[C3C] Early game AI passivity

[Fergei]

As part of my quest to make game settings that match my preferences I have now come towards targeting the issue of the AI generally not being interested in conflict until all easily available land has been grabbed by all the Civs (at least on Emperor, which is my preferred difficulty).

Crucially, I do not want to correct this by making the AI more aggressive in the middle and late game (I'm very happy with that and consider the unpredictability one of the games strong points).

The problem with this issue is that the AIs that start with very early UUs (Aztecs, Egypt etc) rarely use these units until they are competing against superior standard units (e.g. horsemen, swordsmen, spearmen). I also never see AI combat involving chariots vs warriors etc. I don't want this early game fighting to happen all the time, but for it to happen once in a while would be nice.

The only option I can think of is to make the first two columns of techs take far longer to discover so that all the available land is taken up before certain techs are discovered. But I have already done that with Ironworking (which I consider OP given how early it can appear). I could make Bronzeworking, Iron Working and Horseback Riding more dependent on achieving other techs first, but this also doesn't seem to be an elegant solution (and I've already done some of that (e.g. IW requires pottery for kilns) and it didnt resolve this issue).

Firstly, am I right to flat this up as an 'issue'?
Secondly, can anyone think of any alternative solutions?

 

[justanick]

Secondly, can anyone think of any alternative solutions?

First make sure that there are no bonus huts from which you can get techs from. Second make the ancient techs non-tradeable. Those 2 together will slow down early tech advancement by a significant margin.

Still once the settling phase is over and towns grow and possibly grow to city size, 4 turns per tech will become reasonably achieveable and the ancient age will be over reasonbly soon, at least for the more fortunate civs.

The problem with this issue is that the AIs that start with very early UUs (Aztecs, Egypt etc) rarely use these units until they are competing against superior standard units (e.g. horsemen, swordsmen, spearmen).

I believe this to be acceptable. If it is pikes, the issue is different.

Increasing the amount of civs per map size is another option. If there is less territory to grab without war, then wars will occur earlier. Together the 2 measure above this could mean that weaker civs may get stuck in the ancient age with little hope to make up for it. They would be failed nations and due to their relative military inferiority(Spearmen vs. MedInfs) they are likely to be eaten up by the stronger nations.

You could make horseback riding(only 5 base points) more expensive so that Egypt and the Hitties are less tempted to not use their UU.

 

[Civinator]

... and you can set up all ancient units as HN-units. The AI considers them as Barbarians and with Barbarians all civs are always at war.

 

[need my speed]

Personally, I have made all Ancient (and sometimes Medieval) technologies untradeable, and I have increased the minimum amount of turns a technology takes from 4 to 6. I have played with 8 multiple times over the years, but this becomes worse in the Medieval Age, where I find myself quickly outpacing the AIs - but 6 works quite nicely! I also experimented in my last game with putting almost all AIs on one continent, making it quite cramped, and leaving only a few AIs to settle the other continent, allowing them to have comparatively large empires to pose a threat for me later in the game.

Technologies taking longer to research means there's less to build means that land is filled up slightly earlier means that the AI will start amassing units. So far I have always been declared war on at least once, facing sizeable stacks of Warriors and Archers. As for Horseback Riding, perhaps simply make it less likely to be researched, as it remains a potentially very early technology. I'm not if the AI values the cost of a technology in deciding what to research, though? You could play around with the Flavour of the technology, I think the AI does take that into account.

 

[tjs282]

As part of my quest to make game settings that match my preferences I have now come towards targeting the issue of the AI generally not being interested in conflict until all easily available land has been grabbed by all the Civs (at least on Emperor, which is my preferred difficulty).

If you want wars to happen sooner, you could also make more terrain-types non-Settle-able. With fewer 'viable' (i.e. CxxxxC, for the AI) founding spots available, the AIs should switch earlier from building Settlers to building attack-units.

The problem with this issue is that the AIs that start with very early UUs (Aztecs, Egypt etc) rarely use these units until they are competing against superior standard units (e.g. horsemen, swordsmen, spearmen). I also never see AI combat involving chariots vs warriors etc. I don't want this early game fighting to happen all the time, but for it to happen once in a while would be nice.

Cleo and Mursi do build and use their Chariot-UUs (if given the opportunity), but I've also rarely if ever seen anyone else build and use the basic 1.1.2 Chariot. I assume that this is because Warriors are half (and Jags are 3/4) the shield-price for the 'same' A/D stats, and Archers have A=2 for the same shield-cost, and none of them have any terrain-restrictions, so are all 'better' units from the AI's PoV.

If so, the only way to get the AI to build (epic-game) Chariots rather than Warriors or Archers would likely be to make Chariots more powerful -- but if you buffed them up to 2.1.2, then the War Chariot would also need a buff, or its only 'advantage' would be that it started a GA.

 

[vmxa]

Easy solution and you do not have to mod anything, just play Always War. Use max civs for the map size, you will see plenty of Unique Units. My favorite, not too hard for Emp, can be a real bugger at DG.

 

[Fergei]

Extremely useful feedback, thanks. I hadn't noticed I can make some techs untradeable. I will play about with that first (with an exception for Map Making otherwise I will further penalise the stupid AI if it is alone on an island). I'll also maybe make the standard chariot cheaper to build to encourage uptake.

I'll also maybe make one overcrowded continent per custom map, to encourage one location where there is more likely to be a dog fight.

And yes, I should have mentioned Ming the Murciless of the Hittites as an exception. He is so mad he would start a war in an empty room.

If the above doesn't work I'll experiment with some of the other options. The customisation of the game continues to astound and if I can resolve this I'll be delighted. It is the only issue in which I think Civ2 is the superior game (i.e. an AI can pose a potential threat from turn 1).

