"Why Civ 6's AI sucks"

[Agent327]

Put it this way. There will be no excuse in the future for the next civ series to have the same or worse AI than Civ6. From now on there is only one valid direction, an AI that gets better after every patch, expansion and series. 1UPT is here to stay and so they should be able to build up the quality of the AI there is no excuses any more. We know that 1UPT can be made to be a good enough AI (VP mod). So Firaxis are definitely in the spot light on AI development and they can no longer hide away from it by adding "features".

Criterion for a good AI:

• It is competitive for VC which scales with difficulty (as it is in VP mod).
• It plays a tactical game that is sufficient to look as if a prince level human player made the moves. The tactical moves are coordinated and make sense. The tactics don't have to be brilliant, just sufficiently convincing and aesthetically pleasing to watch (as it is in VP mod).
• The strategic AI must be able to conduct multiple fronts simultaneously both in defense and offense because this has already been proven successful in VP mod.
• The AI must not declare repetitive wars just for the sake of limiting better civs because that is tedious even if it is optimal strategy (a problem in VP mod).

Those seem like valid points. Posters may ignore, but Firaxis should not.

The AI may be better if we were happy to wait 10 minutes between turns but we are not.
The AI May be better is more people were happy losing but they are not.
The AI May be better is we were happy to pay £500 for the game, but we are not..

And these are not only not valid points, but quite immaterial. I'm not sure if they even quallify as 'points', seeing as they address nothing mentioned in the OP.

Firaxis should ignore.

I noticed someone mentioning 'these threads have been a drag'. You're neither required to read, nor to post on them.

 

[acluewithout]

I noticed someone mentioning 'these threads have been a drag'. You're neither required to read, nor to post on them.

Sure. But these threads do clog up the General Discussion Thread, and it's now almost impossible to actually find any of the (very few) good posts on this topic because there are just so many of these threads.

People of course can post as they wish (subject to forum rules) and you are right that I'm not required to read or post (and will endeavour to do neither in this thread going forward). But these threads are nevertheless very repetitive. I don't want to stop discussion, but perhaps people could consider whether there are existing threads that their posts could be added to rather than creating yet another AI thread...

 

[Gorbles]

I didn't realise you were the arbiter of what are or aren't valid points, @Agent327

I'm in no way speaking for Victoria (and I think they'd agree that we don't always agree), but I agree with their points here. I'll try to explain them as best as I can, in case that helps you understand why they're relevant:

Increasing the AI's ability to process things increases the time taken by the AI between each player turn. Like all things in a video game, a certain amount of time is spent on each action, and there are margins and thresholds for how long these can take before the game becomes unplayable (for the end user), or it has a knock-on effect on another part of the engine.

The middle point is the most vague, but also entirely fair. Perception is key. Losing a game can be frustrating. The AI can't be unbeatable, at which point scaling the AI for the most competent and dedicated of Civilisiation players I've ever met (I'm referencing this forum, and once again, praising this forum) means scaling up to Deity or perhaps past Deity in terms of player options, and which point it comes back to economic viability for the publisher.

Increasing the cost of the game (assuming the number of people who buys it stays the same, which is the unlikely aspect here) in order to better-fund development is a theoretical way you could increase the quality and attention spent on development (moreso than the ridiculous hours games developers are well-known for putting in), which is precisely why modders can craft the experiences they do (no upper bounds on time spent, no deadlines, and so on, and so forth).

 

[Victoria]

And these are not only not valid points, but quite immaterial. I'm not sure if they even quallify as 'points', seeing as they address nothing mentioned in the OP.

Oh wow
Posing the point that there should have been no version of civ until the AI was better than VI and that stacks of doom was key in all this, then stating 1UPT is nothing to do with it and neither is commercial considerations is odd.
To discard all these arguments and also state this is a serious discussion... seriously?
We are not allowed to bring any counter argument that “the OP” did not mention?

Is the point of the “The OP” that hexes combined with 1 UPT should have made the AI better? Well the number of computations has skyrocketed due to 1 UPT while hex introduction does not have a similar reduction so it is just obfuscation of the facts that computations have gone up massively but so has the cost of producing the game due to graphics. But no “the OP” discards this because they did not say it.

