Whatever they do I hope the AI is substantially upgraded

I think Civ VI still sits in a place where the AI is hurting the reception of the game even if the game obviously sells well. I think it hurts the reputation of Firaxis and 2k in the long run. Personally I don't need an AI that plays at a super advanced level, I'm a fan of AI's that sometimes does make the occasional mistake that the player can exploit. I like the character it gives to the AI. I don't have a problem with the occasional Settler trying to get to a city spot unguarded, I don't have a problem with them occasionally settling in a questionable spot loyalty-wise. Keyword is occasionally obviously.

But Civ VI still has quite a lot of areas where it's just broken. I just had a game where I played an island map, took over one of Australia's cities on the coast though I was pretty roughed up at the end of it. And, as it turned out, Australia just kept churning out galleys in the rest of his cities (this was early on in the game). He had a much bigger navy than I did and what did he do with it? Absolutely nothing. He kept building them and then they just sat there, not moving, not doing anything except allowing my pitiful archers sink them on by one. They could've easily taken their city back, no problem whatsoever. No strategy needed, just smashing into the city.
As a sidenote, he also kept building archers in the abovementioned city before I conquered it, which is good... but, he never let it stay in the city, he always moved the archer out so I could easily kill it. Again, what the hell? It had targets to shoot from within the city.

I don't think that's asking for "great AI", it's just asking for an AI that works. But it's broken, in many ways. And it is in my experience not uncommon at all to see this kind of strange behavoir.

 

I never did and never will understand Firaxis ability to seemingly ignore atrocious AI. How are they capable of fully enjoying their own game?

Just imagine them all working passionately on intricate new expansion mechanics, looking at them mishandled by stupid AI that provides no challenge, and being proud of themselves.

I'd like to enter their heads sometimes and see their thoughts.

 

a) their job is to implement X, enjoying playing with X would be a nice bonus but from a technical perspective, unless explicitly balancing the feel of a faction (which accounts for a minute percentage of the overall development effort for most games) they're not really going to be playing X.
b) "ignore" implies they have a choice about development priorities, instead of that coming from management, which comes with a complex hierarchy involving 2K.
c) "atrocious" means different things to different people. Technical things are often evaluated on "does work" vs. "doesn't work", and working badly unfortunately often falls into "does work".

I mean, yours is a weird comment. If something written by someone else that I have zero control over makes the mechanic I did write look bad, I'm still going to be proud of my work. Being proud of myself for work well done is for the work I did, and not the work anyone else did. That doesn't mean you can't feel let down about it, but we're kinda speculating on the personal feelings of however many people work at Firaxis here. Baselessly speculating, even.

 

You want one thing, I want another thing. The devs are coming your way. Fun for you, less fun for me.

 

I think a more competent AI would be beneficial for all players. The expert players would find a challenging game even at Prince level, how it is supposed to be: AI civs shouldn't need to have bonuses in order to compete with average human players...
On the others hand, the beginners can play with bonuses but against a smart AI they will learn faster how to play properly.
AI and good balance are the two aspects which impact on the gaming experience more than everything else.

 

Exactly! I agree with these 2 posts. For me the main problem with game difficulty is not that the AI gets better and plays better. They just stay stupid as before but gets a crazy ammount of advantages to start ahead, so it is harder for the human to catch up and it takes longer for us to be a threat to the AI - and even then it depends on the civ we are playing, if you play with Gilga you just smash the first AI civ you find with your War-Carts. I hate that the AI just play better because it starts with more troops, a free settler, a couple of worker, bonus beakers and culture, and more production. It doesn't play better, it cheats to play the game better. It is like we played the Sims with klapaucius (which we never did... )

 

I don't expect AI be "smart", but it should be OK and not be "stupid" and immersion breaking...

there are some simple rules that AI cuold follow, no matter what new things come with expansions or DLC,
like, just for example
combat:
- kill weakened enemy unit
- move and then attack with ranged units, if cannot attack, don't leave ranged unit within range of enemy melee for next turn
it involves using combined force: melee and ranged to support each other, being in a "group"
- conquer cities, even with some loss
- be better at trading /OK, it is not combat /
- use AIR FORCE...

 

I would say it's broken. In my last game, the AI did not use its units. It had plenty of units, but it did not attack me (defend itself). Impossible. I was hitting its units, and they were just standing or moving around. Even the Civ 1 AI did not have such a problem. LOL.

Really strange because it did attack me before on the same map before reloading the whole game. It even captured a city of mine.

 

in this whole conversation the keyword phrase was "sells good enough". why should firaxis spend more money if the AI is good enough to sell the game good enough. they will concentrate on additional content to sell more DLC/expac and then move on to CIV7. Hopefully after they tell/explain the modders how to improve their game.
good enough AI in this context is of course not to the players who are able to play at the highest difficulties and are looking for more challenges. Just check out the steam achievements to see how low the number of players is that can defeat the highest difficulties.
unfortunately the player community reading/writing in this forum is part of that little community and thus this AI perception feeds of itself and created this endless loop of they can do better. Of course they can do better and if left to the developpers I am sure they would want to , unfortunately the reality is that the managers dont see financial gain in it and would not allow the resources of their developpers used for this

 

I completely understand why Firaxis doesn't sink more money into improving AI, b/c of the financial incentive. I also understand that non-hardcore gamers don't care about the AI esp if they don't play that much. What I do find strange is that we hear during some livestreams that the developers are playing the game all the time, over and over again. I don't know how some one who plays this game every day and all day doesn't scream in frustration at the AI. Isn't it torture for them to play the game repeatedly and deal with bad AI?

