Can we do anything to make Firaxis devote more resources to improving AI?

[Krajzen]

Title is self-explanatory.

I think a lot of people, maybe even a majority on these forums, consider terrible AI to be the main weakness of Civ6. Meawhile we do never get even a single enouraging word from devs on this issue, they simply avoid this "politically incorrect" problem.

Which is ruining challenge of playing the game and enjoyment of many. Can we do anything to express our collective discontent with incompetence of AI?

Personally I have long ago lost faith than Firaxis will suddenly significantly improve AI in some new expansion or iteration of the franchise, at least under current leadership (damn I fantasized about Ed Beach stopping being lead dev if his replacement would devote more resources to the goddamn AI). They are allowed to run away with this absurd incompetence for way too long.

And please do not justify them, there are aleady simple mods which improve AI a bit (mods which should be a part of the vanilla). In civ5, once it got its dll code released for modders, we got amazing mods such as community patch or vox populi, which GREATLY IMPROVED AI, the difference was STELLAR.

So it can be done, why Firaxis is not doing it?
Because customers allow it.

 

[stealth_nsk]

I think people on such forums are very small minority of civ players. And I believe if we'd ask all players, the problems with AI would look to be a very small issue.

The main thing here is - what to consider a good AI:
- For fans a very big part of AI is immersion. Soif AI looks stupid, it takes out a lot of game fun. That makes AI bad looking, unless the amount of work on AI is 20-30 times more it's now (see vox populi). That would never pay out.
- For casual players (the word is probably wrong as it implies they are weaker players, which is not the case, they just aren't fans) it's important for AI is to provide challenge and it does this more or less ok. Yes, Civ6 Deity is probaly an easiest Deity since Civ3, but this could be solved by adding a couple of difficulty levels.
- For developers themselves it's imprortant for AI to highlight all gameplay areas for players and thus things like not enough aircraft usage are likely to be seen improved as not enough aircraft usage for AI means no reasons to use anti-aircraft for human players, etc.

 

[Ryika]

Who's "we"?

What you as an individual can do is really simple, stop spending money on the game and give it a bad rating because of its terrible AI. Let the world know that it's an important issue for you and follow your words with actions.

But "we" as a community? Not much, as most people outside of the public discourse don't seem to care too much, and as long as most people don't care about the bad AI, or are willing to put up with it, there is little incentive for Firaxis to care.

Honestly, the main problem here is the lack of competition, there just isn't any game that even tries to compete with Firaxis in the "run through history"-type of 4x genre.

 

[Siptah]

Honestly, the main problem here is the lack of competition, there just isn't any game that even tries to compete with Firaxis in the "run through history"-type of 4x genre.

I think FreeCiv is still around, but jumped off that long ago...

 

[GoodSarmatian]

I think the only thing anyone can theoretically do is get rich, buy a lot of Take2 stock and order Firaxis ti make a better AI.

Honestly, the main problem here is the lack of competition, there just isn't any game that even tries to compete with Firaxis in the "run through history"-type of 4x genre.

I've been wondering for a while if it would even be possible to make a Civilization game in all but name. Obviously all the civs, leaders, units etc. are public domain and many of the game concepts are de rigeur for 4x games, but you'd have to change a lot of things just to look different and not be accused of plagiarism.
Then there's the market saturation problem. How many historical 4x games can be commercially successful at the same time ? If a new one is better than Civ and can prevail against brand loyalty and marketing, Civ 7,8 or 9 won't sell enough for Take2 to keep the series alive and we'll just have a new monopolist.

 

[Disgustipated]

Probably not, since we continue to buy the game despite whatever state the AI is in. Perhaps we could convince them to release the code source for the AI so we can fix it ourselves.

I actually don't have a lot of complaints about the AI. The only issues I want to see fixed are the lack of upgrading of obsolete units, and the AI not using certain units like air units and siege towers/battering rams. Only the very high levels does the AI use air units. Where as in past civ games middle levels the AI used them. I still think the cost of aerodromes and hangars need to be reduced significantly.

 

[LastSword]

I am not so sure about GREATLY and STELLAR, because back in the days I was testing every/most dll claiming to improve AI - still it was rather improvement. The issue is a bit more complicated.

I am not sure civ is designed to be a strategy game, it is more a sandbox (look at those shiny cities and wonders!) with AI as background. Which is quite a shame for people that except not to be the only actor on stage. I am writing about single player, eventually multi-cooperative.

