Status of Civ 6 AI

[KayAU]

the vast majority of players struggle on settler difficulty: so there is no incentive for 2k/Firaxis to work on AI

If someone is struggling with Settler difficulty (or Chieftain), that is probably not due to the strength of the AI, but rather that they are struggling to grasp the gameplay and mechanics of the game. Which is fair, as Civ VI is a fairly complex game. I don't think that is an excuse not to provide proper AI which knows how to reasonably play the game though. I am not talking about it mastering "gamey" min-maxing techniques like placing down districts just to lock the production cost, or waiting for an era change before harvesting. But I would like it to understand things like building a large enough force with a sensible composition before attacking, or knowing how to effectively move units around to attack or defend. I'm talking about basic competence at playing the game.

As for incentives for Firaxis, I think they exist:

• Pride in their work. There is no doubt in my mind that Firaxis is a great game developer, and it is lead by the great Sid Meier. I do believe they generally care about what they create.
• Pleasing their core fans. The players who care the most, and talk the most, about Civilization are generally "advanced" player, who care about the AI.
• There are currently loads of negative Steam user reviews, many of which are complaining about the AI. That can't be good for sales.
• Even if it doesn't have a huge immediate financial impact, it is important for the long term success of the franchise that they address the most signifiant flaw of the game in its current state.

Don't get me wrong, various issues aside, I do still think Civ 6 is a great game, which I thoroughly enjoy.

 

[PhilBowles]

go have a look at "game of the month", it's mostly "high end" players so the average civ player is struggling on... ok maybe Chieftain not Settler
Does not change the argument, though

Isn't Game of the Month a forum thing? All that suggests is that the people playing it are a self-selecting minority keen enough on Civ to follow forums and also to download 'scenarios' to play through. There's going to be a strong correlation between that demographic and high skill with the game, but it's hardly a representative exercise.

 

[HF22]

@Ferocitus

Again, you can like the development which has occurred. Your choice, no one is arguing with it.

But the question is if future development will address the concerns others have. And the track record, from both Civ 5 & 6, suggests not.

So it is perfectly reasonable for people to start assessing Civ 6 as is, without others demanding they wait several years for mythical improvements we have no historical basis for expecting.

 

[PhilBowles]

@Ferocitus

Again, you can like the development which has occurred. Your choice, no one is arguing with it.

But the question is if future development will address the concerns others have. And the track record, from both Civ 5 & 6, suggests not.

So it is perfectly reasonable for people to start assessing Civ 6 as is, without others demanding they wait several years for mythical improvements we have no historical basis for expecting.

Firaxis' track record, in fairness, is to devote large parts of most patches in both games to AI fixes, which I doubt would be the case if the AI wasn't at least a consideration for the majority even if it isn't a make-or-break issue for most. During Civ V's life they revamped several entire systems - such as hit points and naval combat - to benefit the AI. You can't reasonably draw a straight progression between Civ V and Civ VI, in part because Civ V got progressively more complex over time and AI improvements were needed just to keep pace, and in part because Civ VI is a regression from the point Civ V reached by Gods & Kings. Part of that too is due to changes in game mechanics most of which are hostile to the AI, though most seems just to be bad programming (and I'd argue that it's not any better than Civ V was at this point, it just benefits from a couple of AI-friendly changes - like increased unit hit points and pillaging to heal - that hadn't been introduced at this point in Civ V).

I'd also question how fair it is to suggest that the Civ V track record 'suggests not'. Functionally, what was Civ V's AI unable to do adequately at the end of its life cycle? It still required buffs to manage happiness because it could never master the mechanic, and it was still easily exploitable with gold trading, but most things - city placement, unit positioning, timing of attacks - were at the level of an inexperienced to middling human player rather than overtly stupid decisions (though it never quite overcame a tendency to wear cities down with ranged units while having no melee forces anywhere in the vicinity, and it still embarked and loitered in the sea in range of ranged units). It was capable of planning and effectively building towards a peaceful strategy, and long-term bugbears like failures to escort settlers had been fixed.

 

[Archon_Wing]

If someone is struggling with Settler difficulty (or Chieftain), that is probably not due to the strength of the AI, but rather that they are struggling to grasp the gameplay and mechanics of the game. Which is fair, as Civ VI is a fairly complex game.

