I’m so glad Firaxis addressed the AI problem in Civ 7

[xphantomx]

This issue has been a baseline complaint of the series for awhile now, and rightfully so.

Most of it has been due to the iterative bloating of the system and mechanics. And while I personally enjoyed all the various choices, the micromanaging of units and concepts, the juggling of complex systems, the massaging of weaknesses of a Civ and leader into strength, etc…they all work to overload the AI into relative inaction,

I know some of the choices for Civ7 are getting some pushback…but narrowing the game into segmented gameplay goals of ages, only allowing AI to play civs whose strengths match each age (rather than fumbling a civ in early eras whose strengths are realized in later ages, and vice versa), removing workers, consolidating the throughput of the tech trees, simplifying district placement and resource improving, implementing fast travel “invisible” reinforcements, allowing armies to be bundled to commanders etc etc

I’m excited by all these minor/major changes. Not just because it allows me to experience Civ in a different way that Civ 5/6 offered, but because it all goes to address by biggest critique of Civ for a long time. Can’t wait for Feb!

 

[Disgustipated]

Limiting player choice does not equal improving the AI. I'm not totally against these changes, just sayin'.

 

[Somebody613]

Best game for AI: Rock-Scissors-Paper.
Can't botch THAT level of simplicity, right?

 

[Haig]

One of the designers said in an article that he can beat Civ 6 on deity but cant do it on Civ 7.

 

[Kaan Boztepe]

i also love how the end age crisis has the potential to solve the late game tedium

 

[greenOak]

One of the designers said in an article that he can beat Civ 6 on deity but cant do it on Civ 7.

This is encouraging to hear.

At the same time I disagree with simplifying the game just for the sake of having a more competitive AI. Also things like workers and tech trees were part of civ when the AI could still pose a threat to the player. The AI started sucking with the introduction of 1UPT because it couldn't handle tactical combat - and that problem will still exist in civ 7.

In civ 6, the AI couldn't scale it's economy into the late game but that had more to do with the AI's choice of which districts to build rather than where to place the districts.

 

[AlfaLeporis]

Can't beat because AI is smart, or can't beat because it got some crazy bonuses? These are completely different things, so I would be cautious.

 

[Xur]

Are we at bargaining stage already? (Five stages of grief )

 

[greenOak]

Can't beat because AI is smart, or can't beat because it got some crazy bonuses? These are completely different things, so I would be cautious.

I can't think of a single strategy game where the AI is competitive without bonuses. The problem in civ 6 was that, even with it's deity bonuses, the AI was unable to pose a threat to you beyond the first 50-100 turns.

 

[Siptah]

I can't think of a single strategy game where the AI is competitive without bonuses. The problem in civ 6 was that, even with it's deity bonuses, the AI was unable to pose a threat to you beyond the first 50-100 turns.

Ozymandias AI is quite good without bonuses. But it’s an asymmetric 4X game, so you can make the argument that choosing weaker civs gives the AI a bonus. Point still stands though.

The problem with bonuses à la civ deity is that they turn into bonuses for the player after some time.

 

[Lord Yanaek]

The problem with bonuses à la civ deity is that they turn into bonuses for the player after some time.

That!
Most of the usual deity "strategies" revolves around gaming the AI to turn their advantages into player advantages. I've recently watched a video of someone winning a real fast game on deity. His "strategy" was simple : sell every resources to the AI as quickly as possible for direct gold (not recurring payment) then denounce them so the AI will declare war (if that fails, declare himself) to get his resources back, rinse and repeat with a new AI the turn he meets them (so no negative impact from gieviances). Use all that gold to buy settlers (expanding faster than Deity AI!) and levy troops from a CS to fight of AI units in case they care to actually attack (AI is usually very bad at attacking you if you managed to get a war started before they were ready to actually attack so while this might sound like a bold move, it's actually relatively safe).

 

[Mr Jon of Cheam]

Have to wait and see on the AI, I'm not too optimistic on that front tbh but I do agree that the streamlining of various systems should be a significant help.