1UPT - it is not even 1 UPT, they can stack, it is just not stacks of doom. They added a tactical side to the game by limiting stacking and it has made the game less realistic, much, much harder for the AI to deal with. But this tactical side has made the game more fun and popular. No one said it simplified the game apart from the OP. So many assumptions by the OP that are rather questionable.

Bottom line is more money from getting rid of stacks of doom and more people happy playing the game so as said, 1UPT is here to stay but to state that Firaxis cannot release the game until the AI is ‘better’ than civ 4 is like a child telling their parent they cannot use an Ipad until they can use it better.

The has been so many threads people have got there now. Firaxis do what they can but there are many limiting factors and they are told they have to make money and so release a version that makes money. It just brain dead simple and these threads are just toy throwing. Just like my threads on England and hurricane damage are.

However at least @glider1 was still trying to make something positive with it so I am humbled.

I think we all know the whole idea of this thread was to get the title out there in another thread, the OP’s job is done, but Firaxis is desensitised to it and I am sure does not even bother reading them.

 

[Bibor]

The logical fallacy people are making here (Aristos included, sorry) is that somehow AI is the reason for why beating the game doesn't feel like a challenge.

The reason why it doesn't feel like a challenge is because the game is not designed in a way humans expect a game to be designed. Games that are interesting are interesting because they enable you to beat someone at their own game. Rules can be simple or complex, it's the other side that is providing the challenge. In case of chess, the board, the pieces and the rules are all pretty simple - complexity comes from the opponent. You're trying to beat each other at each other's game.

Intrest in this game comes not from a challenging opponent, but from discovery. You find new ways to know and then bend the rules, which is then multiplied by the randomness of the map. You're not figthing the AI (well you are, to an extent), but primarily you're fighting yourself. How to shave off a few turns and get to knights earlier, that kind of thing.

The problem with these types of games is they become boring real quick, once the novelty of the map (and perhaps leader) peters out. Since simply having a lot of cities is grating you a win anyway, after, say turn 100 or 150 you don't even with to finish the game, because you did win it.

Why repeat the same process over and over again, buliding the 15th campus in that ideal location, when you already did so 14 times in a row? Once you're sitting at 215 gold per turn, a new commercial hub granting a +3 adjacency becomes irrelevant. Once you have your 4th general and 25th army, why build yet another barracks?

These issues cannot be fixed by "fixing the AI", not even DeepMind would provide you with challenge you want, because the game is not designed to provide it. Take this game for what it is, an empire building simulator. If you want challenge, play games that are obviously designed to be a challenge (primarily multiplayer).

 

[Victoria]

If you want challenge, play games that are obviously designed to be a challenge (primarily multiplayer).

Or make your own targets, my best games have been when set my own victory conditions which often have rules along the way like you are not allowed to take a city from another civ but you are allowed to take a free city.

I found the MP environment rather unsatisfying for other reasons.

 

[ashendashin]

Or make your own targets, my best games have been when set my own victory conditions which often have rules along the way like you are not allowed to take a city from another civ but you are allowed to take a free city.

I found the MP environment rather unsatisfying for other reasons.

I don't have any interest in playing Civ as a sandbox game. Well, I did for some time until I got bored of it and stuck to playing with a friend. Did I play with him because I desperately wanted an at least equal opponent?
No. He sucks at Civ. But he made it interesting to play. Puzzles and sandboxes don't interest me near as much as dynamic environments where other players have a stake in the world. Civ is designed so the AI and humans play the same game. Everyone has an empire for you to consider beyond your own. It's an integral part of the series that has never been fully capitalized on by FXS.

VP did it for free. I never lost interest in upping the difficulty as the AI bonuses aren't insane compared to when I tried a Civ3 deity game for the hell of it when I was younger. Turns don't take forever despite the vastly superior tactical AI relative to vanilla (they did at one point though for some warmongers on huge maps. That was a new problem tho and I don't remember if it stuck around).
If you're content with FXS not being capable of anything like that then that's well enough. There are no real barriers beyond their company's position and management.

