Civ VII single player suffers from the same problems Civ VI did

[Disgustipated]

I did have the AI max out the economic legacy path in Modern age a couple days ago. I thought I was going to lose. But luckily for me they didn't seem to know how to use their great banker. I was really surprised they beat me because once I saw they were accumulating points I really focused on it. I have to admit, I'm kind of glad they didn't know how to use the great banker.

 

[frutiemax]

I've thought about this for a while and I realized there was no way for them to make a competent AI : the devs don't even know how to play the game at an expert level! They don't even know the optimal heuristics for an AI, it's just when the game has long been released when players have identified optimal plays. When that time comes, it's too late, the game is already in its maintenance phase so they won't spend a penny to improve the AI substantially.

I've given hope that a game studio such as Firaxis put effort on the important and not so flashy features such as multiplayer stability, engaging AI, good mod tools with complete documentation, etc. Firaxis is all about pleasing their casual player base that is content of passive city building with no engagement from the AI, without feeling cheated on / flip of a coin on deity AI.

 

[cyprusdog]

Thanks for all the replies, guys, which have been really insightful and a lot more matter-of-factly than my own admittedly frustrated rant.

 

[itix]

The problem here is that AI is not really AI. It is just a preprogrammed rulebook. They dont learn, they dont innovate.

Long time ago, i made an AI for an online 2D action game. It was quite good until someone figured out how to exploit it. I spent weeks on improving it, making it invincible, and finally everyone complained it is too good and felt artificial

Making a good AI is really hard.

 

[King Flevance]

I was actually playing a succession game in Civ4 with Blake's AI and trying to provide feedback to him. The AI got so strong, we couldn't beat them, their SoD where just too powerful to us. Let me try and find it

This is one of the succession games, Blake himself is part of it
https://forums.civfanatics.com/threads/acid_04-the-final-frontier.196533/

I loved the mod. If it still exists, I need to install it again. (I still love 4) It took me down in difficulty back when I played it on Emperor. I was sad to see him walk it back. Though I don't play as often as I used to and 4 can still give me a run for my money on Prince now with the basic AI. I would be happy to have Blake knock me down to Noble.

 

[Gori the Grey]

If I were lead designer on a Civ game, my very first brainstorming sessions with the team would be "how can we design the game so that AI can play it well?" I would take each sub-system--city-building, economics, culture, religion, warfare--and ask "how can this be a thing computers can do well?" The big thing computers can do well is calculate, so I think the answer to my questions would often be to make the choices in the game ones where a very precise calculation can ferret out a marginally preferable result (and the sum total of those would keep the AI players competitive).

I think they start with the question "what would be fun for players?" and only then ask "how can we teach the AI to do those things?"

If you could make it competitive, by building competitiveness into it at the base level, then you can dumb it down for the players who just want more of a sandbox experience.

 

[Leyrann]

If I were lead designer on a Civ game, my very first brainstorming sessions with the team would be "how can we design the game so that AI can play it well?" I would take each sub-system--city-building, economics, culture, religion, warfare--and ask "how can this be a thing computers can do well?" The big thing computers can do well is calculate, so I think the answer to my questions would often be to make the choices in the game ones where a very precise calculation can ferret out a marginally preferable result (and the sum total of those would keep the AI players competitive).

I think they start with the question "what would be fun for players?" and only then ask "how can we teach the AI to do those things?"

If you could make it competitive, by building competitiveness into it at the base level, then you can dumb it down for the players who just want more of a sandbox experience.

There are actually a few mechanics where I think they asked that precise question:
-City defenses (no more ranged strikes for cities, easier to take down fortifications)
-Diplomacy (no more abusable tables, you make/receive a proposal and there's only an accept/support/reject option)
-Settling (rivers are no more vital, terrain type isn't essential to understand, etc)
-Buildings (in Civ VI, you always had to choose what you would and wouldn't build, while in Civ VII, you can build practically everything that's available - which means the AI can't mess up, except maybe the order)
etc

 

[King Flevance]

-Settling (rivers are no more vital, terrain type isn't essential to understand, etc)

I am curious to understand your view on this. To me, I feel 7 makes rivers even more vital with the removal of aqueducts. Navigable rivers even have policies and powers greatly benefit your strategy. The Shawnee are even penalized for not settling near them. Plus rivers are needed for gristmills, sawpits, and sawmills.

Also, I have found it crucial to development strategies to understand the terrain for settling. A lot of vegetated or rough tiles means no food production for growth.

 

[James_Champagne]

I think my problem with the CIV VII AI (and I must point out that this is just going on the basis of the one game I played, where I still haven't finished it yet because around 75% into Exploration Era and with most of the world map exposed the game just became too much for my computer to handle) isn't even just the forward settling and the bad military, it was that, for whatever reason, I barely seemed even aware of the AI as other players. And this manifested in a lot of different ways. One of the main things I noticed was that the AI almost NEVER initiated contact with me, aside from a few times in the Antiquity Era where one civ made an alliance with me a few times. It always seemed like I was the one contacting them: asking for alliances, asking for open borders, things like that. And not only did they not once declare war on me (despite my having almost nothing in the way of a standing army or city defenses), they barely even ever went to war with each other. I didn't see one enemy city ever get captured, or razed, or anything.

