AI Bonuses at the highest levels

[NFIH]

Originally posted by deredico



AI starts with an extra 8 hammers @ Noble.

Is that a one-time bonus or is that +8 hammers each turn? Is that per city or just the capital?

Also, do you know which line in the above table refers to this bonus?

[deredico]

The 8 hammer bonus is a one time deal. The above table does not reflect this. It can be confirmed by starting a game in Noble setting, found your first city, end the turn and then use World Builder to go in and look at the AI cities.

8 hammers is about a 4-6 turn advantage at Noble.

[NFIH]

Originally posted by deredico
The 8 hammer bonus is a one time deal. The above table does not reflect this. It can be confirmed by starting a game in Noble setting, found your first city, end the turn and then use World Builder to go in and look at the AI cities.

8 hammers is about a 4-6 turn advantage at Noble.

Well, yes, that's what I did. So doesn't this mean that in fact it isn't true that Noble is the level at which the AI gets no bonuses? Because everyone seems to insist that Noble is the "fair" level. It's a large advantage on certain settings such as huge map/Pangea/11 civs where the civs start relatively close to each other to begin with. It means you get hemmed in quickly and you're forced to play a warmonger style to get back into the game.

Is the level one below Noble the one where the AI gets no starting hammer bonus?

[deredico]

Yes, it is true that Noble is not a completely level playing field, but it is the closest to being one in that the human player and the AI have the same research rates.

There is no 100 percent equal footing on any setting.

[SebP]

Originally posted by NFIH


Well, yes, that's what I did. So doesn't this mean that in fact it isn't true that Noble is the level at which the AI gets no bonuses? Because everyone seems to insist that Noble is the "fair" level. It's a large advantage on certain settings such as huge map/Pangea/11 civs where the civs start relatively close to each other to begin with. It means you get hemmed in quickly and you're forced to play a warmonger style to get back into the game.

You are right that the AI gets some advantages on Noble. Still you are deceiving yourself if you attribute your problems on Noble to those minor advantages. You must be doing something wrong.
The noble AI can be beaten by just being more efficient then the AI in the builder and tech game.
There are a lot of strategies that work on Noble, but one possibility is to get Civil Service early (with the oracle) and to use a bureaucracy/academy-enhanced capital to fund research. About growth: learn to use slavery to poprush settlers. Granaries are mandatory for this. Then, cottages in most cities and farms mines in one city which only builds military will leave you with a great economy and just enough military units.

I just played a game on Noble using the above-mentioned map settings (pangaea huge, 11 civs, pressed coast lines, high sea level and thus less land). I must admit that i had a really good start that lent itself to the above strategy (7 flood plains, which might be a problem with most leaders for the early sickness, but works like a charm with expansive and financial Victoria). I had to clear lots of jungle though and there's also a good amount of desert around where i didn't have to found in the jungle (The lands are still great admittedly, lots off different ressources).
I built 10 cities before place ran out, my capital now produces 596 beakers in 1585AD and i am killing aztec knights with infantry (As usual, it was him who attacked me for no better reason than religious differences) .
Noone else even researched gunpowder yet, most AIs don't have paper. I am now in the lead by 20 or so techs.

There's definitely no need to expand militarily in this one.
Perhaps you want to look over a few of the savegames? Or play it as a comparison game?

[Blake]

Heh. Yeah. Noble AI's are pathetic.
I've tried to figure out a sufficient handicap for a skilled human to bring them down to noble AI level...

I tried playing with absolutely no terrain improvement except roads, and only 1 worker for building the roads (pillaging all captured improvements). I was still #1 to Liberalism (by a large margin) and easily dominated most of the demographic categories like food and production .
That gives you an idea just how bad the AI's city (mis)management is...

Granted a newbie also labors under the same governor and worker automation, but humans can at least learn. The poor noble AI's are stuck that way forever..

[couerdelion]

We know how bad the AI is at city and worker management because we can let it look after our own and see just how poorly it works. We know how bad it is at fighting because we see how it fights against us. It is staggering that the AI is so poor that it can be beaten with no tile improvements. Just think what this means for your happiness and health when you have no bonuses and to think that the AI can’t beat that is just staggering.

I’ve sometimes wondered how long it would take to win a Noble game if I simply went straight for a conquest victory. I would imagine that the only thing that would save the AI from a Pre-Liberalism defeat would be the oceans separating the cities from the human armies.