 

[Somebody613]

That!
Most of the usual deity "strategies" revolves around gaming the AI to turn their advantages into player advantages. I've recently watched a video of someone winning a real fast game on deity. His "strategy" was simple : sell every resources to the AI as quickly as possible for direct gold (not recurring payment) then denounce them so the AI will declare war (if that fails, declare himself) to get his resources back, rinse and repeat with a new AI the turn he meets them (so no negative impact from gieviances). Use all that gold to buy settlers (expanding faster than Deity AI!) and levy troops from a CS to fight of AI units in case they care to actually attack (AI is usually very bad at attacking you if you managed to get a war started before they were ready to actually attack so while this might sound like a bold move, it's actually relatively safe).

This reminds me of my own tactic (Civ 4) of "capturing a city, gifting it to a top civ, then recapturing it back next turn for free research, rinse and repeat until you catch up with everyone".
Makes everyone angry at you for treachery, but most of the time they are angry anyways (I'm talking about the highest difficulty), so it's not really much of a real change there.
Unless that's a mod thing (haven't played Vanilla for ages now), it's a really funny tactic that MAY even be applicable in later Civs (no idea how it works in Civ 6/7, lol).

 

[jbavon]

Does anyone know if the 'extra'/'late game' AI that are revealed in the age of exploration (or modern) are generated when the age starts or have been playing in the fog from the beginning?

I'm curious about this because it was apparent in Civ6, that the AI could not play the game at all. When you invaded them in the late game you would find lots of cities but mainly just a few farms and districts. Very little development of land/infrastructure. This was always very disappointing for me and ruined the illusion that I was playing against something interesting. In reality I was just playing against hidden bonuses. If the 'extra'/'late game reveal' AI in Civ7 are generated automatically at the start of the age I feel there is a chance that they could be very interesting opponents because they get a jump start with more efficient cities/infrastructure to compete against. It also makes it easier for the devs because they dont have to teach these AI to play the entire game. Instead they just have to play the game from the start of their era.

Edit: wording

 

[xphantomx]

Limiting player choice does not equal improving the AI. I'm not totally against these changes, just sayin'.

Gameplay design is all about limiting choices. Especially gameplay design involving the “board game strategy” genre.

Rulesets and restrictions are exactly what makes a particular strategic board game/card game elegant and effective.

Making the Civ game system more streamlined and focused can actually make the game more fun and the AI better, absolutely.

 

[Krajzen]

I am very happy regarding the change to make units move in groups led by a general. The majority of AI problems with 1upt involved logistic coordination of many units moving together in coherent manner to the front line, it could never escape from the mistake of sending individual units or too small groups to their suicidal deaths.

The ability to easily program AI to just pack threatening group of many units (in some balanced package) on one tile and send them to the frontline is going to greatly increase AIs ability to handle warfare, even if it won't be too bright after "unpacking" and actual combat.

I would also like to point out that the removal of individual unit promotions is very beneficial for the difficulty level. In previous two games human player was capable of slowly turning individual units into invincible elite forces, whereas AI was absolutely unable to keep them alive long enough. Here that human advantage is removed - only generals have promotions, applying to entire army at once, so you can actually expect AI to at least partially share xp bonus power with the human player.

 

[Disgustipated]

I forgot about the no unit promotions. I'll miss that satisfaction of getting my scout up to level 7. Not easy to do, but I've done it a few times.

I am really concerned about all these things limiting the human player. Like no more chopping forests. I suppose if the human player can do nothing interesting except press end turn, the AI might have a chance.

 

[Krajzen]

Ehh, I liked unit promotions but I don't feel much about them going away, it's one more layer of micro, which this time is going to be moved to the level of army commander. Perhaps one of the reasons I am fine with that is being generally far less interested in tactical combat than in civ5 where 1UPT was new and exciting.

I wouldn't jump straight to the hyperbole of doing nothing interesting civ7 seems to have my desired design philosophy of cutting down a lot of civ 6 layers upon layers of optimising micro in favour of more strategic options on macro level. Commander skill trees seem to be very interesting for me.