City Development

I am playing a lot and find the AI at a virtual standstill. In my current game Siam and Rome have been at war for about a thousand years. Siam took one city, then lost it. They are pretty evenly matched, though. Speaking generally, there seems to be some imbalance early in the game, so that when I meet the other continent one civ is smaller than the others. Then near the end with artillery another shift occurs. But I have yet to see a runaway AI with the new mod. That's several games.

Hmm this sounds less good. The problem I was anticipating with making city defense strong is that the AI only takes settlements by luck, when they once in an eon use siege weapons and melee units. In vanilla I typically have the impression that after the first wave is blunted, the AI doesn't mount a sensible attack anymore because it doesn't create formations and instead sends their units uselessly to their death. This means that the player, who can work with the system, has another disproportional advantage.

Artillery is really good, maybe this is a pointer to a better system: Give all siege units 3 range but make them weak against field units while strong against cities (I think you already have this except for range). That way you need a lot of them to make a dent in a field army (instead of owning them as you do now) but a couple as support for taking cities is fine.
 
Artillery is really good, maybe this is a pointer to a better system: Give all siege units 3 range but make them weak against field units while strong against cities (I think you already have this except for range). That way you need a lot of them to make a dent in a field army (instead of owning them as you do now) but a couple as support for taking cities is fine.

I really like this idea. Archery units could be kept as they are, but catapults, trebs, cannons, and artillery should have some sort of disadvantage against units. They could still be useful if you had several, and good for picking off the almost dead, but they shouldn't be able to dominate as they do now. Maybe cut their base ranged strength, but add more to the city bonus to bring it back up to where it was?
 
The mod already has the pro-city, anti-unit bias. This could be made stronger for cats and trebs, although cannons on up should keep their ability to bite other units.

I like the idea of 3-range. Since all units are calibrated for their era - rifles are technically range, but they're melee in the modern world - the fact that catapults would have the same range as artillery is no less inaccurate than bows having the same as cannon.

Let's remember that the goal here is to help the AI, right? So the question is whether it would employ siege weapons better if they had an extra hex of range. I think a safe guess is 33-50% better!

On a separate note, I've noticed the AI building more generally appropriate units with Thal's mods. By this I mean Greece having swarms of CC, more horse units in general, and all the AI building multiple ships on a continents map with five continents. Can the Mod possibly be responsible for this?
 
The mod already has the pro-city, anti-unit bias. This could be made stronger for cats and trebs, although cannons on up should keep their ability to bite other units.

I like the idea of 3-range. Since all units are calibrated for their era - rifles are technically range, but they're melee in the modern world - the fact that catapults would have the same range as artillery is no less inaccurate than bows having the same as cannon.

Let's remember that the goal here is to help the AI, right? So the question is whether it would employ siege weapons better if they had an extra hex of range. I think a safe guess is 33-50% better!

On a separate note, I've noticed the AI building more generally appropriate units with Thal's mods. By this I mean Greece having swarms of CC, more horse units in general, and all the AI building multiple ships on a continents map with five continents. Can the Mod possibly be responsible for this?

The AI still has problems with artillery but it seems to handle artillery better than other units. Two disadvantages to giving units 3 range are that you often get blasted by a unit you can't even see, and without indirect fire it's often not very meaningful but great in other cases. Maybe all units should have an extra sight, too, at any rate. I kind of dislike how you so often get killed by units coming out of the fog, especially if they're horses.
 
The AI still has problems with artillery but it seems to handle artillery better than other units. Two disadvantages to giving units 3 range are that you often get blasted by a unit you can't even see, and without indirect fire it's often not very meaningful but great in other cases. Maybe all units should have an extra sight, too, at any rate. I kind of dislike how you so often get killed by units coming out of the fog, especially if they're horses.

I hate getting killed by Persians in a GA.

The AI is more clever with artillery than just about any other unit. If I have a ship messing with their coasts, they eventually move artillery into their coastal cities, and then do some quick, serious damage.
 