Yea, it is what how I define "lose on their own" as opposed to good play from an opponent. A good example is someone not being able to make a basketball shot, or soccer goal on their own. It is their inability to win that makes them lose. That being said, I don't even know mechanics of games I've been playing for years.

But yea we definitely want an AI that isn't like that.

mastering "gamey" min-maxing techniques like placing down districts just to lock the production cost, or waiting for an era change before harvesting

For some reason, I hold things like the former in disdain but don't have a problem with the later. I guess they could be both considered gamey, but the later is less offensive to me, maybe because it is a bit more intuitive.

 

[HF22]

@PhilBowles

Functionally, what was Civ V's AI unable to do adequately at the end of its life cycle?

As with Civ 6, its combat AI was inadequate, including with absolute basics like "move & shoot" and "have melee to take city".

Now granted, that is less game breaking in Civ 5, as global happiness discourages the conquest as exploit issue Civ 6 has.

But it was never fixed.

 

[earlc]

"And next on BBC 2: The World at War, in the style and voices of Rick and Morty".

Good idea, personal preference, or objectively inappropriate idea, would you say?

That's just a false comparison.

Same thing here - it breaks immersion in a game whose entire purpose is to provide a sense that you're re-enacting a historical development.

There's no reason at all that is necessarily the case. I've been trying to focus on things that the V and VI leader screens factually, actually do, and how they interface with the user.

Even the language you've been using to describe it concedes as much - you want characters expressing emotions, not actual civilisations, which is at odds with the game's core concept.

One, those things are not mutually exclusive. And that also isn't what I'm arguing - I'm saying they more coherently, clearly, and smoothly interface with the user over the long haul. And that's just simple logic.

I know if I'm playing an older Civ game that, yes, Germany is led by Bismarck or France by Napoleon, but when I interact with those civs I'm playing as Civ X interacting with Germany or France, whose representatives are there just to fill excess bits of screen with appropriate images, I'm not having a get-together with Otto. They're there to portray the characters in a thematically appropriate way, not to be the main focus of their civs.

Hmm, you might be alone there. I've always taken the leader as the representative, the "player" if you will, of the other civilization. That's their purpose, and the major component to humanizing the game and allowing players to more easily identify the different civilizations they're interacting with (particularly the more novice players who haven't subconsciously memorized color schemes yet). Here you're basically arguing against leaders themselves, let alone leader screens. It's no wonder you prefer V in that case, since they minimize the leader graphic more than any other Civ. Though it does go against your earlier arguments about the effectiveness of V's leaderheads.

The CIv V images were obviously composed as full pictures - the character is wherever it's most appropriate to make the best picture. No, they aren't all well-judged but it's not clear at all why their being in different positions is an issue and it rather seems to argue against your earlier claim that they aren't distinct enough. Why is it inherently better to have Character With Beard 1 in the same place as Character with Mustache as though the whole thing's just an RPG character creation screen with an assortment of different heads and costumes? Master of Orion II was doing that in 1996, so it's hardly a daring, untried aesthetic.

The idea behind their composition goes without saying. As for why it's an issue, I've already made many solid points concerning why V's leader graphics are a failure from a sheer design standpoint, and also in how the game interfaces with the player.

And none of this is any defence of the stylistic choice for the leaders - wherever they are relative to their background, what benefit accrues from weirdly angular exaggerated caricatures when Civ V had success with naturalistic leader figures?

Well that's because it wasn't in defense of the VI leaders, it was a criticism of V's leaders, which I only realized are quite as bad as they are after picking up the new game.

As for why the stylized VI is "better" - I am indeed saying it's far better than V, though not necessarily better than a pure realist representation of the leaders could ever possibly be. Perhaps a highly talented character designer (which is actually how I'd describe VI's designer(s)) could make leader screens just as effective as they are in VI. But I've never seen it, and it's doubtful.

I've already made a lot of statements regarding the effectiveness of the slight exaggeration, though a quick rehash of the main points includes ease of recognition and memorability, ease of visual communication, and humanization/characterization/desterilization.

...And, quite simply, looks. Many of the most popular games in internet circles are popular in part because of their highly communicable and memorable characters that can be memed, drawn, illustrated, even dare I say turned into fan-fiction. People like characters. That goes back I think to my first post on the matter.

Man, even TV shows and film. Even the most serious drama benefits from unique, recognizable characters - unique faces you can easily apply some thought, feeling, emotion, or some other association to. You're right, no, this isn't something new at all: there's a long track record of its success in all forms of media.