The AI isn't worse than in V, but VI is a much harder game for it to play

[Xyriach]

There is a possibility that if a lot of the player base has been using online play, then they may have geared it towards human vs human rather than investing in the AI. I'm not saying it's the case, but consider the COD franchise where the developers have invested less and less in the AI as they know that the majority of players use multiplayer anyway, so the AI doesn't make for a useful time investment.

With Civ, I'd find that fairly hard to believe, however, given the amount of criticism that Civ 5 had over it's AI at launch, I'd be surprised if it's not something that they've addressed. So perhaps there is an element of HvH in this early release that's being watched against what the actual release will reveal later this month?

Sadly, we're in to the connected era, where companies have a bad habit of releasing a game then fixing it later. As a 25 year old franchise, I'd hope that Civ weren't being that cheap.

 

[jekke]

I never seen a single person complaining about "bad AI" to actually define what "bad AI" or "good AI" is.

I, on the contrary, very often see posts where people first explicitly say which AI actions they consider dumb and then summarize it along the general 'AI sucks' lines.

 

[stealth_nsk]

I, on the contrary, very often see posts where people first explicitly say which AI actions they consider dumb and then summarize it along the general 'AI sucks' lines.

Well:
1. Considering AI dumb is quite weird itself, because it doesn't have and is not planned to have any form of intelligence.
2. That's generally the second and non-functional part of AI - the immersion. People complain about AI looking dumb - it has nothing to do with actual gameplay challenge.

 

[Ornen]

Well:
1. Considering AI dumb is quite weird itself, because it doesn't have and is not planned to have any form of intelligence.
2. That's generally the second and non-functional part of AI - the immersion. People complain about AI looking dumb - it has nothing to do with actual gameplay challenge.

It makes no difference to you? I'd much rather play against an AI that can defend its units and has some sense of how to attack and defend than an AI that just produces a massive carpet of units because of its production bonuses.

 

[SupremacyKing2]

Then make the mechanics more ai friendly.

I am against this if it means putting in mechanics that are uninteresting to the player. Remember that the primary purpose of a single player game is to give the human player an enjoyable experience. This means giving the player some challenges, some fun, some highs and lows etc. The game should be designed around the player experience. The AI is part of that experience certainly but the human player should always be front and center.

 

[AlpsStranger]

Is there any possibility that an AI xpac/dlc could be funded? It would be pretty niche, and I am not sure how difficult the "problem" is to solve.

What if there were a $30+ DLC and ALL it did was allow you to play on higher "AI-Difficulties." They'd play as Prince AIs that knew what they were doing, basically.

But I'm under the impression that this is outside the scope of a game project and would represent an actual leap in AI technology.

 

[cazaderonus]

Yeah AIs cannot be made to be actually intelligent, especially in a video game with as much layers as a CiV game. But there still has to be some coding solutions to many of the problems...

I know nothing of programming, but can't you hardcode some stuff for any AI and any circumstances ? They might end up being more predictable but i tend to think that even a predictable solid play is better than a completely ackward surprising one.

Defense for instance, isnt it possible to hardcode for AIs to keep a garison unit in every city close to another civilization ? Then add a subsequent routine over time that asks the AI to keep X% of its unit fortified near the cities the closest to another civ and mitigate those with the diplomatic modifiers ? Or add some basic milestones the civs need to complete, like having X unit when reaching a specific tech or civic or else ? Some triggers that would only come into play to define the minimal required investment for every aspect of the game ?

IE, a civ is flavored toward science and culture and doesnt build much of an army. At various point throughout the game, the AI is checked versus those minimum requirements. If it doesnt match them, it shifts priority instantly. So that culture\science civ focused on the formers and suddenly it receives a hard order to build X unit until it satisfies the routine.

Like i said, i m no programmer nor developper, so i might be wrong. But i have a hard time believing that you can't do anything to force an AI (and thus not let it "think" for itself) to follow some basic rules in how they play the game.

 

[stealth_nsk]

It makes no difference to you? I'd much rather play against an AI that can defend its units and has some sense of how to attack and defend than an AI that just produces a massive carpet of units because of its production bonuses.

It's only small part of what I'm talking about. Anyway, in terms of gameplay both situations provide challenge. Dealing with a lot of badly positioned units has its own fun. And if you worry about carpet of doom - worry not, in Civ6 AI will not build that many military units on higher difficulty levels - it will have direct bonuses to strength. Not sure whether you'll like this or not. I'm definitely looking forward for this challenge.

 

[AlpsStranger]

Like i said, i m no programmer nor developper, so i might be wrong. But i have a hard time believing that you can't do anything to force an AI (and thus not let it "think" for itself) to follow some basic rules in how they play the game.

The problem is, those "forced actions" could themselves become extremely exploitable weaknesses.

Example "Haha, I just left a scout near their empire and half their army stayed home!"

It's only small part of what I'm talking about. Anyway, in terms of gameplay both situations provide challenge. Dealing with a lot of badly positioned units has its own fun. And if you worry about carpet of doom - worry not, in Civ6 AI will not build that many military units on higher difficulty levels - it will have direct bonuses to strength. Not sure whether you'll like this or not. I'm definitely looking forward for this challenge.