@alpaca
In my current game the AIs have wiped out 2 civs and about ~5 citystates. (epic immortal large pangaea) It's typically more than this but the terrain is very unique, large mountain ranges blocking off civs from one another. City-states are the toughest to take since those focus on defense buildings and have a Palace for an extra +4, so it's my baseline to see if the AI is coping well.

I'm playing a standard emperor game on pangea with the latest balance-combined. It is turn 200 and no city states have been captured by the AI and only 1 civ has been destroyed. That was a combo of AI Siam and me though.

My previous game was the same way. No city states captured by AI and no civs conquered. Cities are just too strong to take without siege.
Is this just because the AI is too stupid since I am on emperor instead of a higher level?
 
I think 3 range siege may be the answer here. Playing my current game with these mods I noticed that AI's would usually flood a city with melee units but refuse to attack because base city defense was too high. Only when they got a siege weapon in place were they able to start taking the city down and that could take a LOONG time.

On the other side, I found my war with elizabeth really tough as she had the great wall and spammed longbows and I had a tough time getting past her with arrows raining down all over me. She actually used them quite well tactically, aside from when she tried flanking me with them :smoke:. But that is getting fixed in the patch apparently.
 
I guess I'm in the minority here, but I think 3 range siege is OP.
 
Increasing range but decreasing strength isn't OP IMO, it just levels the playing field as the AI's can now effectively use siege as well as the human. It also prevents some of the ridiculous stalemates which occur quite frequently which prevent the AI from taking cities as their melee units won't attack yet but they're blocking siege unit's ability to bombard.
 
Increasing range but decreasing strength isn't OP IMO, it just levels the playing field as the AI's can now effectively use siege as well as the human. It also prevents some of the ridiculous stalemates which occur quite frequently which prevent the AI from taking cities as their melee units won't attack yet but they're blocking siege unit's ability to bombard.

I was just pitching an idea, it has to be implemented and tested to see if it works. Personally, I think 3 range in itself are strong but not overpowered as long as units don't just die to a siege strike. Siege units have severe disadvantages in that they can't fire the same turn they are moved (except for rocket artillery), making them primarily defensive weapons, and they are vulnerable to melee units. Of course, a range increase allows you to more easily defend siege units with other units so I'm not sure it will work. As I said, it would have to be implemented and tested.
 
is it possible to add a new national wonder to the game that improves logistics in ones own empire (thus also buffing empire defense)

Imperial Road Network (national Wonder)
requires: market in every city (or in 5 cities)
prequisite: Engineering
reduces movement costs across modernized terrain to 1/2 (also possible to 1), bridges inclusive (when both sides of the river are modernized).

to simulate that modernized ressources surely have some basic road access and my soldiers surely don't have to hack through the underwoods to get to my big iron mine.
 
Increasing range but decreasing strength isn't OP IMO, it just levels the playing field as the AI's can now effectively use siege as well as the human. It also prevents some of the ridiculous stalemates which occur quite frequently which prevent the AI from taking cities as their melee units won't attack yet but they're blocking siege unit's ability to bombard.

I don't see how this could possibly be true, unfortunately. It may help the AI a little, but honestly it would just be a buff to the player.

In my experience, when I'm on a warpath and I either start getting the range promotion or upgrade to artillery, the war tilts dramatically in my favor, as I can now have TWO lines of siege behind my defensive line rather than one. I really don't see the AI taking advantage of this like a human can and will. Even with a large penalty against units, if you have 6-7+ siege units (and a great majority of the time, the AI will not have that many) behind a fortified line of melee, the AI is pretty much toast because of it's penchant for moving every turn and/or attacking with bad odds. And then once their army is decimated you have a massive and well promoted siege army to sweep through their empire.

I have read and understand all the arguments proposed here, but this change will help the player far more than the AI IMO.

If someone wants to put up a mod with this I will be happy to test it.
 
I don't see how this could possibly be true, unfortunately. It may help the AI a little, but honestly it would just be a buff to the player...

I have read and understand all the arguments proposed here, but this change will help the player far more than the AI IMO.

If someone wants to put up a mod with this I will be happy to test it.

