Acken's Minimalistic Balance for singleplayer (and AI improvements)

[TheGrumpyBuddha]

Hmm... that should be okay, depending on your opponent of course. Go after softer targets, preferably tradition civs with nice cites and wonders, and not stuff like honor players with a large army (even if they look threatening). As long as they do not have medieval units you should be fine if you can gather like 6 swords and some support units by that time. Even if they also have swords you can use flanking and the fact that they promote cover instead of shock/drill to beat them in my experience even though you might lose 1 or 2 swords. Catapults are necessary to take harder cities efficiently, they are hardly ever targeted if you have a wounded swordsman in range, which you probably will. If you still struggle try the Vikings for 23 strength move 3 (+) units at MC.

I don't play as often as you guys, but I'll tell you that by turn 80 or so the enemy has 3-4 CBs and 4-6 melee units (even if they are just spearmen, it's tough to move fast against that kind of carpet) and I'd say at least half the time I go after the nearest Tradition foe, he pops the Great Wall .

By the time I get to his cap, it's turn 90-100 and his cap is toughness 23+, which means I'm often losing a Swordsman a turn if he's attacking. I may need to try catapults, I haven't built them in forever.

There have been games where I've managed a quicker sweep, but they're when I open Honor, so my infrastructure/science is crap, and I have happiness problems. I'd love to see a Let's Play and/or a posted saved game.

 

[noto2]

I've done a lot of warring in the classic age as a non-Honour civ. The reason is because the AI loves to plop a city down right in my face, far, far away from the rest of his/her empire too often. I respond to that by taking that city. I don't always try to go straight for their capital. I might just take a city or two and weaken them first, then take the capital in a future war. Also, in a close fight, don't go straight for the city. Take out their units first, one by one, and withdraw your wounded units. It takes longer, but as time goes on you'll slowly whittle down their forces, while your army grows from reinforcements and grows in experience. Then you can push forward and take the city.

 

[TheGrumpyBuddha]

I've done a lot of warring in the classic age as a non-Honour civ. The reason is because the AI loves to plop a city down right in my face, far, far away from the rest of his/her empire too often. I respond to that by taking that city. I don't always try to go straight for their capital. I might just take a city or two and weaken them first, then take the capital in a future war. Also, in a close fight, don't go straight for the city. Take out their units first, one by one, and withdraw your wounded units. It takes longer, but as time goes on you'll slowly whittle down their forces, while your army grows from reinforcements and grows in experience. Then you can push forward and take the city.

The problem is that, on Immortal+ at least, you don't have time to dilly-dally because you're competing with that AI on the other end of the map who will have 1500+ science by turn 200 ... best I can tell, if you haven't taken down at least a couple juicy civs by turn 120, you're boned, and even if you have, you still might be boned, it depends what's happening Way Over There.

I will say, though, that giving my guys the double-Cover bonus + trying Catapults has helped with the charges!

 

[Acken]

A bit disappointed in the audio but here are the 3 first parts of my LP:

Byzantium v4 Acken's balance mod. Demigod difficulty Culture:

Link to video.

 

[Acken]

I'm currently brainstorming to make trade posts a more interesting choice. My current idea is to make them a bit better but with restriction:
-Available at The Wheel and give +1 gold
-+1 at currency
-+1 at economics
-+1 with commerce
-+1 with golden age
Possible restriction candidates:
-Only adjacent to the city tile
-Same rules as chateaux

Edit: While working more on it and doing some maths I also can see how fishy the gold vs production is in the mod and in the base game.
Let's try comparing a base bonus of 2Production vs 4Gold.
Without any bonuses the current balance of gold vs production is usually assumed to be 1 Production = 2 Gold for most of my changes. This is a very poor ratio early game and player will usually preffer to focus on food or production and simply gather gold through trade. For example the granary ratio is 340/60=5.6 and the aqueduct 4 (4 gold for every hammer). But you also get your buildings a lot faster to snowball your new cities so it helps.

The bonuses will soon start to appear. Mid game it's very easy to have 50 to 75 bonus in gold. Let's assume 60%. On the other hand production usually is in the 10 to 30% area. This mean that a 2P bonus now gives 2.4 while a 4G bonus gives 6.4G. We have a 1:2.6 ratio, a lot better. If we look at a mid game building, the school, the cost ratio also got better over time. It is 920/300 = 3.1 for the school. On the other hand the snowball effect should be smaller.

This is problematic because without any purchasing bonus we're already closing in on the cost ratio.

