War in Civ7

Kayak06460

Chieftain
Joined
Jul 30, 2006
Messages
25
I enjoy playing conquer the world. Yet, Civ iterations have become worse at providing an ai that can wage war. If you’ve selected Domination as a victory goal then shouldn’t at least one ai civ be following that path? Remember when the ai could actually launch an invasion by sea? My hope is that Civ 7 addresses these issues and makes an ai that can actually build an army, launch an attack and for even peaceful civs to build an army for defense.
 
I hope they do get that fixed. I always turn OFF, diplo, culture and religious victories, yet the AI still go for religious VC.
(stupid waves of apostles...)

I would like to refer to the quote in my signature.
 
I enjoy playing conquer the world. Yet, Civ iterations have become worse at providing an ai that can wage war. If you’ve selected Domination as a victory goal then shouldn’t at least one ai civ be following that path? Remember when the ai could actually launch an invasion by sea? My hope is that Civ 7 addresses these issues and makes an ai that can actually build an army, launch an attack and for even peaceful civs to build an army for defense.
I miss Civ 4's victory condition Conquest, as well as the strategy involved with moving units on to ships, to cross to another continent. You had to have a real navy, to protect transports, and the AI did this. The current feature of embarkation, where every unit is a ship, is a dumbed down approach and makes having a navy mostly unnecessary.
 
The biggest obstacle to creating a challenging Domination / Conquest victory objective is that once you've conquered one civ, you've conquered them all. The rest of the game is just cleaning up. Having to organize a cross-ocean invasion may be tactically interesting enough to be enjoyable, but the outcome is not in doubt. The only ways I see to put the outcome in doubt is either:

1. Create enough penalties associated with conquering civ one that conquering civ two, civ three, etc. becomes harder, rather than easier. This has mostly proven true in the real world, but mechanics to reflect this in a 4x game would likely be unpopular, at least amongst a large portion of the game audience. It could create a real domination challenge, though, that would be welcome by some, if each new war became harder and harder to win.

2. Create an AI that (a) ruthlessly pursues aggression so that during the end game the player is facing a unified second half of the globe that's as powerful as they are, and (b) is tactically smart enough (or given enough AI-only bonuses) to potentially beat the player at the complex late game warfare. See Leyrann's quote of stealth_risk for why this would likely be unpopular and see any conversation about Civ 6 AI for why it's also almost certainly technically unfeasible without AI "cheating" (which itself would give birth to such a black hole of discontent on Steam/Reddit that all positive Civ 7 vides would be sucked into it, disappearing forever beyond the event horizon).

3. Load the AI civs up with enough of a starting bonus on Deity that surviving long enough to conquer civ one is enough of a challenge that players are content to then enjoy their laurels as they mop up the rest of the map. This is likely where we'll land.
 
The biggest obstacle to creating a challenging Domination / Conquest victory objective is that once you've conquered one civ, you've conquered them all. The rest of the game is just cleaning up. Having to organize a cross-ocean invasion may be tactically interesting enough to be enjoyable, but the outcome is not in doubt. The only ways I see to put the outcome in doubt is either:

1. Create enough penalties associated with conquering civ one that conquering civ two, civ three, etc. becomes harder, rather than easier. This has mostly proven true in the real world, but mechanics to reflect this in a 4x game would likely be unpopular, at least amongst a large portion of the game audience. It could create a real domination challenge, though, that would be welcome by some, if each new war became harder and harder to win.

2. Create an AI that (a) ruthlessly pursues aggression so that during the end game the player is facing a unified second half of the globe that's as powerful as they are, and (b) is tactically smart enough (or given enough AI-only bonuses) to potentially beat the player at the complex late game warfare. See Leyrann's quote of stealth_risk for why this would likely be unpopular and see any conversation about Civ 6 AI for why it's also almost certainly technically unfeasible without AI "cheating" (which itself would give birth to such a black hole of discontent on Steam/Reddit that all positive Civ 7 vides would be sucked into it, disappearing forever beyond the event horizon).

3. Load the AI civs up with enough of a starting bonus on Deity that surviving long enough to conquer civ one is enough of a challenge that players are content to then enjoy their laurels as they mop up the rest of the map. This is likely where we'll land.

I like this idea. There could be a choice in game setup…call it “Real World War “.
 
@Trav'ling Canuck , these options focus on the core of what Civ games have been, and the 4X genre overall is. The advice given to new Civ players all through the 2000s and most of the 2010s was, "land is power." The eXpand part of 4X is rewarded; more land means more citizens, more resources, more commerce. Civ5 provided some penalties for expansion; I started playing with Civ2 and played a LOT of Civ3. The penalties kept me from playing as much Civ5. The penalties included a redefinition of what the military VC consisted of, moving counter to the military victories in Civ1-4. If Civ7 does not stay true to 4X and "land is power," I probably will play less Civ7 after buying it and keep playing other games in the franchise.