 

[Victoria]

I am very content in the idea that Firaxis has made a game I can play as a sandbox.
I have a problem playing the same old moves just so I can say I won which ultimately is where a highly difficult game ends up. Horses for courses.
I am also very happy that eventually someone will mod the game to be smarter but I feel I understand why this is not Firaxis. I do feel for the mod community here, it is at this stage that 7 should be starting to be made and the code For 6 released.

 

[Civinator]

The problem with these types of games is they become boring real quick, once the novelty of the map (and perhaps leader) peters out. Since simply having a lot of cities is grating you a win anyway, after, say turn 100 or 150 you don't even with to finish the game, because you did win it.

Why repeat the same process over and over again, buliding the 15th campus in that ideal location, when you already did so 14 times in a row? Once you're sitting at 215 gold per turn, a new commercial hub granting a +3 adjacency becomes irrelevant. Once you have your 4th general and 25th army, why build yet another barracks?

These issues cannot be fixed by "fixing the AI", not even DeepMind would provide you with challenge you want, because the game is not designed to provide it. Take this game for what it is, an empire building simulator.

Yesterday I played my favourite scenario Athendore in AOW 2 SM and came exactly to the point when I had conquered nearly all of the cities and let my enemies in the game only survive to see all places in the map of that scenario: The game was won and only the interest in exploring the last secrets of the map kept me playing.

And than it happened: So there shouldn´t be no risk of a rebellion in all my cities and with all my different troops and a booming economy, some random parts of my big and 'unbeatable' army revolted. They fought against me and I lost some bigger cities of my empire. I reloaded the game to see if I had done a mistake, especially ignoring some signs of a revolt of those troops and the people behind those troops that had revolted - but I had not ignored signs of a revolt. The next turn other troops of my army revolted. It seemed the revolts were randomized.

I´m thinking this could be a good setting to overcome the cliff in the game, that Bibor has pointed to. When your empire becomes too big and there seems to be no other relevant enemy on the map, coupled with a randomized factor, you suddenly become your own biggest enemy in the game - and the game becomes interesting again. The arguement for such a mechanism in the game could be a sudden loss of control over an oversized empire. I think this mechanism could be done to games of the Civ series, too without overstretching the demands about the AI. Concerning the AI, I think the AI opponents at least shouldn´t ignore the most evident situations of combat in a game. If this happens frequently, the game in my eyes is not good.

 

[Victoria]

, you suddenly become your own biggest enemy in the game

Yh, was thinking along the same lines this morning. Things like razing cities should cause an increase in barb camps in the area aggressive to you, that wide empires do get unhappy and corrupt.

 

[glider1]

The logical fallacy people are making here (Aristos included, sorry) is that somehow AI is the reason for why beating the game doesn't feel like a challenge.

The reason why it doesn't feel like a challenge is because the game is not designed in a way humans expect a game to be designed. Games that are interesting are interesting because they enable you to beat someone at their own game. Rules can be simple or complex, it's the other side that is providing the challenge. In case of chess, the board, the pieces and the rules are all pretty simple - complexity comes from the opponent. You're trying to beat each other at each other's game.

Intrest in this game comes not from a challenging opponent, but from discovery. You find new ways to know and then bend the rules, which is then multiplied by the randomness of the map. You're not figthing the AI (well you are, to an extent), but primarily you're fighting yourself. How to shave off a few turns and get to knights earlier, that kind of thing.

The problem with these types of games is they become boring real quick, once the novelty of the map (and perhaps leader) peters out. Since simply having a lot of cities is grating you a win anyway, after, say turn 100 or 150 you don't even with to finish the game, because you did win it.

Why repeat the same process over and over again, buliding the 15th campus in that ideal location, when you already did so 14 times in a row? Once you're sitting at 215 gold per turn, a new commercial hub granting a +3 adjacency becomes irrelevant. Once you have your 4th general and 25th army, why build yet another barracks?