 

That has always puzzled me also. People here have said that the majority of the developers either really don't play or they don't play on anything harder than King, but that doesn't explain it either. Others argue that they don't finish their games so they don't know how much it can lag at times. Regardless, it still baffles me.

I would love to watch on a livestream that either of the developers, and especially Carl, would get their butts kicked by the AI for once, and not necessarily only in combat. That would go a long ways. That would be so much fun to see.

 

https://www.cnn.com/2019/01/24/tech/deepmind-ai-starcraft/index.html

Perhaps now is the time to mention deepmind AI.

Perhaps one day developing strong AI's will be cheaper, but obviously the above example took a whole lot of resources to develop. And I don't think any of us want an AI that good. I know I don't. I would lose every single game.

 

I was just reading about this. Well, there are masochists among us I am sure!

 

Right, the AI is reliably bad. It's like knowing your opponent's "tell" in a game of poker.

I was thinking about when the AI builds an Archer in a city that I've surrounded... and moves it out of the city, basically committing suicide. There's simply no chance that's working as intended.

I wonder if AlphaStar can be tuned to different levels of skill and different tendencies? They describe using replays of human player's SCII games to train the AI's "neural network." If so, then maybe it's possible to have it mimic human players, not just play the most efficient game that it can. If it could do that, not only could it present different levels of difficulty, it could present different playstyles. It seems like DeepMind's goal was to create a StarCraft AI that wins, not to provide a human player with a fun game; it may be that the latter would present design challenges that the former doesn't have to address.

 

in order to make the perfect play you need to be able to tell the worth of each possible play and thus a difficulty level can be implemented. Then again at the moment not likely. Maybe if civilization was an esport or had a wider twitch audience then google would do it to showcase its strength.
I imagine the developpers mostly play to see if their implementation works and if they can actually win on a lower difficulty against the computer with the given bonuses and the AI guy/team are instructed to provide a challenge to the casual ( 90% ) lower level difficulty of the players.
Again these guys repeatedly state that their best guy is Carl and he plays on emperor and he is treated like the ultimate player to beat that difficulty which is very easy once you start to understand how the AI plays.
I usually play on emperor because i like building wonders ( personal preference not gameplay or minmaxing ). Otherwise they are usually only a waste of production that causes the AI to lose.
The way the computer approaches war is to defend first , so it will call back everything just because there was a barb scout from a camp 20 tiles away , which will cause the expert player to call it dumb. Even the deepmind AI for starcraft2 does it and it cost it the latest game. the human player abused it and made google look really dumb.

 

I just wonder what people's expectations are for the AI. If you want the AI to stop doing a particular thing that is dumb or you want the AI to start doing a specific action that is smarter, I think that is a realistic expectation to put on Firaxis. But if you are expecting a "super AI", that is not realistic. Plus, we need to realize that Firaxis is not interesting in making a "super AI" because their goal is not to create a super competitive game where the AI can crush you, their goal is to create "personalities" for the human to interact with. The AI is basically there to create characters in your civ story.

 

Maybe they are even worse than their AI in their game? Seriously. Watching the last few streams (offline), they overwhelmingly repeated and repeated that "Carl the destroyer of worlds blablabla" was at the driving seat of the gameplay, and presented him as some sort of guru of civ 6. Then, you watch the game being streamed.

I did not see almost any difference between "Carl" and the AI I use to confront in my games... maybe they are good enough developers (?), but are really mediocre at their own game, to the point that they cannot see the immense failure that the AI is.

And I'm being very, very polite here. Not speaking my complete mind at all. Not even close.

Back in the late 80's/early 90's, what you call "masochism" now was the whole force driving video games. We played BECAUSE it was hard, and the best games (as perceived by the communities) were the ones that were the hardest to beat. Granted, there was probably a lot of "cheating code" in there ala civ 1's "IF human is winning DoW" one liner... but nobody cared. If it was hard, it was good. If it was impossible, it was dope.

Now the instant gratification generations want immediate reward for almost no effort. That is the market now. That's why it will only get worse until something happens that changes everything, and with it the markets (all of them, not only gaming).

Sad, but true.

 

Because they are not looking at the AI, they are looking at the game more from a design point of view (is the world congress resolutions interesting? Is this wonder OP?) So they don't "see" the bad AI.

 

I think people still want a challenge. Even Ed Beach said "the game must be challenging" or something like that. The thing is the game is not challenging at all, so it seems that they've failed to provide the experience many people expected, esp. on higher difficulties. I think many people dropped Civ VI after a few games because there is no challenge.

Challenging games have been good because they are absorbing, intriguing, and respectful etc. In the late 80s one of the things games could offer was challenge. There was no great technology, graphics, animations. Games were simpler, but still that one feature could absorb people for so many hours.

To make the game challegning, the AI does not need to be super intelligent. Vox Populi level would satisfy me, to be honest. The thing is the AI in Civ VI does not work as supposed to. It's not only immersion breaking but discouraging. Why would I do things in the game (use the available features) when I do not have to. I can simply ignore it, and still win the game. Why would I want to prepare for wars and fight if the AI does not attack me at all? Even archers just wonder around not attacking my units. It is so demotivating to continue playing the game. I am just losing my interest. It is pointless. I don't need to plan strategy or use available options if things aren't working as supposed to. I can blindly hit the next turn button.

I just hope that one day some honorable people will fix Civ VI because there are things that I like about it more than in Civ V. I would donate them if necessary. Though, I know that Gazebo rejected himself to be donated.