In one of the oft-cited videos on this site, Soren Johnson talks about designing an AI that is "fun" versus one which tries hard to "win." I've lost games of Civ3 and Civ4 to AI players who both expanded better than I and who pursued a VC, e.g., diplomatic, better than I. A ruthless AI would not be fun for players new to the series/game, so it would need to be selectable or scalable somehow. AI who are competent in tactics has been seen in the series before. In Civ4, I had to deal with runaway AI (growth and conquering) and I did not play on the higher difficulties.

Option 3 (starting bonuses) has been the pattern in several games in the franchise. In Civ3, the AI at "Sid" level -- beyond Deity, LOL -- get two settlers and several military units on turn 1. More importantly, they get discounts (compared to the human player) in producing more units and discounts to trade techs with each other. I believe that Civ4 had similar starting bonuses for the highest difficulty levels. I agree that is probably where we will land.
 
Isn't 1UPT (or even xUPT) the biggest hurdle for a "competent" AI? SoDs are easy enough to understand, but even those seems hard for a Firaxis official AI so we needed BetterAI mod in civ4. AI is a complete misnomer though, should be NPC.

As for "Ruthless AI" seems to me to be an option toggle, or if a gradually more ruthless AI could be made as you progress up in difficulty.
 
Isn't 1UPT (or even xUPT) the biggest hurdle for a "competent" AI?

I don't know if it is the biggest hurdle for a good AI. But 1UPT did add a new hurdle by making moving units more difficult since units could block each other's path. This issue does not exist with xUPT since units can occupy the same tile so it makes moving units easier. Units can move through other units and don't block each other anymore. Also, stacks of doom did make things easier for the AI since the AI did not really need to do any tactics. The AI could just put a lot of units in a stack and there was a good chance it might win just from brute force. With 1UPT, the AI now had to actually be smart about where to place units and how to attack. So with stacks of doom, the AI could be "dumb" and it mattered less. With 1UPT, a dumb AI really sticks out more.

The truth is that Firaxis could make the AI great if they wanted to, especially now with the advances in machine learning. I just don't think it is a high priority. Polls show that most civ players don't play on the highest difficulty level. Most civ players are "casual". They want a little challenge but they want to win. Moreover, they don't care about playing optimally. They care more about the story of their civ as they play and want to have fun. Additionally, making a great AI would require a lot of time and effort which could be devoted to other parts of the game. It does not really make sense for Firaxis to devote a ton of time and effort into something that only a small percentage of the player base cares about. Better to focus on providing a good game experience for the majority of the player base.
 
bad AI is a feature of 1UPT, not a bug. If the AI was remotely competent in military the human player would need to move 20+ units in every defensive war every turn. It quickly becomes very boring if you have to spend 80% of your time on trivial unit moves (like getting reinforcement units to the frontline through chokepoints.
 
I don't know if it is the biggest hurdle for a good AI. But 1UPT did add a new hurdle by making moving units more difficult since units could block each other's path. This issue does not exist with xUPT since units can occupy the same tile so it makes moving units easier. Units can move through other units and don't block each other anymore. Also, stacks of doom did make things easier for the AI since the AI did not really need to do any tactics. The AI could just put a lot of units in a stack and there was a good chance it might win just from brute force. With 1UPT, the AI now had to actually be smart about where to place units and how to attack. So with stacks of doom, the AI could be "dumb" and it mattered less. With 1UPT, a dumb AI really sticks out more.
Its dumbness does stick out more when asked to do more complex stuff. And the "dumb AI" is why I label it NPC, especially when Firaxis focus more on "immersiveness" and gameplayability. I'm not saying that's wrong. I'm a civ4 monarch/emperor player myself (who struggles at high levels) who likes just that immersiveness and storytelling modes and abhor multiplayer warfests. I don't really want a perfect NPC, but it really breaks immersiveness when stuff like the carpet of doom situations happen.