This actually seems like a better solution. It's also scary and not just an annoyance to the player.

 

[stealth_nsk]

The problem is, those "forced actions" could themselves become extremely exploitable weaknesses.

Example "Haha, I just left a scout near their empire and half their army stayed home!"

Yeah. In Civ3 AI didn't have fog of war and his highest priority of attack was undefended city. Human players moved units out of the city, waited for AI army to appear nearby, after this moved units back to the city and did the same with another city on the other side of the continent. AI armies turn and walk toward new target.

Hardcoded AI conditions is very exploitable approach.

 

[cazaderonus]

The problem is, those "forced actions" could themselves become extremely exploitable weaknesses.

Example "Haha, I just left a scout near their empire and half their army stayed home!"

Well, i guess it'a question of prioritizing routines. If a civ doesnt decide to go to war, i see no problem having half of its army staying home. Acutally i'd expect a higher number.
Also, AIs know everything in the game, and they use it for sure. They know how many military you have etc... They could easily use those numbers as grounds to those routines to avoid any exploit.

Or in your example, if you park one scout near their borders but your overall military is 2 unit, you shouldnt see much of an army in that place.

 

[cazaderonus]

Yeah. In Civ3 AI didn't have fog of war and his highest priority of attack was undefended city. Human players moved units out of the city, waited for AI army to appear nearby, after this moved units back to the city and did the same with another city on the other side of the continent. AI armies turn and walk toward new target.

Hardcoded AI conditions is very exploitable approach.

Yeah, but i wasnt suggesting anything that big. What i had in mind was more about defining minimal requirement for AIs to follow and then avoid any huge hole in how they play. Like the lack of units, or walls etc..

 

[Magil]

It's only small part of what I'm talking about. Anyway, in terms of gameplay both situations provide challenge. Dealing with a lot of badly positioned units has its own fun. And if you worry about carpet of doom - worry not, in Civ6 AI will not build that many military units on higher difficulty levels - it will have direct bonuses to strength. Not sure whether you'll like this or not. I'm definitely looking forward for this challenge.

Really?

Spoiler :

Maybe they're not military units, but they're just as annoying.

 

[stealth_nsk]

Maybe they're not military units, but they're just as annoying.

Yeah, we know AI building priorities are messed in this build. It has nothing to do with the thing what AI gets less bonuses to production and direct combat bonuses instead on higher difficulty levels.

 

[Magil]

Yeah, we know AI building priorities are messed in this build. It has nothing to do with the thing what AI gets less bonuses to production and direct combat bonuses instead on higher difficulty levels.

Fair enough. Still, I'm not sure what exactly leads to that kind of apostle spam. It's not exactly a good sign for the AI and its coding.

 

[stealth_nsk]

Fair enough. Still, I'm not sure what exactly leads to that kind of apostle spam. It's not exactly a good sign for the AI and its coding.

Apostles can only be bought with faith. This means too much focus on getting faith, too small focus on spending it elsewhere, like patronizing Great People (maybe to not break the game for human players). Maybe the issue is actually outside the AI and the game needs another faith siphoning mechanics for AI to use. Or developers gave AI some faith bonus on Pronce for some reason or by mistake and they could just rebalance it. Or AI has too high priority for faith buildings and this could be just balanced with priorities. One of those reasons.

 

[juanpavo]

Yeah. In Civ3 AI didn't have fog of war and his highest priority of attack was undefended city. Human players moved units out of the city, waited for AI army to appear nearby, after this moved units back to the city and did the same with another city on the other side of the continent. AI armies turn and walk toward new target.

Hardcoded AI conditions is very exploitable approach.

I somewhat agree with you, but you have to admit that this tactic is valid, to feint an attack to draw the enemy out of position. If I saw the AI approaching in strength from one side, I would certainly send my military there in defense. And if the AI then attacked for real from the opposite side, I would say "well done".

 

[Magil]

Apostles can only be bought with faith. This means too much focus on getting faith, too small focus on spending it elsewhere, like patronizing Great People (maybe to not break the game for human players). Maybe the issue is actually outside the AI and the game needs another faith siphoning mechanics for AI to use. Or developers gave AI some faith bonus on Pronce for some reason or by mistake and they could just rebalance it. Or AI has too high priority for faith buildings and this could be just balanced with priorities. One of those reasons.

That's an absurd amount of apostles (costing an even more absurd amount of faith). They would've needed to have been purchased in waves. I have to believe there's something else going on there other than high Holy Site/faith building priority.

 

[stealth_nsk]

That's an absurd amount of apostles (costing an even more absurd amount of faith). They would've needed to have been purchased in waves. I have to believe there's something else going on there other than high Holy Site/faith building priority.

Yes, I agree with you. It's probably some kind of bug in faith production.

 

[Acken]

I am against this if it means putting in mechanics that are uninteresting to the player. Remember that the primary purpose of a single player game is to give the human player an enjoyable experience. This means giving the player some challenges, some fun, some highs and lows etc. The game should be designed around the player experience. The AI is part of that experience certainly but the human player should always be front and center.

This is not an exclusive situation.

And in my opinion not many of civ6 features are truly problematic for the ai. Many of the problem seen to the ai can be easily fixed.