Now things look really imbalanced to me if we add the buying reduction. Let's say you get BB and Mercantilism: These two bonuses are additive and will reduce the cost by 40%. This effectively mean that 1G is now worth 1/0.6=1.67 G, or a 67% increase. So our 6.4G is now worth 10.7G. We now have a 1:4.5 ratio in a world where things actually have a cost matching a 1:3 ratio at that stage (early industrial).
In other words the gold strategy is now approximately 50% more effective. This is even worse when we add ideology reduction and better cost ratios from the end game. I don't mind a gold strategy using a lot of gold but this looks a bit out of hand to me if you can reach double effectiveness with the added benefit of getting the building immediately.

But this ratio is something from my mod. Chumchu has been warning me there's simply too much gold and indeed. If we use what the base game has a tendency to do and compare 2G vs 2P instead we then get 2.4 for production and 5.3gold. This is a ratio of 1:2.2. Which isn't that good but will get better in the later era when the cost ratio will diminish.

To sum up:
-A 1:1 comparison makes early gold really really bad.
-A 1:2 comparison makes late gold way too efficient with hurrying cost modifiers

One thing to note however is that if without cost reductions the ratios arent very good, gold still has a benefit for city states, maintenance, trade and the occasional rush.

Possible solutions:
-Reduce the hurrying cost modifiers
-Reduce the slope of the cost ratio so that late game buildings cost more
-Go back to a 1:1 design but diminish the slope of the cost ratio so that early buildings are cheaper
-A more radical change could be to make the cost ratio a constant and then have gold and production follow similar bonuses and reduce the effectiveness of cost modifiers to reasonable levels.

 

[Ravellion]

I find that sentiment a bit strange. It's true that the slot increases the early game science speed but by only +2 science. Compared to v3, at university you now have 2 slots for a total of +4 instead of +3. Not that significant.

The first GS would come faster though so it may be an explanation. That's the problem with GS, they will accelerate whatever era they spawn in.

A possibility would be to simply go back to 2 slot universities or simply slow down GS spawn rate by limiting it to 3 slots and 4 with the NC.

As others mentioned, it is indeed the early academies that the scientists give. Everything's fine till the medieval era, but then the university boosted academies give quite a significant boost.

 

[chumchu]

Re: gold why not aim for 2p:3g for yields and try to stabilise puchasing cost at 1p:4g across the board?

 

[TheGrumpyBuddha]

Acken -- in your Let's Play, what mod are you running other than your v4 -- to show extra information about individual spaces, etc.

 

[Acken]

Re: gold why not aim for 2p:3g for yields and try to stabilise puchasing cost at 1p:4g across the board?

Ultimately it would make late game stuff cost a lot more and some ancient/classic era stuff less. A granary would cost 240 for example while a school would cost 1200.
With previous examples this would mean that with 60% gold bonus and BB+mercantilism we get 2.2p:8g which is almost the 1:4 (1:3.6) ratio. Without the rush diminution we get a 2.2:4.8 ratio (1:2.2 ratio).

I think the accumulation of rush buying reduction still creates a problem where gold is a bit too good with it and not enough without. 3 Policies in Commerce have the potential to increase the effectiveness of gold by 67% (BB+Mercantilism) that's not very healthy I think.
With something like skyscrappers, the gold you save for buildings has an effectiveness increased by 150% (ideology bonuses are multiplicative so the bonus is (1-0.15-0.25)*(1-0.33333)=0.4, 1/0.4=2.5 or 1+150%).

I think I'll spend some time with Excel coming with a good model.

Acken -- in your Let's Play, what mod are you running other than your v4 -- to show extra information about individual spaces, etc.

Enhanced user interface

 

[Dahebbie]

I think I'll spend some time with Excel coming with a good model

Thx for all this effort. So cool you spend so much time on these details that many of us find important too.

Basicaly imo: trading p. shall come sooner (the weel sounds good), and MUST be worth building from the start. Of cause balanced, but still.
Idea: make them really good, but limited. Maybe 2 pr city or something.
River bonus makes sense to me as well. (Prioritize between food/gold)
Dono, so many ways to go, sure youll find the balanced one.

Thx again

 

[Nigel_Tufnel2]

If you limit TP per city, I would prefer X per city radius; rather than adding a geographical condition, such as in the first ring.

 

[Acken]

Difficult because of city borders not being exclusive to the core. Most possible rules are about adjacency or features of the terrain itself.

 

[Lord Yanaek]

I'm currently brainstorming

Since it's the brainstorming stage, i guess i can throw some silly ideas here just in case one of them is not totally silly. If you think there is a balance issue between early/late game buildings purchases that makes it hard to have gold that's not useless early or Op late, maybe rather than try to reach the perfect balance you could avoid the issue by changing the logic behind buildings purchases.

