Official announcement: Hot off the presses. Next Civ game in development!!!!!!!

[bbbt]

It's not that hard in principle - higher difficulties always choose the best option, while easier difficulties randomly take one of the best three options. Like a chess AI that doesn't (always) take the highest rated move. Vox Populi did it that way for a long time (though they got rid of it in newer releases, don't know why).

For the divide of NPC AI vs competitive player AI, there's a divide of behaviors towards the player. They actually tried to add it in a sloppy way with the "they hate us because we are winning modifier", but basically a competitive AI could turn around and nuke your capital after a game-long alliance just because you are ahead. A NPC AI would not.

They could add that as a "slider", the more you dial up competitive, the more the AI refuses to do behaviors that benefit the player and more it's attitude gets worse towards the player when they lead. Not sure how many players actually want a truly antagonistic AI though.

Also the way the win conditions are done in civ could lead to another variant - how much the AI targets the steps of a particular win condition (ie steps) vs just following flavors (gaining science)

 

[Eagle Pursuit]

The more nuance you give the AI decision-making, the longer the turn times.

 

[InsidiousMage]

Not to discount the need for challenge in a game but, at some point, players will get past the difficulty curve and if that's the only thing that makes the game interesting, what do you have left after that? Not much. The fact is that the AI can only provide so much challenge for any skilled enough player so there needs to be something else in the game. Difficulty alone does not make for a good or even interesting game.

 

[Kwami]

Because the AI is pretty bad. You're creating a false dichotomy. The overall incompetence in regards to even the most basic aspects of the game, like when the AI goes the entire game not fixing pillaged tiles which occurred in the Ancient Era, both makes the game less challenging and less immersive.

It's not that bad, though. It's capable of winning by turn 300 on Deity difficulty, which is probably just about right in terms of balancing a game where the player should usually win if he makes reasonable choices.

It's not perfect. Sometimes the AI does stupid things, like not fixing a tile. But overall, it works pretty well. And, the turn timers aren't ridiculous, even later in the game and on systems with less compute power (e.g. Switch, iOS, Android, older PCs). I can't say the same for Vox Populi.

 

[Krajzen]

AS far as AI goes, the key detail is focus: is the AI trying to "Build an Empire" or "Win the Game". This philosophy determines all underlying decisions about how the game is played, especially the diplomacy aspect.

If the AI is trying to win, it should disregard all past beneficial relations/agreements once it is clear you are in a position to pull ahead (which happens in board game sessions, btw.) Even if it means a mostly hopeless attack or severing mutually beneficial trade ties, they should all dogpile on the score leader at a certain point. This means there can never be long-term friendships, the AI will betray you once it is clear you have a strong likelihood of winning. It would be stupid AI to do anything else. At the same time, it can be miserably frustrating to be buddies with someone for 2,000 years and then out of the blue they hate your guts and all your peaceful play goes into the crapper and you're just grinding out military units.

In a simulation of building an empire, the calculus changes dramatically. There, the AI can be your buddy for life if you invest the time and energy in reaching that level. However, with proper planning you can get everyone to like you so much that you can just cruise on to a peaceful victory with little endgame drama. In Civ V the easiest win is to play a small, mostly peaceful empire with strong defensive positions, then let science/culture/gold flow in. By the time the game is winding down victory is a foregone conclusion.

Very insightful comment, honestly I have no idea how I'd solve this terrible dilemma as a game developer.

The essential problem here is the very notion of 'historical victory' which simply doesn't exist IRL, but this kind of game needs it still (and Humankind's attempt to revolutionize this aspect has been divisive to say the least ). Do Japan, Germany and US hate each other for having great scientific output? Well why the hell should they? Similarly, why should world's most culturally influential countries (whatever that means) rival each other? What does it even mean to achieve 'diplomatic victory' (that's the question I have always had). Don't get me started on the 'religious victory' (imagine it today, with half of the world being heavily secularised by modernity, and almost no country being recognized 'the leader of a religion').

Actually I have the idea.
1) Design the game so AIs aren't 'playing meta to win the board game' but 'trying to hit the optimal balance between safety, and power' which are relatively universal state goals
2) Make drama emerge from the game's mechanics for example
- regional competition for power and resources creates rivals willing to ally with the outsider against each other
- religions being actual source of great controversy
- industrialization really benefits from agressive colonial empires so they rival each other as they try to expand
- lategame finally has proper world wars and cold wars thanks to big alliances, partially because...
- ...we come back to civ5 Ideologies on steroids and make global liberals, marxists and nationalists hate each other guts just like IRL
3) Make the victory conditions depend on either a) hard to achieve cooperation b) hegemony which breeds 'natural' global opposition thanks to 1 and 2 c) Overall Achievements of Your People Across History

So maybe this way you have both immersive not infuriating AIs and interesting drama?

 

[Gedemon]

That makes it more difficult to understand.