I mentioned this earlier. However, the majority of changes helps the player more than the AI. The benefit of the adding 3 range is to create some give and take. However, I'm wondering if the answer may be to simply reduce the buff to walled cities a little from where it is now.

In my current game Russia is on an island alone with two CS, and has yet to take either. I am on a continent with the Aztecs, French, Songhai, Mongols and Chinese. The Aztecs attacked me early. The French attacked them briefly mid-game. Their war was inconclusive - and that is all the fighting I've seen. Not one CS fell until the usual nineteenth century burst of artillery-fueled expansion. I'm sure there are games where some AI become more dominant, but it has really fallen off with the buffed city defenses.
 
Thal,

Do you mean an upcoming v10, or one I downloaded last week that has yet to be approved?

That version was a mistake. The release version was at the present time v8, I incremented it to 9 then 10 in development (too far). The current release version is v9.

Artillery did get nerfed quite a bit in the combat mod. Most siege units have -20%:c5rangedstrength: and artillery have -25%:c5rangedstrength:. This puts Cannon at 21 and artillery 24, a more reasonable increase. The primary advantage now is just the bonus range and terrain-ignore. Artillery in the mod also require an additional tech more than vanilla.

In my current game 3 out of 9 civs got killed by AIs. In vanilla it's usually more around 6-7 out of 9. They still can competently attack each other and conquer cities... at least on immortal difficulty level. What I've been thinking is on prince-emperor difficulty settings the lower production bonus of the AI means they can't field as many units, thus have a harder time capturing cities. It's not a bad thing though if you think about how AIs capturing more cities does make the game harder, the purpose of difficulty levels.

What frustrates me to no end, as I'm sure some of you have seen me talk about, is the fact buildings cannot alter city HP. This is the perfect solution... make killing a city take longer but don't alter the base defensive strength beyond vanilla values (5/7.5/12). It's not possible yet since city max HP is currently stored as a global value. :(

I'll just change defensive buildings to a 1.5x modifier as a temporary solution, and increase city hitpoints to 40 (currently 30 in the mod). It'll probably make cities too easy to capture again, but at least will give the AI a little boost.
 
They still can competently attack each other and conquer cities... at least on immortal difficulty level. What I've been thinking is on prince-emperor difficulty settings the lower production bonus of the AI means they can't field as many units, thus have a harder time capturing cities.

I'll bet you're right that this is why you're getting different results from me. Could you lower the city defense bonuses incrementally, so that we can inch our way to a pre-C++ medium?
 
It would be somewhat ugly, but I think you could mod a trick to simulate cities' HP going up from buildings. Set the max city HP to what it would be with all the defensive buildings up. Then, once per turn, "damage" cities that lack the proper buildings. The City:SetDamage() call exists, and probably works (haven't tested it). From the city health bar, it would look like cities have say 10/40 hit points at the start, but if you can get over that, it would achieve the goal. The other problem is that it would take a while for a city to "heal up" after it built a defensive building. You might decide that's cool. If not, then you'd have to monitor when this happens and raise the hit points artificially in the same piece of code. There would a 1-turn delay in getting the health up.
 
I don't see how this could possibly be true, unfortunately. It may help the AI a little, but honestly it would just be a buff to the player.

In my experience, when I'm on a warpath and I either start getting the range promotion or upgrade to artillery, the war tilts dramatically in my favor, as I can now have TWO lines of siege behind my defensive line rather than one. I really don't see the AI taking advantage of this like a human can and will. Even with a large penalty against units, if you have 6-7+ siege units (and a great majority of the time, the AI will not have that many) behind a fortified line of melee, the AI is pretty much toast because of it's penchant for moving every turn and/or attacking with bad odds. And then once their army is decimated you have a massive and well promoted siege army to sweep through their empire.

I have read and understand all the arguments proposed here, but this change will help the player far more than the AI IMO.

If someone wants to put up a mod with this I will be happy to test it.

