather recent Sid game and it does seem the AI had a quite a lot of regular units in the Ancient Era, which is more than I remembered, However, in the Middle Ages pretty much every single unit did seem to be veteran. Also for another Deity game I played, first being on a Large map, the Deity on a Standard one. These were both unmodded games.

The thing is, even when built as horsemen, these knights could have been built as veterans. Why? Because when the AI expands, it doesn't work the tiles optimally, and so there is a lot of waiting time for population to catch up to get out the settlers. They could have easily squeezed in a barracks in the meanwhile, esp if they're militaristic.

 

[Predator145]

I just discovered that "Capitalization" also includes anything that increases commerce output. When I had it ticked as "never build", the C3X AI priority ranking had Marketplaces at a huge negative and no amount of C3X during game perfuming would change that. So I'd say Capitalization means: Wealth, Marketplaces, Banks, Stock Exchanges, Commercial Docks.