Is AI still brain dead?

[Karpius]

Yes let's hope so. Maybe if the community refused to buy more DLCs until the game was fixed 2K/Firaxis would take notice. Sadly, until that happens change is unlikely.

You mean Firaxis might not try to change things until they can make more money from it? Are you saying they are in it for the profit? You mean they would rather cater to those casual players who keep buying their product rather than those who know better? So sad.

 

[liv]

I think the casual player would benefit just as much as the so called "experts" It breaks the enjoyment of the game to watch AI settle one square from a river for no good reason. Or when they declare war on me without a military. Or when they send two knights to take a city with wall when four would probably have done it. It would just be a better experience overall if some of these things were dealt with.

 

[Spectrum]

Very disappointed with the game. I can forgive the cartoonished art, but the horrible AI is really anoying. No intention to buy the expansion or any other Firaxis game in the future. Not at release at least.

 

[Gorbles]

@Gorbles

what's the point in fluffy features if they just don't work properly?

Refer back to saying "lol stuff is already broken" is not an argument. If you want to discuss an actual thing in the actual game that isn't speculative hyperbole, I'm all ears. But regardless, taking people off of mechanics and putting them onto AI is also going to mean more unpolished mechanics as well. Because, well, less resource. So you're not actually asking for an improvement, even assuming the goalpost you just moved is accurate.

No, just no.

Civ5 got FANMADE community patch mod which made ai really good when compared with base game. I actually played with it and its ai was actually okay and actually much better than vanilla civ5 ai. It was not humanlike but nobody here wants genius or humanlike ai, just ai which doesn't look like a complete immersion-breaking idiot.

It is possible. It can be done. It is just Firaxis incompetence or unwillingness.

Civ6 didnt get such mod yet (due to lack of released dll code) but it already got simple basic ai upgrade mod (which just tweaks values without actual source code work) which is, indeed, markedly better than civ6 ai.

No, just no. We don't need neural networks and turing machines for civ 1upt ai to not be depressingly moronic.

I'm a games modder, and I find it curious that you think just because a mod team with zero constraints on their working hours can improve something that people put into the game in the first place, this means the people that put it into the game in the first place are unwilling or incompetent.

Really, all it shows me is that you have no idea what you're talking about.

You're saying "just because the people that can improve the source code Firaxis wrote in the first place, the people that wrote the source code which gives the modders the tools they need to make the mod in the first place are somehow bad". This is rather simplistic (if not outright offensive) on a number of levels. It is an order of magnitude easier to mod a game than it is to create it from scratch. When I say "AI programming", I mean "AI programming a complete solution from the ground-up". I don't mean "improving something somebody else already created from the ground-up".

It's a completely different thing. I'm not knocking the effort the people behind the Community Patch put in, either. You're the one assuming incompetence, and the only reason that springs to mind is that you simply don't understand the work involved, I'm afraid. There are no other explanations.

 

[salty mud]

@Gorbles you've assumed that people are asking for less staff working on features and more on AI. Where did you get this impression? I think most people would argue for more people in both departments. If anyone moved a goalpost, it's you.

It must be pretty lofty up there on your high horse. @krazjan wasn't implying what you were saying at all. He is rather lamenting Firaxis's either inability or unwillingness to improve the AI after the release of the game. No where did he criticise the actual making of the AI. His point stands true - if the community can improve the AI with the limited resources on offer, there's no reason Firaxis can't do it.

 

[hdbhdbvGFG541]

Refer back to saying "lol stuff is already broken" is not an argument. If you want to discuss an actual thing in the actual game that isn't speculative hyperbole, I'm all ears. But regardless, taking people off of mechanics and putting them onto AI is also going to mean more unpolished mechanics as well. Because, well, less resource. So you're not actually asking for an improvement, even assuming the goalpost you just moved is accurate.


I'm a games modder, and I find it curious that you think just because a mod team with zero constraints on their working hours can improve something that people put into the game in the first place, this means the people that put it into the game in the first place are unwilling or incompetent.

Really, all it shows me is that you have no idea what you're talking about.

You're saying "just because the people that can improve the source code Firaxis wrote in the first place, the people that wrote the source code which gives the modders the tools they need to make the mod in the first place are somehow bad". This is rather simplistic (if not outright offensive) on a number of levels. It is an order of magnitude easier to mod a game than it is to create it from scratch. When I say "AI programming", I mean "AI programming a complete solution from the ground-up". I don't mean "improving something somebody else already created from the ground-up".

It's a completely different thing. I'm not knocking the effort the people behind the Community Patch put in, either. You're the one assuming incompetence, and the only reason that springs to mind is that you simply don't understand the work involved, I'm afraid. There are no other explanations.