[NFIH]

Originally posted by SebP


You are right that the AI gets some advantages on Noble. Still you are deceiving yourself if you attribute your problems on Noble to those minor advantages. You must be doing something wrong.
The noble AI can be beaten by just being more efficient then the AI in the builder and tech game.
There are a lot of strategies that work on Noble, but one possibility is to get Civil Service early (with the oracle) and to use a bureaucracy/academy-enhanced capital to fund research. About growth: learn to use slavery to poprush settlers. Granaries are mandatory for this.

OK, except that this seems to be somewhat map dependent. So far I only play huge maps, pangea/terra/11 civs. This means we all start pretty much on top of each other. I can choose to rush settlers to try and get breathing room. But I'm at a disadvantage in this early game because of the AI's 5-turn lead--its settlers simply get to the spots before mine can. It also means I can't run research at 100% or so because I'm already paying maintenance costs. This in turn means I fall behind on research and can't get the Oracle slingshot or civil service early.

Or I can choose to not expand and stay competitive with research. But then I don't have enough cities to pump out enough units to stop the inevitable early declaration of war by an AI neighbour.

I feel confident that all of this would disappear if I lowered the number of players on the map so that I had a bit more room, which would in turn negate the AI's turn advantage. But I don't want to. Not yet anyway.

As it is, I play warmonger and beat the AIs that way. But it gets tedious.

Then, cottages in most cities and farms mines in one city which only builds military will leave you with a great economy and just enough military units.

Hmm, I find one city alone producing military units is simply not enough. I will be attacked by the AI.

I just played a game on Noble using the above-mentioned map settings (pangaea huge, 11 civs, pressed coast lines, high sea level and thus less land). I must admit that i had a really good start that lent itself to the above strategy (7 flood plains, which might be a problem with most leaders for the early sickness, but works like a charm with expansive and financial Victoria). I had to clear lots of jungle though and there's also a good amount of desert around where i didn't have to found in the jungle (The lands are still great admittedly, lots off different ressources).
I built 10 cities before place ran out, my capital now produces 596 beakers in 1585AD and i am killing aztec knights with infantry (As usual, it was him who attacked me for no better reason than religious differences) .
Noone else even researched gunpowder yet, most AIs don't have paper. I am now in the lead by 20 or so techs.

Amazing. Are you saying you founded 10 cities before even running up against an AI border? That's incredible. That has never happened to me. Ever. If I could manage to found that amount of cities without border issues, then yeah, I'd be cruising, too.

There's definitely no need to expand militarily in this one.
Perhaps you want to look over a few of the savegames? Or play it as a comparison game?

Hmm, don't know what to tell you. I can attest to starting games and within a mere five turns of movement, the scouts of two civs have already encountered me--meaning their civs are so close that there isn't possibly any chance of founding something like 10 cities.

But if you have saves I can look at, sure, I'm up for looking at them.

[Quillan]

Try something other than pangea maps then. I usally play huge/continents/medium sea level, and under normal circumstances I can fit in 8-10 cities before really being crowded. They aren't all where I'd like them to be, because usually while I'm placing a city in site 1 one of the neighbors grabs something close enough to site 2 that I can't get full usage out of it.

With proper development, even paying maintenance costs doesn't hurt your research much. It sounds to me like you aren't generating enough commerce. Cottages, cottages, cottages! You have to get them down early and often, and work them. The commerce generated from a bunch of these takes up the slack. Sure you won't be running 100% science, but when your GNP is twice that of your closest rival, even running at 60% science still outresearches him. Concentrate on commerce development alongside everything else you're doing. You definitely need it.

[SebP]

Originally posted by NFIH


OK, except that this seems to be somewhat map dependent. So far I only play huge maps, pangea/terra/11 civs. This means we all start pretty much on top of each other. I can choose to rush settlers to try and get breathing room. But I'm at a disadvantage in this early game because of the AI's 5-turn lead--its settlers simply get to the spots before mine can. It also means I can't run research at 100% or so because I'm already paying maintenance costs. This in turn means I fall behind on research and can't get the Oracle slingshot or civil service early.

It worked for me. i even got alphabet before getting code of laws and then traded for mysticism/meditation and researched priesthood(I originally didn't want to do the sling shot, i only learned it when starting on monarch and didn't want to resort to such unfair means as deep tech beelines, but when i saw the tech scene when i got alphabet... I decided to still try it).

Or I can choose to not expand and stay competitive with research. But then I don't have enough cities to pump out enough units to stop the inevitable early declaration of war by an AI neighbour.

Perhaps you adopt religion too early. I only adopted confucianism long after most of my northern neighbours had. Mostly to stay on jewish Montezuma's good side for a while longer. The way i raced through the techs i didn't need that many units really. I alos gave him tribute more than once. I didn't see why not with teh cheap techs he demanded and NONE of the AIs bothering to get alphabet for a long long time.