I can't help but compare things to the Ai opponents in CIV VI, who feel more present and vital, I guess I'd say. I could always count on them to frequently pop up when they wanted to trade resources, or become friends, or request open borders, or just to leave a little comment about you and your actions (and how it reflected on their own individual agendas). I felt like they played a greater part in the story (and the fact they they were speaking directly to YOU rather than talking to your leader avatar also made it seem more personalized). I pretty much always knew where I stood with them, and thanks to things like diplomatic visibility and the whole gossip system, I always had a least some basic grasp or idea of what they were up to, and thus felt more engaged with them.

And I know I'm not the only person who feels this way, because the Aftermath review complained about the same thing: "A new diplomacy system has commoditised everything in the game, while at the same time removing the way Civs would often pop in just to say hi, exclaim at your progress or just threaten you, which makes the entire experience feel so much lonelier than previously. I never felt like I was one part of a wider human race in this game, more like one player tending to their little zen garden while everyone else tended theirs." I think that hit the nail on the head . . . it all just felt kind of lifeless I guess, with each civ just quietly working on their areas of land and ignoring the rest of the world.

 

[ThERat]

I tend to agree with the notion that giving modders access to the code would help a lot. Just look at Blake's AI for Civ4 or VP in Civ5. Whenever I hear players praise Civ5, it almost exclusively due to VP. For some time now, we have a modder who is working the Civ3 code having solved a lot of long standing issues (houseboat bug, the submarine war declaration bug, the fact that the AI wasn't able to use siege weapons of armies), resulting in a much better experience for me as a player. It can be scary when the AI suddenly appears with a stack of 24 infantry/artillery pairs, which can even take down your armies in one turn.
I am no sure what keeps Firaxis from doing the same for Civ6. To me, their business model has been to throw an unfinished product at us, knowing that fanatics would improve the game for them for free. Just look at the soon to be released patches, it seem as if it is taken entirely from customer feedback. Very convenient.

 

[pokiehl]

Just look at the soon to be released patches, it seem as if it is taken entirely from customer feedback. Very convenient.

You are criticizing the developers for listening to player feedback?

 

[StargazingDog]

WHAT is the point of introducing new and exciting systems if your opponents cant use them against you?

Once again, the debate on whether AI should present competition or fun. ALL Civiization games have made the AI to be fun. They are speed bumps on your way to victory. They do not play to win. If you need that to have fun, you need to find a different game.

This might be a controversial take, but I'm not fully convinced that the AI is actually an opponent in civ games.

It's not controversial at all. It is a stated goal of the developers.

 

[civlateintothenights]

Once again, the debate on whether AI should present competition or fun. ALL Civiization games have made the AI to be fun. They are speed bumps on your way to victory. They do not play to win. If you need that to have fun, you need to find a different game.


It's not controversial at all. It is a stated goal of the developers.

The AI shouldn't be incompetent thought - the perception of threat was there pre-Civ V. In Civ IV - the AI expanded well, provided a military challenge, and probably most importantly - they behaved how they were expected to behave diplomatically. Allies acted as allies - villains acted as villains. But it was still an extreme challenge to win on higher difficulty levels. No one was ever beating Deity on their first campaign. The sense of immersion was there. I've never felt remotely threatened by the AI in Civ V or VI - and diplomatically, the game just felt incredibly random at times. The sense of immersion was not always there.

 

[Leyrann]

I am curious to understand your view on this. To me, I feel 7 makes rivers even more vital with the removal of aqueducts. Navigable rivers even have policies and powers greatly benefit your strategy. The Shawnee are even penalized for not settling near them. Plus rivers are needed for gristmills, sawpits, and sawmills.

Also, I have found it crucial to development strategies to understand the terrain for settling. A lot of vegetated or rough tiles means no food production for growth.

Almost every settle location in the game has access to at least one river tile for a Gristmill and Sawmill (Saw Pit doesn't require a river). Navigable rivers are definitely useful, but are barely any different from coast in that manner. Shawnee are indeed the one exception, but this only applies to navigable rivers.

Thing is, everything rivers do is something that you also get if your city isn't on top of them or adjacent to them, with one exception - the 5 happiness from fresh water. Meanwhile, in Civ VI, the thing rivers gave you was an extra three housing, which was absolutely vital to get a city up and running in a reasonable timeframe (unless you were rich enough to just turn 1 buy a Granary, but that was more of a lategame thing).

In Civ VI, a settlement without a river was gimped (unless it was adjacent to a lake), and one that didn't have an Aqueduct spot was terrible. In Civ VII? You're losing out on some happiness. That's it. Might matter in a crisis if you're also over your settlement cap, but it probably won't.

As for understanding the terrain, it's again quite simple for me. Everything gets settled (unless it's too far away but that only lasts until I settle closer or the AI settles it). If there are good adjacency bonuses, the settlement becomes a city. If there are no good adjacency bonuses, it becomes a town. If it becomes a city, the urban yields are what matters. If it becomes a town, I build fishing boats and farms, and once it takes 10+ turns to grow (which is almost guaranteed to be before it runs out of fishing boats and farms to build), I specialize it.