I was also interested in R&F after hearing that Ed "AI is fine" Beach is less involved, but after watching part of first live-stream I have no positive exceptations. There will be no fall (for human players at least), but more power creep and filler mechanics (clicky, clicky, sixty bucky). Isn't it interesting that civ which has a disabled access to game mechanic is doing significantly better than others? Same with human who can simply ignore unneeded stuff, while AI is coded to develop in all areas, which leads to quicker catch up. In civ5, AI with proper flavours could become run-away, but flavours are gone (mixed feelings about it).
Anyway, I remember players better than me writing that BNW reduced difficulty by two levels (in comparision to GK) so we can even except R&F to make AI worser, which may be actually what we need to happen.

There is a lot to write about, timings, catching up as only challenge, AI being a sack of gold for human, is good AI fun at all (imagine situation, in which noone can conquer each other, sitting in own counter till one civ acquire [let's say] nukes and manages to win), hardcoded AI and mods, agendas are not a path to follow, game mechanics should be AI friendly. Blah, blah.

And finally, would it be possible to advertise expansion with "AI overhaul" theme + 9 new civs. Something with "mind" in the title, best strategy experience ever in civ and so on... I am not sure the majority of community would like it as much as I do. I believe the last expansion (major patch is not enough) should be spend on polishing product and I would support it with cash. Modders cannot handle it. Back to economy, how would they sell civ7 without claiming that "Civilization VI's oft-maligned AI is no more!"? Joke.
Or can they handle decent AI improvements + some new content?

I would say that with AI bonuses there is a need for two other options: AI competitiveness (let's say current stage would be the worst level - this one requires the most work) and Attitude towards Human (neutral - doesn't treat human any different than other players, friendly, harsh - the main reason for AI player to exist is to eleminate human player). Because there is a need of AI spectrum behaviour to meet the exceptations of the greatest amount of players.

 

[Ryika]

I've been wondering for a while if it would even be possible to make a Civilization game in all but name. Obviously all the civs, leaders, units etc. are public domain and many of the game concepts are de rigeur for 4x games, but you'd have to change a lot of things just to look different and not be accused of plagiarism.
Then there's the market saturation problem. How many historical 4x games can be commercially successful at the same time ? If a new one is better than Civ and can prevail against brand loyalty and marketing, Civ 7,8 or 9 won't sell enough for Take2 to keep the series alive and we'll just have a new monopolist.

Well, Civ VI is currently competing with (and when it comes to player numbers losing to) Civ V, but Firaxis is still pushing out lots of DLC, so I would assume that there's enough money in the market to sustain 2 historical 4x. I think looking at the many SciFi-4x that managed to be, more or less, commercially viable in the past, also shows that the 4x market in general can be a very good place to make money.

The simple reality is (or at least that's what it seems to be) that Civ is full of fluff that is just there to make the game look "flashy" but not really needed for a good gaming experience. The tons of music that was created just for the game, the many models of stuff on the map, custom, very detail-rich animations for the leaders... all of that probably eats very much into the budget. A 4x that cuts back on those things and focuses on gameplay and AI could probably be very profitable, after all people still play Civ V, even with leader-mods that in 99.95% of all cases come with still images as leader screens.

___

Not sure how to transition here, but the problem with AI in Civ is also not only that it's bad, but that the rest of the game isn't really designed with the AI in mind. That's especially true for the movement system, and district management, and tile-management, etc. - the AI could be a lot better in relation to the player if the game was more designed towards long-term strategies, and less towards micro-decisions, and I feel like Firaxis has never really attempted to explore how you could combine fun gameplay with mechanics that are manageable for the AI.

 

[Morningcalm]

The AI at present do not aid in immersion (weird agenda reactions, hypocrisy re: army units in borders, code-bugged trade deal offers which if reofffered by you are rejected, weird hypocrisy regarding warmonger penalties, esp. if you take cities of an enemy that is attacking them).

The AI also does not perform competently to threaten the human, with the possible exception of Kongo (though they do can be rolled over militarily with ease at most difficulty levels).

 

[leandrombraz]

Firaxis is quite aware that people are discontent with the AI, as a matter of fact there's a good chance that Ed Beach himself will read or at least take a glance at this thread. As Ed put it, they see the development of Civ VI as a marathon, not a sprint, so you can expect small improvements to be made with each patch, making more noise won't really change that.

I think the only thing anyone can theoretically do is get rich, buy a lot of Take2 stock and order Firaxis ti make a better AI.

Then realize that things are not that simple, find yourself in the position where you need to make a tough call and let Firaxis release a game that you know is less than ideal, hoping that Firaxis can improve it with time, which will set you as the next evil money guy who ruined a beloved franchise out of greed.

 

[Vandlys]

I think a lot of people, maybe even a majority on these forums, consider terrible AI to be the main weakness of Civ6. Meawhile we do never get even a single enouraging word from devs on this issue, they simply avoid this "politically incorrect" problem.