• CBP uses an accelerated production when you "buy" a building. You'll build it twice as fast but you still need production (don't know how it works in other countries but here in Belgium even when the autorities pay plenty of money to get something built, it's still finished severl months/years late )
• If possible from a coding p.o.v. rushing could be possible only for buildings from the previous era. This way you could get a mid-game city productive fast by rushing granary, workshop, library, but you won't plop Public Schools as soon as they are made available (probably with an exception for Jesuit Education). Alternatively you could get neglected buildings fast (workshops or lighthouses after you prioritized Education) but you'll never get the brand new tools immediately built.

As for the Trading Posts limitation i think Chateau-like makes the most sense (it will effectively limit the number, and they would be placed near the production site of some valuable resource which would give some meaning to the "trading" part of the name). 1st ring is tricky as some cities will have a lot of resources in R1 and coastal cities could have very few available R1 tiles.

Thanks for the LP, will have a look.

 

[Acken]

I'm aware of CBP's idea but I have never understood what problem it really fixes. Hoarding is a decent strategy. That's what we do with unit rushes too. I don't have a problem with it.

The second point is easy to do but it creates weird mechanics I think.

In the end my problem isn't really that players will hoard some gold to rush some schools. My problem is that focusing on gold before doesn't come at a cost from mid to endgame. You're basically saving hammers at a too efficient rate.

I just think there are two problems:
-Efficiency of early gold vs late gold
-Commerce power vs Non-Commerce power

Basically Commerce is probably by far the best secondary tree right now with the highest happiness and a very high increase in gold efficiency coupled with more gold.

 

[noto2]

On gold topic: The slope of gold cost doesn't make any sense, and it might as well be removed. Earlier buildings should be a bit cheaper, and later buildings should be much more expensive. Also, since the purchase reductions are additive, they need to be nerfed.

Another topic: I've just played 5 immortal games in a row, and won all 5, which is something new for me. What really irritates me, however, is that in every single game I play at immortal, I spend the vast majority of my game at war and the reason I win is always because of warfare, whether I'm playing Korea or Egypt or whatever.

At immortal, even if I'm playing as Babylon and I get a fantastic start and build tons of wonders, I fall far behind in everything if I try to play a peaceful game. The Zulu will get an era ahead of me in tech.
The only way to win at the upper levels is by constantly being at war all the time. This has been the case with all iterations of Civ since I can remember.

Low level Civ (Prince/King) ? You can win peacefully or by military force

High level Civ? (Immortal +) Sorry, forget culture or space victory, forget fighting defensive wars... you have to conquer many, many civs to win. Every single game is a warmongering game. You're no longer playing Civ, you're playing Risk or Warcraft


Why is this the case? Because as we go up in difficulty the AI gets economic bonuses but zero military bonuses. Given the same army, the AI will perform equally well at settler as at deity difficulty.

So, what's my crazy idea? Maybe instead of giving the AI economic bonuses only, we could give it a combat bonus, and then tone down its economic bonuses. This would lead to a more balanced game.

For example, you could take the emperor difficulty and give it only monarch bonuses, but then give it a +15% to combat across the board, to compensate. This would make it so the AI wasn't so ridiculously overpowered in terms of economy, and a little more threatening on the battlefield. 15% would be enough that the player would need to be more careful in battle and probably field a slightly better army, but not enough to make the AI unbeatable at war.

Thoughts?

 

[Lord Yanaek]

The second point is easy to do but it creates weird mechanics I think.

Guess it could be abused by rushing next era so you're probably right.
Could also be that you could only buy something you've already built at least once (need some "prototype" but not sure it would solve anything either). Just brainstorming here. It's true that hammers loose much in the late game, maybe you can find the right balance, you seem to be much better at putting numbers together than i am.

At immortal, even if I'm playing as Babylon and I get a fantastic start and build tons of wonders, I fall far behind in everything if I try to play a peaceful game. The Zulu will get an era ahead of me in tech.

Well, i would argue that building tons of wonders won't help you science-wise but i get your point. Without the mod, the big warmonger was often lagging in tech and would soon be crushed by multiple opponents with smaller army but better tech. Now they seem to always snowball into tech domination. Seems like we need some middle ground where going to war can be beneficial, but isn't THE way to go.

The only way to win at the upper levels is by constantly being at war all the time. This has been the case with all iterations of Civ since I can remember.

Low level Civ (Prince/King) ? You can win peacefully or by military force

High level Civ? (Immortal +) Sorry, forget culture or space victory, forget fighting defensive wars... you have to conquer many, many civs to win. Every single game is a warmongering game. You're no longer playing Civ, you're playing Risk or Warcraft

Not without the mod. It's pretty easy to beat Deity CiV with peaceful tradition 4 city science turtle (unless you get forward settled by Pocatello of course). I think it's good that the mod makes aggressive AIs more aggressive (no longer can you win Deity with 3 comp bows at the end of the game ) but you should not be forced to be the aggressor.