 

[Gorbles]

Settler AI is one of the things targeted by the devs in the roadmap.

Anecdotally, somebody has been making AI changes in VII, and one of the things they fixed was settler AI. Apparently that alone made the AI a lot more competent (because they weren't wasting time on silly settling choices).

To me, their business model has been to throw an unfinished product at us, knowing that fanatics would improve the game for them for free.

Modders can always do more than what developers can do. Modders are not constrained by working hours, code reviews, other work-related processes, and so on. Nor does their continued existence as a modder rely on not being fired, or being made redundant. Because they're not employed!

(though it's very, very funny that you say this in the same post where you say the devs should give us fanatics the tools to improve the game for free)

You are criticizing the developers for listening to player feedback?

There is a theme at times, isn't there?

 

[protocol7]

Modders are constrained when we don’t get source access, though. They haven’t done that since 5 I believe.

 

[aelf]

It's not necessarily true that the AI can't fight, in my experience so far. I'm not sure if it's due to AI personality or an arcane confluence of factors, but I have once faced a Xerxes who spammed trebuchets. I never knew what menace such an army could be because I underestimated the pure ranged damage they could inflict.

Sure, the AI's tactics weren't anything exceptional, but I was only barely able to hold it off with a large treasury and suiciding expensive units. And in the end, the city was saved by the age ending.

You are criticizing the developers for listening to player feedback?

At this point, you can't win with some people. Just because they disagree with certain major design decisions, anything Firaxis does is given a negative spin by them.

 

[King Flevance]

Almost every settle location in the game has access to at least one river tile for a Gristmill and Sawmill (Saw Pit doesn't require a river). Navigable rivers are definitely useful, but are barely any different from coast in that manner. Shawnee are indeed the one exception, but this only applies to navigable rivers.

Thing is, everything rivers do is something that you also get if your city isn't on top of them or adjacent to them, with one exception - the 5 happiness from fresh water. Meanwhile, in Civ VI, the thing rivers gave you was an extra three housing, which was absolutely vital to get a city up and running in a reasonable timeframe (unless you were rich enough to just turn 1 buy a Granary, but that was more of a lategame thing).

In Civ VI, a settlement without a river was gimped (unless it was adjacent to a lake), and one that didn't have an Aqueduct spot was terrible. In Civ VII? You're losing out on some happiness. That's it. Might matter in a crisis if you're also over your settlement cap, but it probably won't.

As for understanding the terrain, it's again quite simple for me. Everything gets settled (unless it's too far away but that only lasts until I settle closer or the AI settles it). If there are good adjacency bonuses, the settlement becomes a city. If there are no good adjacency bonuses, it becomes a town. If it becomes a city, the urban yields are what matters. If it becomes a town, I build fishing boats and farms, and once it takes 10+ turns to grow (which is almost guaranteed to be before it runs out of fishing boats and farms to build), I specialize it.

That is an interesting view of it. I hadn't considered it from that lens. I can see now what you meant. (I recognized the saw pit error in-game about an hour after I posted it.)

I personally prefer maps that don't have many pockets of wasteland not viable to settle. I would actually prefer the whole map be viable to settle. I do agree though that removing some minor rivers would probably make it more interesting. Although I prefer game design that offers different types of choices rather than making obvious good choices just less plentiful. Otherwise start location and map generation wins the game for you via RNG lottery. I would still like aqueducts to return and/or adding a versatile system where founding a city with no fresh water gives you a different benefit like +5 gold or something. Perhaps even differentiate between salt water and fresh water. We also don't have tundra in this edition.

I certainly agree that terrain should be more dynamic. However, I will give Firaxis credit that adjacency boni make it a lot more interesting than in VI. I have seen people say that you just plop down buildings with the obvious placement, but I have found that good adjacency is uncommon enough to encourage city specialization and even suboptimal placements just to be able to fill a need. I don't think the system is perfect but I personally see these districts and map exploitation as a step forward compared to VI for my tastes. I will now be considering your view in future playthroughs too though, for sure.

 

[queenpea]

I find the AI to be very uneven, and I suspect that what I’m seeing is the result of a system of unbalanced civs (Maya for example is OP, so if Maya is controlled by AI then that player usually excels throughout further ages) and map generation (just finished a play through where Xerxes as Persia settled only three settlements the entire game. Xerxes spawned on an island off the coast of the main continent).

As a human player, it’s obvious that some civs are unbalanced. Combined with the choose your own leader mechanic, where many leaders are also OP, it creates a system where you are incentivized to take advantage of OP leader and civ combinations which enables an easy snowball to victory.

I never played Deity in VI, and now I am finding Deity in VII to be too easy.

 

[protocol7]

I also play on Deity, and the AI does very strange stuff. It has a lot of units so you have to be efficient attacking, which is great, but they will also do things like wander units out of the city and just stand there outside while you take it. They will send in commanders as cannon fodder first. They don’t seem to bother targeting damaged units. It can still be tough given their bonuses, but they aren’t playing in a coherent way.