The first civ 6 ai mods were released after a week and took less then an day to be made.All they did was put some hard coded rules and fixed bugs leading to ai not utilizing some game options.That doesn't require a new ai team just common sense when it comes to programming.
And lo and behold most of the new official ai updates are hard coded rules and bug fixes.(don't attack without siege,use airplanes,etc)

Also if you are a game modder you do understand that game modders without access to the source have a harder time then the original devs who have access to source and can change stuff easily.
Look at stuff like bloodlines and stalker which had no sdk and modders have to work with buggy game code and they still accomplish amazing things.
Oh and i would love a example " zero constraints on their working hours" mod team.Most modders have secondary jobs or families,while firaxis gets paid to work on the game and spends a lot more time on the game then modders.That is why most mods never come out or take years,they have limited time and money.

 

[Thormodr]

The AI could be improved, yes, but I hardly think it is terrible.

It’ll get better as Civ VI continues it’s life cycle.

Just the same as every other iteration in the franchise.

 

[civtrader6]

I think the casual player would benefit just as much as the so called "experts" It breaks the enjoyment of the game to watch AI settle one square from a river for no good reason. Or when they declare war on me without a military. Or when they send two knights to take a city with wall when four would probably have done it. It would just be a better experience overall if some of these things were dealt with.

This settler issue is actually a bug that became prominent after the Nubia patch - where the algorithm would recommend subpar locations even when fresh water is available. So it's much easier to fix than complex AI behaviour, yet it still persists (albeit slightly less now).
Last time I pointed this issue and how it handicaps the AI, I got pretty much slammed with stories how AI can build an aqueduct ah...

 

[Gorbles]

@Gorbles you've assumed that people are asking for less staff working on features and more on AI. Where did you get this impression? I think most people would argue for more people in both departments. If anyone moved a goalpost, it's you.

It must be pretty lofty up there on your high horse. @krazjan wasn't implying what you were saying at all. He is rather lamenting Firaxis's either inability or unwillingness to improve the AI after the release of the game. No where did he criticise the actual making of the AI. His point stands true - if the community can improve the AI with the limited resources on offer, there's no reason Firaxis can't do it.

I did assume such because I literally described this in my original post. People replied to that. That was the initial scenario that I asked people to respond to.

More people is always an ideal situation, but real life doesn't work like that. There isn't infinite money in the pot (and if you want to criticise CEO wages, feel free to, but that's not going to magically make them hire more people). Their point is not true, because developers have a billion more constraints on their time than modders do. Developers implement features from scratch, they don't implement something on top of someone elses' work. I mean, I'm not expecting much considering your silly and aggressive attitude, but I recommend you actually read my posts before replying. I'm just repeating myself at this point.

The first civ 6 ai mods were released after a week and took less then an day to be made.All they did was put some hard coded rules and fixed bugs leading to ai not utilizing some game options.That doesn't require a new ai team just common sense when it comes to programming.
And lo and behold most of the new official ai updates are hard coded rules and bug fixes.(don't attack without siege,use airplanes,etc)

Also if you are a game modder you do understand that game modders without access to the source have a harder time then the original devs who have access to source and can change stuff easily.
Look at stuff like bloodlines and stalker which had no sdk and modders have to work with buggy game code and they still accomplish amazing things.
Oh and i would love a example " zero constraints on their working hours" mod team.Most modders have secondary jobs or families,while firaxis gets paid to work on the game and spends a lot more time on the game then modders.That is why most mods never come out or take years,they have limited time and money.

There's a few points here.

1. "common sense", actually, no, it requires time. If you don't have the time, you're not going to be able to put in the obvious polish that is required. You think developers are unaware of game issues? Yeesh.
2. It's not harder to do things without the source. It simply means you can do less things, in the proper way at least. If you want to talk about common sense in programming, hacking somethin via mod tools instead of doing something "easier" (i.e. properly) with access to the source is not common sense Not that I was ever arguing that, anyway.
3. Modders do accomplish amazing things. Never said they didn't.
4. Mods do not have deadlines, your argument is invalid. The ultimate resource on anything is time. A developer might work on something as a 40 (or 60, or 80) hour job, but a) they have many other responsibilities that you're not aware of, and b) the time they have to work on a feature is a lot shorter than a modder can take to do things.

 

[Acken]

Right now you should stick to older games, especially with mods, for a challenge.

The biggest problem with AI in general is that its an itterative process in a fixed environment to make it competent. If you keep changing the rules during devlopment and then through expansions youir AI will never catch up without strong dedication. First because the developer(s) is just drown in just coding the AI to use the system so lets forget about optimization. And then the guy is sent on another project after the last expansion.

Thats why modding usually work. The modder can just focus on the optimization and itterate pretty quickly with no one impacted by changes.
But for good modding you need to release the source of the AI code.

Finally for strategy games like civ. The other issue that make an AI seemingly bad is that the AI doesnt learn what is good and what is bad. While a player understands what is a good building, a good card etc an AI doesnt and just follows the rules. Thats where balance has a lot of impact on your perception of AI strength. Make the options the AI picks be competitive over those of a player and you will directly feel a stronger AI.
But balancing is an itterative process that also requires a strong understanding of the game which is not something developers are usually very good at either while fans will be able to quickly categorize what needs to be fixed and once again mod it.