Hmm, I find one city alone producing military units is simply not enough. I will be attacked by the AI.

Just make sure the attack won't come from several directions at the same time -> religion.
I admit this game felt a lot like civ3 and alpha centauri in terms of unit counts. I had very few and it still worked. I attributed it to the level. However, i can't really know this. I don't play noble.

Amazing. Are you saying you founded 10 cities before even running up against an AI border? That's incredible. That has never happened to me. Ever. If I could manage to found that amount of cities without border issues, then yeah, I'd be cruising, too.

Exactly. I may have gotten lucky: Upon further inspection my distance TO AI capitals is bigger than the distances between the AI capitals. And my northern neighbours (Americans) got an especially wretched start.
I don't usually play Pangea maps. In any case. 8 cities and only 6 oft them good would have been enough too, to run away with the game in just the same manner.
Also, note that on Terra, the old world should be way more crowded than a Pangaea. Terra maps are bigger than pangaea ones, but i think the old world is still smaller than a pangaea.
However, since the AI doesn't race for astronomy, terra on noble is child's play.

I'll attach the 4000BC save, and the other 3 saves i found as well. They are not from the best possible turns i guess. The early years are not represented, earliest save is 275BC.

possible spoilers: (well after reading it again, the above has probably more spoilers in it than this. whatever)

The 1655AD one really shows the difference between the Ai and a human. Compare the city of Texcoco which i just conquered (remember,

the aztecs attacked ME) with the ones of Canterbury, Warwick and Newcastle. Texcoco was founded after Canterbury, but before Warwick. Or else i would have founded there myself. (And i got rid off the jungle really late too, it's still there in the 800 AD save.)
I didn't build quite enough workers, i needed to make sure confucianism would spread to the western neighbours at least.
Consequently i switched to serfdom early, and didn't use slavery to it's full effect. I pretty much used it only early to get the settlers built. That was pretty stupid. Why switch to serfdom if you only have 6 or so workers? I could have just built 3 more for the same effect. And still, for all the errors i made, look at how much i'm leading... I also still don't use the prioritize growth/production/commerce buttons but micromanage my cities myself, but only sporadically, for instance i don't check citizen allocation after every new pop point right away... so there, probably another error too, word has it that knowing how to use those buttons is actually better than halfheartedly micromanaging citizens)

I was kinda racing through this game and didn't think too much. I didn't even bother to build a great people city but just cottage spammed. If you find any grave errors in my gameplay (except the ones mentioned above), i attribute them to this. Off course i usually play WAY better . No really, i guess i don't, not by much.

[SebP]

more saves:

[SebP]

How can you add more than one attachment?

[SebP]

Oh well. Let's build up some uh... post count. (Hopefully i don'T get a reset)

oh, and v.CHR is BC in german, n.Chr is AD

[couerdelion]

There still has to be something fundamentally wrong with the approach if you are getting DoWed at Noble level. I think the last time someone DoWed me was in a Monarch level game with Mao when I managed to get three DoWs. Two of these were while I was at war with the only exception being when Jules declared war early in the game and sent a massive invasion force of about three archers and a galley.

In more recent games, I’ve sat next to several aggressive neighbours and not had a single declaration against me. All this time I am selecting one civ after another to pick holes out of – pre-emptive strike to monopolise Iron and/or Ivory being among my favourites.

I think the principle is quite simple. As long as you have a reasonably strong army, they won’t dare to declare war. Early on, a couple of archers is normally enough. As things progress you’ll need chariots and/or axemen to dissuade potential invaders.

But a pre-slingshot DOW from the AI is quite rare at any level. You ought to have time to complete it before any sizeable force comes knocking and once you have the production advantages of CS, you should be able to build up sufficient forces to dominate the area. If CS slingshots are possible at Immortal level – I don’t think I’ve heard of one done at Deity level – then they ought to be safe at Prince or lower levels.

[SebP]

Yep. From what i remember from a thread about the xml files, Montezuma won't attack if you have 131% of his power (or maybe the values were interpreted the other way around, gandhi will only attack with 131% your power, while montezuma will attack with even 70% your power... in any case the principle is the same). I had very few units. I never built the heroic epic.
When the aztecs attacked me in my industrial age i had only 3 macemen and 3 archers total in the 2 border cities, as well as 6 catapults and only 1 archer or warrior in other cities. It happened in this game because i was so running away with the techs that frankly i didn't care. When the aztecs came knocking, i upgraded 8 units to infantry and that was that.
Keep in mind i was trying to NOT expand by force, just to prove the point
I have played only 4 or 5 noble games in total so i was kinda trying out.