K-Mod: Far Beyond the Sword

[Charles555nc]

I dont like the military support increase idea at all personally. Not at all.

 

[stingo]

I dont like the military support increase idea at all personally. Not at all.

Just curious how do you manage 500 units in late game? It is too good to start a war in late game because gains are massive. Fully developed cities on a plate, just take them all. Wars must be costly at modern times or may be ideas for some other penalties?

 

[Charles555nc]

I dont know what difficulty you are playing on...but the higher you go, the easier it is to sit back and build culture and tech than it is to make a larger navy and a large army and invade other continents (that might have nukes and a religious love for each other). War also can set you way behind in the tech race.

Of course you have to war some, early and mid game, but civilization isnt a peace run. Or you could select "always peace" in the options...and you dont get fully developed cities...alot of buildings are destroyed capturing cities...

 

[noto2]

Karadoc I'm sorry if I confused you earlier. This is really all I want: I want it to take more than 1 turn to raze a city to the ground. I think it's horribly imbalanced that a city can be razed in 1 turn, especially a capital or legendary city. Also note that both Civrevolutions and Civ5 changed this, and made it so that razing takes several turns, dependent on city size. I'm not sure if you could code it, but I would suggest razing do something like -5 population per turn. So a size 20 city would take 4 turns to raze completely, for example.

 

[stingo]

I dont know what difficulty you are playing on...but the higher you go, the easier it is to sit back and build culture and tech than it is to make a larger navy and a large army and invade other continents (that might have nukes and a religious love for each other). War also can set you way behind in the tech race.

Of course you have to war some, early and mid game, but civilization isnt a peace run. Or you could select "always peace" in the options...and you dont get fully developed cities...alot of buildings are destroyed capturing cities...

I play regular noble, but it does not really matter much. Here is the stack AI built to go to war with me (this is only one stack there is much more units this AI has). I am not sure I can sit back. No I do not want always peace since it breaks lots of game mechanics such as diplomacy and expansion. This is only standard map and not even modern time. This bogs down performance and hurts the gameplay by making you slave of micromanagement. Transporting this number of units to a different continent is not my idea of spending free time, but not for AI.

I want to play on large map with many civs and still be able to handle the end game even if my relationships are terrible with inca.

 

[karadoc]

Looks like they built too many cannons.


One thing worth keeping in mind about late-game wars is that although it's typically much easier to take cities (because you can have huge armies which are extremely mobile, making it difficult for anyone to defend); the tradeoff is that taking a city isn't worth nearly as much - because it's already late in the game. Any city you capture will take a significant amount of time to come out of revolt, and then time to build based infrastructure, and it will need units to defend it, and so on... and then the game will be over and you probably won't end up getting much value for your investment. That kind of effect could be thought of as the counterbalance to wars becoming cheaper...

(By the way, one problem with making unit cost scale with era is that there will be a big jump in costs every time you change era... and that would be weird & bad. If you want later units to be more expensive then I think it would be better to just change the cost of individual types of units, so that, for example, tanks would cost more than archers - rather than the cost of archers going up in later eras.)
--

Also, regarding this:

Few random thoughts after playing a game last night.
1. I have noticed a thing which was pretty obvious but somehow escaped my attention before: AIs don't make a resource for gpt trades with each other. In some situations it limits their ability to manage economy significantly and, generally, it seems wrong if a player can do something what an AI can't do. Can it be fixed somehow?
2. It looks like that AI doesn't transport great people through ocean. I've noticed this long ago but this game reminded me: one of civs got a holy city on a remote island. It was never shrined, though this civ generated few great prophets. I suspect that by the same logic they don't put great merchants on boats for trade missions. Once again, can it be fixed? Also, it looks like that AI doesn't transport missionaries, though he does transport corp executives. One of civs spread its religion agressively but only few of island colonies it had was affected (and I suspect it was an auto spread).
3. In the course of a game I attempted to raze a city with a corporate headquarters which belonged to a civ I already was at war with. I expected weak opposition as this city was very far from the main front. My strike force, however, was detected in three turns before attak. And when I reached my target city it had a huge army in it, about thirty highly experienced units. I had 32 marines in my fleet, 7 battleships, 12 destroyers, 4 carriers full of fighters but still had to drop my plan.
Hence I have questions. Am I correct that AI can understand the threat of naval strike during war? Can he, then, also be made to understand it during peace? And is it possible to teach it to employ such tactic (sniping down important cities like corp headquarters and holy cities)?
4. During the game Mao captured Orleans. It had a rich Islamic shrine in it, an even more rich headquarters of Standart Ethanol and Mausoleum of Maussolus, one of the best late game wonders. Also Orleans was right next to Chinese borders. And yet Mao razed it, though this city could, I think, double his GNP. Why? The only reason I can think of is that Orleans was close to a legendary status and France was heading to cultural victory. But France was falling apart and had no real ability to recapture it.
Actually, I think Mao razed every French city he captured during that war.
5. What do you think about the ability of bribing AI players to start wars in which you don't participate? During my last game I found that it is very exploitable. Basically the game was decided by me bribing Suleiman first to attak France to stop them from winning by culture and then by bribing him to attak my main competitor.
6. Oh, one more thing. I have noticed that AI doesn't leverage corps agressively enough. Particulary, it seldom trades for relevant resources and, as I noticed in the beginning, never buys resources for gpt.