How does this problem connect, at all, to political correctness? 0.o

I mind the AI at points, and agree there is a lot of room for improvements. Why I, as a customer, allow it though? Because I like the game despite it, and I'd rather have Civ VI with bad AI than to stop playing it. Simple as that.
Why do you allow it?

 

[MooFreaky]

We could all spam Elon Musk to get his people to do the AI? Though then we would all lose over and over again...

 

[GoodSarmatian]

Well, Civ VI is currently competing with (and when it comes to player numbers losing to) Civ V, but Firaxis is still pushing out lots of DLC, so I would assume that there's enough money in the market to sustain 2 historical 4x. I think looking at the many SciFi-4x that managed to be, more or less, commercially viable in the past, also shows that the 4x market in general can be a very good place to make money.

That makes it even worse. A new game doesn't just have to compete with the current Civ, but also with the last entry which has had years of expansions and patches. The problem with the 4x game market isn't that there's no demand. It's that people will buy one game and play that for hundreds of hours and only get others that are sufficiently different and take place in another setting (I'm stll waiting for a decent new MoO like game with proper tactical combat).

The simple reality is (or at least that's what it seems to be) that Civ is full of fluff that is just there to make the game look "flashy" but not really needed for a good gaming experience. The tons of music that was created just for the game, the many models of stuff on the map, custom, very detail-rich animations for the leaders... all of that probably eats very much into the budget. A 4x that cuts back on those things and focuses on gameplay and AI could probably be very profitable, after all people still play Civ V, even with leader-mods that in 99.95% of all cases come with still images as leader screens.

That's a big issue with the games industry in general. It's too focused on marketing. You can't put good AI or writing in a trailer, so big studios double down on graphics and sound and lose control of their budgets. Firaxis is one of the least bad offenders in this regard.

Not sure how to transition here, but the problem with AI in Civ is also not only that it's bad, but that the rest of the game isn't really designed with the AI in mind. That's especially true for the movement system, and district management, and tile-management, etc. - the AI could be a lot better in relation to the player if the game was more designed towards long-term strategies, and less towards micro-decisions, and I feel like Firaxis has never really attempted to explore how you could combine fun gameplay with mechanics that are manageable for the AI.

That's true, but I like microdecisions in a game and don't want to see it dumbed down for the AI.
Just make the AI better.

 

[comatosedragon]

Yeah, as many have stated, the problem is many of us think we are 'the norm'. We are not. We are civ Fanatics. I doubt if we represent 5% of the player base.
Most 'average' players are not into dissecting every aspect of the game as we are.

 

[UncivilizedGuy]

I wouldn't mind seeing more improvement of the AI. I never really had any complaints of Civ4 and Civ5 but Civ 6 suffers to such a degree that it affects immersion. The AI just does not take a balanced approach at all. Many civs build an excessive number of units. They are terrible at building improvements in the early game. Sometimes I see Builders lingering around for a number of turns without doing anything. The game definitely needs improvement in this area.

 

[Temppu]

The thing with the AI is that in a complex and dynamic game such as Civ, developing a great AI requires loads of trial and error iterations. Its not like a single great developer can just come up with all the code required for building a machine that somewhat optimally plays a super complex game that a Civ is. With this is mind, because the process is certainly tedious, developers probably dont want to make it multiple times.

Every time the game is changed and tweaked, it affects the AI. As a result, it is somewhat unreasonable to devote tons of developing resources to optimize the AI as long as the game is still under development. These resources would be wasted to some extent, because new changes in the game would decrease the optimality of the decisions the AI does.

 

[bbbt]

We could all spam Elon Musk to get his people to do the AI? Though then we would all lose over and over again...

They'd have to make Civ an online game first, which probably wouldn't go over well (see the final SimCity). Though maybe that's what Civ 7 will be!

 

[Rosty K]

Got to figure out what data the 2K and/or Firaxis marketing teams use to make decisions about focusing on the fancy visuals. Then everyone should do the opposite to what causes them make those decisions Though that 'opposite' is probably 'not buy the game', which means we (those who want gameplay over graphics) are an overwhelmed minority. And I've said it before too, I think...

 

[Bechhold]

Only chance of that is if/when Paradox Interactive makes a Civilization game.

 

[PYITE]

I would love some more AI improvements, but considering they only budgeted for one AI engineer, I’m not holding by breath.

I wrote an AI mod that improves build orders for the AI. No tactical improvements (definitely needed but seems we are going to need the dll. for that), but it definitely makes the AI more challenging.

https://forums.civfanatics.com/resources/civ-flavor-deity.26561/

http://steamcommunity.com/sharedfiles/filedetails/?id=1206813450&searchtext=

 

[Bechhold]

I think the OP meant diplomatic wise as well. Example, AI not denouncing other AI for aggression, unequal trade deals, and so forth.