I've probably read some of these points before. Could you just write some bullet points? I mean, I associate size and empire building with Civ 4 more than with Civ 6, because unpacking the cities makes me more focused on each city in Civ 6 (and managing a large empire eventuall becomes dull imo). But I don't think that's what you mean, or at least not just that.

That would require a whole thread. Your point here is part of it, among others.

The only thing I can really think of that makes Civ 4 less "gamey" than Civ 5 or 6 is the AI ostensibly is designed not to "win" the game. Other than that, I don't see how any Civ game can be credibly viewed as anything close to a simulation.

Indeed, or they are all "simulating" something anyway. I try to avoid "realism" and "simulation" vs "gamey" because it is confusing as the game is not, and can't, by it's scale, be a realistic simulation of Human's history.

But "building an Empire to stand the test of time" has always been my motivation to play civ, and that's a feeling, it's a bit similar of "suspension of disbelief", when you perfectly know that a movie or a game is not reality (Sci-Fi quick and easy to understand example: faster than light ships or sounds in space), but you don't mind as long as it stay in the limits of your own willingness to suspend disbelief.

That's something that will be different for each players, some will have that feeling immediately broken by the immortal leaders, others will not care until the 4th wall is broken by a leader saying he doesn't like us "because you're winning the game".

But usually it will be the sum of multiple small details that will make the difference.

The AI playing to "win the game" at the cost of diplomatic interactions, leaders agenda (not a bad idea in itself) making no sense sometime, gameplay limitations to prevent exploit, unstacking units, unstacking cities (ho, I need to reserve that tile for when I can build something 3 eras later...), forced micromanagement (personally I don't care if the UI use cards or lists, but the evolution of the UI has followed the gameplay changes: IMO it was easier to manage a the Empire level in civ4 than civ5/6, unless you use mods), AI not being able to use some units, smaller maps, etc...

All those are such details in my case.

 

[The_goggles_do_nothing]

If the AI is trying to win, it should disregard all past beneficial relations/agreements once it is clear you are in a position to pull ahead (which happens in board game sessions, btw.) Even if it means a mostly hopeless attack or severing mutually beneficial trade ties, they should all dogpile on the score leader at a certain point. This means there can never be long-term friendships, the AI will betray you once it is clear you have a strong likelihood of winning. It would be stupid AI to do anything else. At the same time, it can be miserably frustrating to be buddies with someone for 2,000 years and then out of the blue they hate your guts and all your peaceful play goes into the crapper and you're just grinding out military units.

There's a model for how to do this, well, almost do this. Civ 6 religion. There's a limited number of players that can pursue a religion victory. In civ6 if you don't get a religion you just play without out it and try a different victory. But if you changed the game so you could "capture" a religion and religious victory was the only VC it would "feel natural" that at some point everyone without a religion attack those with one.

 

[PiR]

If the AI is trying to win, it should disregard all past beneficial relations/agreements once it is clear you are in a position to pull ahead (which happens in board game sessions, btw.) Even if it means a mostly hopeless attack or severing mutually beneficial trade ties, they should all dogpile on the score leader at a certain point. This means there can never be long-term friendships, the AI will betray you once it is clear you have a strong likelihood of winning. It would be stupid AI to do anything else. At the same time, it can be miserably frustrating to be buddies with someone for 2,000 years and then out of the blue they hate your guts and all your peaceful play goes into the crapper and you're just grinding out military units.

This is thinking there is only one victory condition, but there are many.
You might be the lead on a science victory, and there might a friendly civ on the lead to a religious victory. Why become suddenly enemies with them? Makes no sense, just hurry up to win with science.

 

[Eagle Pursuit]

One of the things I'm eager to learn about Civ 7 is how they are going to do the Vikings. In 1-4, the Vikings were Vikings, and that was kind of silly because Viking is an occupation, not a cultural identity. You don't have a Plumber civ. With 5, they corrected this by making the Vikings Denmark. Good choice. In 6, it was Norway. Now they've hit the two main homelands for Viking raiders.

So, what to do for 7? Do they move Sweden from Renaissance to Medieval to make them the Vikings? Stay at Norway? Switch back to Denmark? Or do we begin to explore the various states established by Viking invaders? Iceland? Normandy? Kingdom of Dublin? Norman Sicily? The Danelaw?

My personal preferences would be Iceland or Normans.

 

[Leucarum]

For both vikings and celts in particular I'd love to see them rotate the civs represented... I'm sure there are other examples, it's a good way to highlight the various first nations too.

 

[Connor_CivFan]

I'd love to see a swap between the Viking Nations. I would be interested to see what they would do with a less viking orientated Norway/Denmark and vice versa with Sweden.

 

[Xandinho]

They can keep Norway or bring Denmark back, I don't care which one they decide. However, Sweden should continue to be represented as a Protestant power.

 

[Alexander's Hetaroi]

They can keep Norway or bring Denmark back, I don't care which one they decide. However, Sweden should continue to be represented as a Protestant power.