These issues cannot be fixed by "fixing the AI", not even DeepMind would provide you with challenge you want, because the game is not designed to provide it. Take this game for what it is, an empire building simulator. If you want challenge, play games that are obviously designed to be a challenge (primarily multiplayer).

What you say is good but I think you are being a bit too black and white. We don't want the AI to provide the challenge. We want the AI to be aesthetically pleasing in the choices it makes then let bonuses do the rest. Firaxis focus on graphics aesthetics but neglect AI aesthetics too much. The AI should appear to the human to be making reasonable choices for its own benefit even if they are not optimal. Those choices should be down to the fine grain of tactical moves. VP for civ5 proves it can be done without radical changes to the engine. Currently the AI in Civ6 is the equivalent of an eye sore, a bug splat on another wise clean windscreen. That is why people are constantly complaining and they are right because they can't clean off the stain and it is frustrating!

 

[darko82]

I didn't realise you were the arbiter of what are or aren't valid points, @Agent327

I'm in no way speaking for Victoria (and I think they'd agree that we don't always agree), but I agree with their points here. I'll try to explain them as best as I can, in case that helps you understand why they're relevant:

Increasing the AI's ability to process things increases the time taken by the AI between each player turn. Like all things in a video game, a certain amount of time is spent on each action, and there are margins and thresholds for how long these can take before the game becomes unplayable (for the end user), or it has a knock-on effect on another part of the engine.

The middle point is the most vague, but also entirely fair. Perception is key. Losing a game can be frustrating. The AI can't be unbeatable, at which point scaling the AI for the most competent and dedicated of Civilisiation players I've ever met (I'm referencing this forum, and once again, praising this forum) means scaling up to Deity or perhaps past Deity in terms of player options, and which point it comes back to economic viability for the publisher.

Increasing the cost of the game (assuming the number of people who buys it stays the same, which is the unlikely aspect here) in order to better-fund development is a theoretical way you could increase the quality and attention spent on development (moreso than the ridiculous hours games developers are well-known for putting in), which is precisely why modders can craft the experiences they do (no upper bounds on time spent, no deadlines, and so on, and so forth).

And how does it refer to Vox Populi? The turn time etc. ??? OK, the AI might not still be perfect, but it is so much satisfactory in this mod - how huge difference it is to the vanilla or Civ VI AI? Unbelievably good. All those points are totally inrrevelant in terms of Vox Populi. And I totally disagree with @Victoria. Vox Popull denies those "points" and proves they are wrong.

 

[Mahi]

The AI should at least be able to do the most simple combat moves, like using archers probably in combat, using rivers, jungles, etc. to get better stats. The numbers of directions that can be taken shouldn't be a problem. There is always a tile that is equal, worse or better to move to, no matter how many choises you get. The input output decesion shouldnt be affected.
There may be a stastitic negative impact on better choises an AI can take but the AI's decesion making should still be the same.

 

[Victoria]

Vox Popull denies those "points" and proves they are wrong.

Are you aware of how much effort has gone into vox?
And when noobs go into the V forum and ask for advice on how to play, no one suggests they start playing with vox, it’s about learning on something easier.

I have played Vox, it’s a vast improvement... I have no issue with this. I have issue with people who like a very hard challenging wargame thinking that Firaxis have the same agenda as them. They do not. And the OP’s deity like approach to who can say what in a thread that is entitled “Why civ VI’s AI sucks” makes me question agenda.

I am saying Firaxis make something sellable and it’s up to the modders to make a harder version for the hard core. That creating another “AI sucks” thread when the last one rolls off the front page is just peeing into the wind

Also have you ever thought that the people that like a hard AI are mostly on other forums and the people that like an all round game with variety rather than repetition have embraced VI.

Yes the AI sucks, no one disagrees. No Firaxis is not going to ‘fix’ it. thank god we have modders that have done things like vox. You just have to wait for the dll and I have always supported that. These threads are not about debate, they are about agenda and I again ask the mods to move them to a sub forum because this single subject now has enough threads to support it. I have about as much chance as Firaxis making the AI near unbeatable.

 

[leif erikson]

Moderator Action: This thread has run its course as civility has ceased. Troll posts deleted. Thread Closed.
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889