For #1, and #2. You're right that the AI currently doesn't know how to do those things. It would probably help them if they could, but I haven't gotten around to teaching them that stuff. Trading gold for resources wouldn't be hard to set up - but it's hard to say how much it would help the AI, because they aren't very good at determining how much a resource is -really- worth, or when a better deal might be available from someone else... they might end up just wasting their money on resources they don't really want anyway. (I often sell my junk resources to AI players who probably don't need them. Maybe it would be good if they AI sold its junk to one another...)

But using boats to transport great people can be confusing and potentially quite risky. With execs and missionaries its easy, because it doesn't matter if they go to the 'wrong' city first, and it doesn't matter much if they get killed en-route; but with great people it matter a lot. They have to get to the right city, safely, and without taking way too long to get there... Anyway, it would be good if the AI could do it, but it's not an easy thing.

For #3... the AI only has very limited understanding of imminent threats. In your case, they probably just fluked it. They mostly just try to defend their most valuable cities and their border cities.

#4. It's almost certainly because Mao was trying to block a cultural victory. Maybe Mao could have held the city anyway... but when someone is close to victory, they don't risk it - they just raze it.

#5. It's exploitable. Yes.
I've been half intending to change the evaluation process for war trades and peace trades... I just haven't gotten around to it. (When I rewrote the AI's evaluation for its own war targets, I was intending to use the same functions to evaluate peace deals and war trades -- but I just haven't done that yet.)

#6. That's true.

--

Essentially all of your points are areas where the AI could potentially be improved -- but none of it is really easy.

 

[noto2]

@stingo: I too have been frustrated at the number of units in play in the late game, and have wondered if there was a way to reduce it while still maintaining balance. I suppose one possibility would be to make late game units more expensive in hammers but also more powerful. Classic age wars are a lot of fun because on a standard map you might have 20-30 units in your main assault stack. By the Napoleonic era you might have 40-60 in your stack, so it's already becoming somewhat cumbersome. In the late game, however, when playing with jet fighters, carriers, modern armour, gunships, etc, nations will have 200, 300, maybe even 400 unit militaries. It just becomes a chore and the game is less fun.

 

[stingo]

Looks like they built too many cannons.


One thing worth keeping in mind about late-game wars is that although it's typically much easier to take cities (because you can have huge armies which are extremely mobile, making it difficult for anyone to defend); the tradeoff is that taking a city isn't worth nearly as much - because it's already late in the game. Any city you capture will take a significant amount of time to come out of revolt, and then time to build based infrastructure, and it will need units to defend it, and so on... and then the game will be over and you probably won't end up getting much value for your investment. That kind of effect could be thought of as the counterbalance to wars becoming cheaper...

(By the way, one problem with making unit cost scale with era is that there will be a big jump in costs every time you change era... and that would be weird & bad. If you want later units to be more expensive then I think it would be better to just change the cost of individual types of units, so that, for example, tanks would cost more than archers - rather than the cost of archers going up in later eras.)
--


Essentially all of your points are areas where the AI could potentially be improved -- but none of it is really easy.

AI does not understand that war is too late and it is often not too late if you can take the entire country or erase several crucial cities. It is quick to compensate with massive number of hammers/sceinece/money you will get in return and war barely slows down your or AI development. You can not get behind in research or anything. It is cheap to create units and maintain them. Besides there are usually as many defending units as atacking since you and AI are both aware of army sizes. This drives total in game units number to the sky . There is no point keeping 50 units stacks. The AI built so many on the screenshot because it is smart. It is the safest and the most advantageous strategy. If you build units the other team also busy building them so you cannot get behind in research or anything.

I tested with many games the proposed change where unit maint cost is increased linearly with era (+ inflation). This is has some realism to it because with more modern units life infrastructure is improved and costs more. It does not matter that much if it is an archer or rifleman. This change has really nerfed AI armies down. Some leaders still create armies larger than free unit support threshold, but they try to stay within limits. They realize the impact of the cost, which is great. I also decreased unit upgrade cost to encourage upgrades. It would be smart from AI to disband some units after war, but it does not do it. Overall the game runs faster and interesting to watch because the largest empire does not win as often.
It is also important to have this by era and not turn number (like inflation cost) because every player is affected differently depending on its advancement level. This has great rebalancing effect. The penalties you mention are way to weak to take into account in the late game.

@stingo: I too have been frustrated at the number of units in play in the late game, and have wondered if there was a way to reduce it while still maintaining balance. I suppose one possibility would be to make late game units more expensive in hammers but also more powerful. Classic age wars are a lot of fun because on a standard map you might have 20-30 units in your main assault stack. By the Napoleonic era you might have 40-60 in your stack, so it's already becoming somewhat cumbersome. In the late game, however, when playing with jet fighters, carriers, modern armour, gunships, etc, nations will have 200, 300, maybe even 400 unit militaries. It just becomes a chore and the game is less fun.