 

[ThunderLizard2]

Very disappointed with the game. I can forgive the cartoonished art, but the horrible AI is really anoying

I rushed into buying Civ5 at release and feel I got burned - holding off on Civ6 (and needless to say Civ7) until it has a decent AI which could be a very long time.

The biggest problem with AI in general is that its an itterative process in a fixed environment to make it competent.

OK - but it's going on two years now and it seems there are still dumb mistakes the AI is making that modders have fixed using their limited tools in hours/days of effort. Maybe 2K/Firaxis should take a few modders on staff to do pre-release (or event post-release) fixes?

"common sense", actually, no, it requires time.

Time? We're not a launch plus 1 month. It's now 2018 and no fixes are in sight.

 

[Gorbles]

Time? We're not a launch plus 1 month. It's now 2018 and no fixes are in sight.

So now the goalposts have been moved to talking about a standard development cycle to analysing post-release support?

Alright. A simple answer would be "because other stuff needed fixing more".

 

[Ferocitus]

The AI could be improved, yes, but I hardly think it is terrible.

It’ll get better as Civ VI continues it’s life cycle.

Just the same as every other iteration in the franchise.

My thoughts too.
It also depends a lot on the type of game, the map size, and the
number of civs. AI mistakes by one civ hardly matter when you play
against 20-25 civs on a ludicrous size map at epic or marathon pace.

 

[liv]

How does it not matter that the AIs do know how to win the game?

 

[ThunderLizard2]

How does it not matter that the AIs do know how to win the game?

A challenging game is when the AI is competing against the player either to win or at least making it a challenge for the player to win. Otherwise what you have is a glorified SimCiv which is fine for some folks but not what the Civ franchise has been for those playing since Civ1

 

[hr_oskar]

A challenging game is when the AI is competing against the player either to win or at least making it a challenge for the player to win. Otherwise what you have is a glorified SimCiv which is fine for some folks but not what the Civ franchise has been for those playing since Civ1

Wait, I've been playing since Civ1 and don't think the AI was ever truly good. Actually I recall thinking that the Civ3 AI was a huge improvement to Civ1 and 2. Civ5 AI is frickin' Sun Tzu compared to Civ1 AI. Haven't played Civ6 yet but honestly... poor AI has been a complaint in every single-player game I ever played, including all the Civ games. Not saying it's fine, but we obviously still play all those games and enjoy them.

 

[liquidox]

The AI in Vox Populi is actually quite a challenge. Has Civ6 still not managed to match that after a year post-release?

 

[aimlessgun]

I think the casual player would benefit just as much as the so called "experts" It breaks the enjoyment of the game to watch AI settle one square from a river for no good reason. Or when they declare war on me without a military. Or when they send two knights to take a city with wall when four would probably have done it. It would just be a better experience overall if some of these things were dealt with.

Yes! I would like challenge but honestly my biggest beef is just how crappy it makes the game feel. I just watched Teddy approach an undefended Norse city with 6 units. Several turns later he hasn't even damaged the city and all his units are dead or dying. It's like a giant middle finger to anyone attempting get immersed in the game.

 

[Tech Osen]

Weirdly, I have the opposite problem where the AI will send a lot of force against me that would pose a challenge and I kill one unit, so it retreats (mild exaggeration for effect).

Often when they declare a joint war you just have to kill one or two units and they will beg for peace after the ten turn cooldown. Unfortunately you often have to send some units halfway across the world to actually find mobs to kill.

 

[Greasy Dave]

The issue for me isn't if the game has decent AI but more if the game challenges experienced players at higher difficulties.

I've played a lot of hours of Civ 5 and find it challenging even now. I play Immortal to be challenged but win , I play Deity to lose. The AI in Civ V is not extraordinary. It's as dumb as Civ 6. The balance of the game though is fairly well refined.

I came to Civ 6 just before Christmas and with no experience spanked the game at immortal without breaking a sweat on my first playthrough. The first 50-100 turns were interesting and offered the same level of intensity as I've come to expect from Civ 5 but after that it sank into dull repeated end turn button mashing (reminded me a lot of Civ BE ).

The reasons for this are many - but the biggest reason is how simple it was to win through war. The AI completely sucks at war after turn 100. This isn't an AI issue - because they set Civ 5 up to be challenging if I went domination - it's somewhere else in the design. (presumably, how the AI builds and upgrades its armies). This isn't about designing the game for "casual" players. Because Civ 5 clearly had both fanatics and casuals in its sights and it challenges me. It's about recognising where the problem is and trying to fix it.

I'm waiting to see what R and F brings to the table - but having read some of the devs remarks about the current state of the game, I'm not optimistic. I've read nothing that shows me that they're even aware that the game isn't challenging. ( I wonder who is playtesting the game and telling them that what they have is fine).

Currently I regret the purchase of Civ 6, so OP, don't bother. When R and F comes out, I will buy and play (the mechanics they are introducing interest me). But I'm expecting to play with a hand tied behind my back and only play peacefully to make the game at least somewhat challenging -I hope I'm wrong. But I'm not holding my breath.