OK so maybe I shouldn't have said that the Ai's will use the siege weapons as well as humans, obviously, they don't have our level of tactical ability. What I really meant was that comparitively, the AI's fair much better when they have 3 range units at their disposal. This is why you generally see the AI being much more effective at war once they hit artillery. And going up against Liz's longbows is surprisingly difficult given how poorly everyone claims the AI fares at warfare.

Here's the two main reasons i see why the AI's need 3 range and why this boost will be more significant for them than for us.

1) 1UPT is the main cause of this. Typically, we can see small human, or even city state armies fending off larger no.s of AI armies as 1UPT severely restricts the amount of units the AI can have fighting in the front line. We often see them swarming cities or chokepoints with no chance in hell of getting through yet in doing so, they block their own siege units, the only ones capable of actually doing some damage. The extra range gives the AI the ability to actually leverage somewhat their extra numbers.

2) The AI can't defend its siege units. Personally, I have little trouble coordinating things so that my 2 range catapults are out of reach, so I don't feel that the extra range will help the player too much in this regard. The AI however, constantly places these units in extremely vulnerable positions. With 3 range, the AI tends to be setting up their siege for attacks BEFORE they get too close and so this problem will be alleviated somewhat (from what i've seen)

At the very least we can test out if what i'm saying will occur or not but from what i've seen I think it will help quite a bit.
 
Thal,

Would it be too hard to leave the Emperor/Deity setting alone, and mildly adjust the mid-range one, so that you offer two versions of the Combat (or whatever) mode?
 
@Perkus
Ouch, I see your idea... though simple in concept it'd be a pain to implement. You'd also get healing numbers showing up on every single city... not sure that'd be an ideal tradeoff of gameplay vs visuals, but it's creative.

@bobbyboy29
On the topic of AI tactical command, supposedly the next patch will make it so the AI properly puts ranged units behind melee.

The problem is calculating optimal unit placement is not just processing-intensive, it's completely infeasible as the number of units rises, regardless of the quality of the AI or your computer. It's an np-hard algorithm. Unfortunately this leaves me pragmatic of the likelihood of any significant AI improvements in this regard in the near future, we probably won't get to something resembling the competency of the Civ IV AI until a year or so after release.

@Txurce
Generally any changes will affect difficulties equally, with more civs getting conquered, so it shouldn't be a big deal. It both makes the game easier (you can take things) and harder (bigger rivals) at the same time.
 
@Perkus
Ouch, I see your idea... though simple in concept it'd be a pain to implement. You'd also get healing numbers showing up on every single city... not sure that'd be an ideal tradeoff of gameplay vs visuals, but it's creative.

@bobbyboy29
On the topic of AI tactical command, supposedly the next patch will make it so the AI properly puts ranged units behind melee.

The problem is calculating optimal unit placement is not just processing-intensive, it's completely infeasible as the number of units rises, regardless of the quality of the AI or your computer. It's an np-hard algorithm. Unfortunately this leaves me pragmatic of the likelihood of any significant AI improvements in this regard in the near future, we probably won't get to something resembling the competency of the Civ IV AI until a year or so after release.

@Txurce
Generally any changes will affect difficulties equally, with more civs getting conquered, so it shouldn't be a big deal. It both makes the game easier (you can take things) and harder (bigger rivals) at the same time.

You can set the heal rate, too, if you like (at least there's a command for it) :lol:

As for AI: What I would do as a game designer would be to abstract most of the AI vs AI combat. What a player doesn't see doesn't fazz them (and it's actually annoying waiting for the AI units to duke it out), so you only really need to model AI movement and combat properly if it's against the player. This saves a huge amount of processing, obviously, and allows a much better AI against the player.

If you don't want to go quite that far, there's still the option of two different tactical AIs: A simplistic one for AI vs AI, and a better (more resource-hogging) one for AI vs player battles. It doesn't need to be chess-like good; in Civ5, projection one turn into the future should be sufficient.

Strategic placement is more difficult, of course, but an abstraction would help here, too. Of course, the players might feel fooled by AI units that are just smoke&mirrors and only actually appear when they get into the area.
 
Top Bottom