Is it too much to ask for Viking Norway, Kalmar Union Denmark under Margaret, and Protestant power Sweden?

If so I'd take Cnut the Great and Margaret for Denmark.

I also have a feeling for the "Celtic" civ we might at least get Ireland. Not sure if they'll do a pre-Medieval one like Gaul again or not.

 

[Eagle Pursuit]

I would love Ireland. I wonder if they would be open to the idea of Grace O'Malley as leader. Female pirate and minor noble? Big personality? Pen pals with Elizabeth I?

If that's a little too off-track, there's always Brian Boru.

 

[Takhisis]

If I had one feature I want to see in Civ 7 it's this:

An easy Build-a-Civ tool, using premade features.

You would choose the name, demonyms, leader's name, etc.
Pick an icon and color scheme
Pick an ability from a drop-down menu (for your leader too)
Pick a unit to be made unique
Pick its ability from a drop-down menu
pick a tech/civic for it to be unlocked
Name the unit
Pick a unit icon from a list
Pick a building/district or "new improvement" to be unique
Pick a trait for it from a menu
Pick an appearance for it from a menu
pick a tech/civic for it to be unlocked
Finally, an option to create a city list, or just name them as you go

So basically you want the civIII editor to make a comeback, but upgraded for individual civ-packs. Nice.

How exactly? Does not the cultural tradition and mastery of bowmanship and horse-riding made possible the conquests in the first place? How can your "cultural high point" exclude the culture itself?

In its original appearance in (again) Civ3, culture was meant to mirror soft power, to stop players from being warmongering barbarians. As present-day runs on the game continue to demonstrate, it didn't work that well, but when properly tuned (off-the-shelf Civ3 is a beta, but check out The Rood and the Dragon's win conditions) it actually does help players diversify play.

I'm not sure whether horsemanship and weapons training counts as such, except that in warrior cultures it is, well, their culture. It could replace or complement other buildings in expanding a civ's ‘cultural value’.

 

[InsidiousMage]

Or do we begin to explore the various states established by Viking invaders? Iceland? Normandy? Kingdom of Dublin? Norman Sicily? The Danelaw?

Depends on what you consider as a "civ." I would think that the Normans are too "small" for Firaxis to do on their own, and the same with Iceland honestly. Norman Sicily is way too separate from the Vikings for them to be considered a "Viking" civ. You could probably make the same argument from the regular Normans as well. I've seen plenty of people describe William as a "descendant of the Vikings" but I don't think I've ever seen anyone call him a Viking. I think not repeating yourself is important so I'm fine with Sweden being the "viking" civ in the next game.

That said, I think if Firaxis is willing to expand the idea of what a "civ" is the game, then you've got more options. If you expanded to include things that aren't something akin to a "state" then I think you could include peoples like the Normans as their own civ and even do something like Ragnar Lodbrok leading the Norse.

 

[Eagle Pursuit]

Depends on what you consider as a "civ." I would think that the Normans are too "small" for Firaxis to do on their own, and the same with Iceland honestly. Norman Sicily is way too separate from the Vikings for them to be considered a "Viking" civ. You could probably make the same argument from the regular Normans as well. I've seen plenty of people describe William as a "descendant of the Vikings" but I don't think I've ever seen anyone call him a Viking. I think not repeating yourself is important so I'm fine with Sweden being the "viking" civ in the next game.

That said, I think if Firaxis is willing to expand the idea of what a "civ" is the game, then you've got more options. If you expanded to include things that aren't something akin to a "state" then I think you could include peoples like the Normans as their own civ and even do something like Ragnar Lodbrok leading the Norse.

The Normans were definitely a state, albeit a duchy instead of a kingdom. The king of the Franks had very little control over the Norman duke, to the extent that William, Duke of Normandy decided to take over a whole kingdom on the other side of the channel and brought himself to rival the Frankish power.

 

[Leucarum]

I think the biggest problem with the Normans is that England is definitely going to appear, so you'd probably want a Norman civ that specifically leaves out the English portion of their history. There's plenty of historic space for that, the normans did get around, but disentangling the two could leave things a bit messy.

 

[queenpea]

I think the biggest problem with the Normans is that England is definitely going to appear, so you'd probably want a Norman civ that specifically leaves out the English portion of their history. There's plenty of historic space for that, the normans did get around, but disentangling the two could leave things a bit messy.

I would love a Civ "Brexit"...

Moderator Action: Let us please not get into current events/politics in the game threads. Thanks. leif

 

[InsidiousMage]

The Normans were definitely a state, albeit a duchy instead of a kingdom. The king of the Franks had very little control over the Norman duke

I feel like under this definition every duchy and most counties in France and Germany at the time would count as state, which feels like an overly broad definition. The lack of central control was a defining feature of Western and Central Europe at this time and I don't think most people would consider the Duchy of Aquitaine a state during the time period.