 

[tjs282]

As part of my quest to make game settings that match my preferences I have now come towards targeting the issue of the AI generally not being interested in conflict until all easily available land has been grabbed by all the Civs (at least on Emperor, which is my preferred difficulty).

I usually play at Emperor as well...

It is the only issue in which I think Civ2 is the superior game (i.e. an AI can pose a potential threat from turn 1)

...but at DG, I've been stomped by AI-controlled Warriors as early as Turn 20 (Small Pangaea, 6 Civs: I rolled the Incas and got killed by the Iros, 1 turn before my first Settler completed -- Hiawatha already had 4 towns).

So upping your difficulty level might also be an option...

 

[Fergei]

Ah yes, I could increase difficulty but I prefer the game with a consistent difficulty throughout (cost factor) rather than a big spike in difficulty at the beginning (AI starting units) and a very straightforward industrial period onwards (once you've overcome the spike).

I've tried the the changes I said I would and made all techs in the first two columns of the ancient era untradeable. This alone seems to increase aggression and there were early game fights (admittedly courtesy of Hittites and Vikings) long before all land was taken up. It makes me wonder if the AI likes rivals more if it trades techs with them and, without being able to do that, if the AI then gets more irritable? I was certainly being threatened more than normal too in the early turns. But one game isnt enough of a sample, although I am optimistic.

I kept the goody huts for now and it had a consequence I didnt expect. The combination of a scout with defence 1, no additional starting units for AI (both my preference), no tech trading early on and goody huts massively assists Expansionist Civs early game to get superior tech and the other Civs don't all catch up quickly as is normally the case via trading. I might actually have a game where Expansionist Civs aren't an afterthought by the medieval stage. In terms of my aim to have AI combat involving starting units I probably do have to reduce the number of goody huts though. Thanks again all.

 

[Fergei]

In my next game Japan made the first AI standard chariot I have ever seen. They then promptly declared war on me while my best military unit was still a warrior (me being greedy). It made me lose the game, but mission accomplished! Thanks again for the tips - there is a lot less predictability in the early game now and greed is punished more. Feels more like the intensity of multiplayer.

 

[Puppeteer]

Crazy idea, and I've never tried this: What if you make a late-AA tech a prerequisite for producing settlers and then do something like have the palace produce a settler every 5, 10, or 20 turns that obsoletes when you get the settler tech?

If the AI are unable to make more settlers maybe they'll attack more.

 

[Theov]

I made deserts, tundra and swamps unsettlable. Makes for a far more realistic map, and those tiles are more than youd think.
I like the idea od the early techs to be untradable.
I also made the mongolians to behave like barbarians, so there will always be action with them.
But if they are around, they will hinder development of other players.
Addung units with the enslavement options works too, slaves develop land so yu need less workers/population to do it.

 

[Theov]

Oh and I turned off map trading throughout the game.
It forces you to discover more yourself and takes away the explpit of you trading maps every turn. And it lowers ai on ai trading, which is a good thing. If they are not friends, theres more chance of them fighting, Id guess.

 

[Civinator]

Crazy idea, and I've never tried this: What if you make a late-AA tech a prerequisite for producing settlers and then do something like have the palace produce a settler every 5, 10, or 20 turns that obsoletes when you get the settler tech?

If the AI are unable to make more settlers maybe they'll attack more.

This - and a lot more - is exactly done in the CCM mod (over 1800 downloads in a lttle bit over one year).

 

[Fergei]

Crazy idea, and I've never tried this: What if you make a late-AA tech a prerequisite for producing settlers and then do something like have the palace produce a settler every 5, 10, or 20 turns that obsoletes when you get the settler tech?

If the AI are unable to make more settlers maybe they'll attack more.

My memory is so poor I forgot that restricting settler spawn was recommended to me years ago. I think I imagined such changes would be beyond my ability with the editor, so didn't pursue it at the time.

I made the palace spawn a settler every 20 turns and proper settlers appear at mapmaking (which is greatly delayed by preventing tech trading of early AA techs) and cost a lot more shields and 4x population.

Pros:
- higher chance of AI (and player) aggression as greater focus on producing military units and the AI loves to use units once they have built them (so mission accomplisheď in terms of this thread.
- far better use of workers and hooking of resources by AI
- AI explores maps with Curraghs regularly
- Chariot based UU Civs think about war long before the UU loses its advantage.
- different AI Civs focus on military, city improvement or territory based on when they learn map making, so lots of variety early on with AI priorities
- any human or AI focussing on an early Wonder are not jeopardising their medium term growth anywhere near as much as normal

Cons:
- not being able to build settlers feels weird, as does your capital city growing so much early on
- a freshwater start feels a little overpowered as you can quickly get to 8+ pop as palace spawned settlers use no population
- granaries don't feel so valuable until you can build proper settlers, so city improvements are sort of delayed, so players are kind of forced towards units production until map making comes and you can build settlers

Neutral
- quite a lot of juggling of wealth and managing unit support
- occassionally an AI can be wiped out before having 4x cities. Or can lose their capital early (I increase the number of Civs on the map to try and counter this)
- with default difficulty settings the AI starting units enabling more rapid expansion are a key element of generating difficulty, so you will either have to tweak the difficulty settings or go up a difficulty level since AI expansion is curtailed.

I would argue restricting settler output is the single greatest variable in shaking the AI (and player) out of obsessing about expansion.

It is partly my rustiness, but I am getting beaten by the AI with these settings. Even from a seemingly commanding position. Spamming warriors to make the AI think you have a strong military does not seem to work so I will have to switch off the autopilot and actually engage my brain to navigate the early game.