The idea of making unit hummer costs does not work though. What it does it just slows down building of city improvements and in the end research time and everything is affected. AI does not realize relative hummer costs of building units. Besides there is always number of units on free support and slowing down unit building has negative effect for weaker civs making them even weaker. It is also nearly impossible to build units quickly to defend and breaks balance with chopping and drafting.

 

[noto2]

I guess I'd just rather micromanage fewer units, but perhaps that's just me.

 

[rkade8583]

For the Dun, culture. It keeps the original intent without unbalancing.

 

[Charles555nc]

Any updates? Life or otherwise

 

[noto2]

"the idea of making unit hummer costs does not work though. What it does it just slows down building of city improvements and in the end research time and everything is affected. AI does not realize relative hummer costs of building units. Besides there is always number of units on free support and slowing down unit building has negative effect for weaker civs making them even weaker. It is also nearly impossible to build units quickly to defend and breaks balance with chopping and drafting."

Well you could limit the changes to non-draftable units, for example. You could boost the strength and hammer cost of tanks, all planes, marines, etc. I agree with you that upkeep costs become nearly irrelevant by the late game. In the early game military upkeep has an impact. If you sit there building sword after catapult after axe, turn after turn, your economy slowly grinds to a halt as unit upkeep gobbles up a significant portion of your budget. But in the late game it seems I can pump out literally hundreds of units without even noticing upkeep costs.

 

[Axios]

For the Dun, culture. It keeps the original intent without unbalancing.

I quote.

 

[Charles555nc]

I wouldnt be against the bonus exp for melee units, just not on the celtic warriors because it might be overpowered. And add the culture in, is ok by me.

 

[noto2]

the problem with the celtic UU and UB is that it's geography based. So is the jaguar, but forests are very common in the early game - when it appears. Also, the jaguar has other advantages, such as being cheap and resource independent. Even still, the jag is routinely rated as one of the worst UU's in all those polls conducted.

The Celts have it even worse. If playing the Celts and you wind up in very hilly land, well, you'll be glad you're Celtic! All you archers and GW's get major bonuses for fighting on hills and anyone would find it nearly impossible to fight you on your home turf. But how often does this happen? Aside from playing a global highlands map, it just practically never happens. The Dun and the GW are great if you have cities on hills. Those cities will be invincible until the medieval age. But how likely is it that you can found every border city on a hill? Basically impossible. So that is the problem with the Celts. Their entire concept is geared towards hills. They'd be overpowered on a global highlands map, and otherwise are underpowered. I think either the UU or the UB should be changed so that it has nothing to do with hills. One of them being hill dependent is enough, don't make them both. So if we're leaving the GW as is, then make the dun have nothing to do with the guerilla promotion, or visa versa. That's my two cents.

 

[rkade8583]

I quote.

Huzzah! My first!

 

[Charles555nc]

*Robotic voice*

Need updates....

Processing...

 

[karadoc]

the problem with the celtic UU and UB is that it's geography based. So is the jaguar, but forests are very common in the early game - when it appears. Also, the jaguar has other advantages, such as being cheap and resource independent. Even still, the jag is routinely rated as one of the worst UU's in all those polls conducted.

The Celts have it even worse. If playing the Celts and you wind up in very hilly land, well, you'll be glad you're Celtic! All you archers and GW's get major bonuses for fighting on hills and anyone would find it nearly impossible to fight you on your home turf. But how often does this happen? Aside from playing a global highlands map, it just practically never happens. The Dun and the GW are great if you have cities on hills. Those cities will be invincible until the medieval age. But how likely is it that you can found every border city on a hill? Basically impossible. So that is the problem with the Celts. Their entire concept is geared towards hills. They'd be overpowered on a global highlands map, and otherwise are underpowered. I think either the UU or the UB should be changed so that it has nothing to do with hills. One of them being hill dependent is enough, don't make them both. So if we're leaving the GW as is, then make the dun have nothing to do with the guerilla promotion, or visa versa. That's my two cents.

I agree with this. In general, I think it would be better if the UB and UU didn't both rely on the same kind of situational map features.

 

[SyntaxError]

Yay! Thanks for the update!

 

[Charles555nc]

1.37 small bug...

When building 3 missionaries...I go to another city to build a 4th missionary, and highlighted in red it says "needs a christian monastery" instead of something like "missionary capacity reached". (this is under organized religion as well)

Suggestion/Idea

I notice the ai doesnt add great people to cities, except in rare instances like great artists for culture wins, and maybe Ive seen a single great merchant or great engineer. Im not including great generals in this distinction. Getting metal casting or theology instantly, with a great person, especially early game, is much better than the option of adding him to the city. I would like the choice to be more even.

My idea: A great person (but not great generals) added to a city now also provides +1 happiness. If this is considered too strong in multiple great people, then maybe hard limit happiness to +1 through this method.

I love you Karadoc! An update! I was so bored. Could you post that link again for people who want to donate?