The truth is that Firaxis could make the AI great if they wanted to, especially now with the advances in machine learning. I just don't think it is a high priority. Polls show that most civ players don't play on the highest difficulty level. Most civ players are "casual". They want a little challenge but they want to win. Moreover, they don't care about playing optimally. They care more about the story of their civ as they play and want to have fun. Additionally, making a great AI would require a lot of time and effort which could be devoted to other parts of the game. It does not really make sense for Firaxis to devote a ton of time and effort into something that only a small percentage of the player base cares about. Better to focus on providing a good game experience for the majority of the player base.
I'm sure they could given enough resources, but apart from the 1UPT-combat stuff, it's more often than not modders and experienced high-level players who can make more competitive NPCs. They probably know the game's mechanics better than the developers, and especially how to find synergy (or cheeze, depending on who you ask).
 
bad AI is a feature of 1UPT, not a bug. If the AI was remotely competent in military the human player would need to move 20+ units in every defensive war every turn. It quickly becomes very boring if you have to spend 80% of your time on trivial unit moves (like getting reinforcement units to the frontline through chokepoints.
That's a problem with 1UPT, and I highly doubt that's intentional, even if it "helps" 1UPT. The problem in that regard is that 1UPT breaks the strategic level the rest of the game is played at. I really hope we'd get a better tactical combat system than 1UPT, as it's the worst of both worlds IMO. A system for moving as (limited size due to supply) stacks and zoom-in to tactical map would also be easier to program for the NPC/AI.
 
A system for moving as (limited size due to supply) stacks and zoom-in to tactical map would also be easier to program for the NPC/AI.
Saying that it’d be “easier to program” just seems like pure speculation. Regardless, I would not want a separate tactical screen. Civ doesn’t need to devolve into some tactical combat simulator. Combat should remain on the map and remain abstract.
 
Anyone who thinks the AI can't properly handle 1UPT should play Old World because it is more than fine there and can actually challenge the player. The biggest difference between Civ and OW is that there is a lot more room actually move units around in OW so the AI doesn't have as many issues.
 
Last edited:
Yup, 1UPT is better the bigger the map, or rather the smaller the tile area.
 
I enjoy playing conquer the world. Yet, Civ iterations have become worse at providing an ai that can wage war. If you’ve selected Domination as a victory goal then shouldn’t at least one ai civ be following that path? Remember when the ai could actually launch an invasion by sea? My hope is that Civ 7 addresses these issues and makes an ai that can actually build an army, launch an attack and for even peaceful civs to build an army for defense.
Strongly agree. Civ VI, due to the Ai, often felt like SimCity to me. Due to the really fun designs (versus boring design of most Civ V civs) that was somewhat permissible but I truly wish historically expansive or conquest minded civs/leaders were effective in that role. I want to see an AI player begin to snowball because they’ve swallowed up their two nearest neighbors and now posses the tactical skill and required army to threaten to me and I need to act one way or another.

I honestly think my two biggest wishes would be

1) Expanded and more deep diplomatic options and capabilites
2) AI Capable of effective warfare (effective being crippling my economy at minimum but preferably able to capture cities from players).

I think the fears that new players would quit if an AI was able to actually threaten them are unfounded. When I was a total noob I lost territory and cities and lost games to aggresive civs in civ 4… I didn’t suddenly quit because of that.. it motivated me to get better and also frankly made the game a hell of a lot more immersive…

Even on Deity, the Ai often feel like actors designed to simulate having competition while 85% of the game is just internal min/maxing
 
Anyone who thinks the AI can't properly handle 1UPT should play Old World because it is more than fine there and can actually challenge the player. The biggest difference between Civ and OW is that there is a lot more room actually move units around in OW so the AI doesn't have as many issues.
Using mods that remove city strike and rebalance walls also helps so the AI doesn’t just lose half their army shuffling around your cities for 10 turns.
 
Blaming basically everything except the AI for the state of the AI isn't really the right way to go about it.
It's not 1UPT, it's not walls, it's not anything else. The AI just sucks, it's old and rusty, to be frank: we just need a new system with plenty of effort and nuance in its algorithms.

We don't blame the design of a fighting game for bad AI, we blame the code of the AI. Same goes for shooting games, RTS, board strategy games, racing games, basically anything with an AI.
 
We don't blame the design of a fighting game for bad AI, we blame the code of the AI. Same goes for shooting games, RTS, board strategy games, racing games, basically anything with an AI.
Soren Johnson, the lead developer of Civ4 and Old World, has made the point multiple times that the game needs to be designed for the AI and the player. You have to make a game the AI can effectively play in addition to designing a good AI.
 
Soren Johnson, the lead developer of Civ4 and Old World, has made the point multiple times that the game needs to be designed for the AI and the player. You have to make a game the AI can effectively play in addition to designing a good AI.

I'm not sure how that's relevant to the talking point. I was in agreement with your earlier comment - the fact that the game has 1UPT is not the sole guiding reason behind the bad AI.

In other words, there's no point shirking the bad AI onto other mechanics. The truth is the AI sucks because there is not enough effort behind it, and that's just how it is.

As you suggest, if an AI can play Old World perfectly well then why not Civilisation?
 
Top Bottom