For example, you could take the emperor difficulty and give it only monarch bonuses, but then give it a +15% to combat across the board, to compensate.

CBP (again, or maybe it was the old Communitas mod) did this. It's not a bad idea but quite visible. However AI already gets better military on higher levels (more troops and free promotions) so not sure it would really solve the issue. Economic bonuses might have to be toned down as AI becomes actually smarter thought (Firaxis balanced the high level science bonus with an AI that would forget to use scientist slots )

 

[noto2]

I wasn't talking about the ability to win peacefully based on a passive AI.

In Civ we are supposed to have different options to victory. On Prince, for example, I could do two things:

strategy 1: establish a small empire of 5-6 cities and focus on my economy, getting a tech lead and winning by space ship. I would fight wars but they would be primarily defensive in nature, or defending a city state ally. I wouldn't need to war to conquer enemy capitals, for example.

strategy 2: forget building any wonders at all. Focus 100% on the military and conquer everyone and everything in sight. Win by conquest.


The problem is that on Immortal, strategy 1 is out of the question. I like aggressive AI's. I like having to defend my empire from invasions. What I don't like is when I play as Korea, have an empire churning out 1000 beakers per turn at turn 175, have academies and scientists up the wazoo, have 7 cities, avg city pop is 30.....

and yet, the freaking Huns, who have no academies, not much more population than I do, no universities, are still 5 technologies ahead of me.

Can I win such a game? Of course I can. All I have to do is build artillery and battleships and blast the living crap out of everyone. That's why I won a continents game with England on something like turn 240 (using ships of the line, not even battleships)

But can I win without conquering everyone? No. Is it just me, is this just my problem? On high difficulty levels, I can't win without conquering at least a few civs. In my last game, for example, I was Babylon and I was determined to win by space victory. No conquest. Byzantium hit the modern age before I hit the industrial age. Without using my military to smash everyone in sight, the game seems just impossible to win above emperor. The AI is impossible to catch via peaceful means, the only way to beat it is on the battlefield. That's why I think the AI's battlefield performance needs to be buffed, and its economic performance nerfed.

 

[chumchu]

Korea game should easily be winnable by space with that high pop cities. With rationalism and working max specialists you get 2000+ science (250 pop + 100 specialists meaning 800 base science without jungle or academies and a bonus of at least 100%) for a SV before turn 250 unless social policies are totally off. Send me that save if you please. =)

It is very efficent to conquer a player or two but it can hurt your diplomacy and require an investment in military, I like it that way.

 

[Ninakoru]

To me the biggest flaw about gold bonuses is how gold bonuses stack. When you get commerce + order/autocracy you get ridiculously low prices for buildings/units respectively. Calculating multiplicatively those bonuses instead of adding it to a percentage reduction would hinder this exploit.

For example if you get the 15% + 25% + 33% bonus pack, summing percentages reduces the base cost by 73%, but applying them multiplicatively, the reduction goes to 57%.

I'm more with maintaining the original gold ratio with minor boosts (what I did in my balance mod long time ago: No costs on monument/shrine, boosted customs house and GM gold); is good enough, as boosting the ratio is against any minimalist effort, you have to re-balance any gold-related aspect of the game : Hurry purchases, CS investment and unit/building maintenance, gold in policies, gold boosts on land, gold boosts from buildings, techs, promotions, civ traits...

Anyways, about the mod: I think this mod if far from minimalist right now, but still keeps the game core more or less the same, I did similar changes long time ago, I must say I like most of the changes, I'll give it a few tries soon enough, thanks for your hard work!

 

[Lord Yanaek]

In Civ we are supposed to have different options to victory. On Prince, for example, I could do two things: ...

I got this. However you wrote it's always been this way in Civ games. It's not true in CiV5 without this mod. The CB example isn't just an image, I actually won once on Deity without upgrading further than CBs (not sure i had 3, maybe 4 - one in every city).
Totally peaceful game was perfectly possible without this mod as the AI got their late game techs much slower on Deity that they do on Emperor with Acken's great work.

As for the situation with this mod, i actually never played higher than Emperor (or not long, i quickly realized i had to drop the difficulty till i adjusted to the changes) so i can't judge how late game techs go. It's still a WiP. Acken managed to have an AI that actually uses their specialists slots. The high level bonuses were made for an AI that wouldn't use them. Of course the bonuses might need adjustments.
On the same line of thought, i wouldn't start adding artificial bonuses to AI military power until all actual improvements to it's logic are done.