Basic question re AI: RTS versus TBS

[kaltorak]

If that's an actual reason for poor AI decision making, why not add an option to allow the AI longer to finish it's turns? I'd happily let the AI have as much time to think as it needs, if that would result in a better game experience.

Lol would be great, but it's the opposite. When a game is real time, and you have to to click 4 buildings for constructing units, order 15 units to advance, and send 3 workers gathering different materials while planing an expansion, all at the same time, then the AI shines. Not the opposite. The more time doesnt make the machine better (at least not much). But the less time does make the player worse.
If starcraft would be a game that is paused and just moves as you click advance, no AI would win a pro player ever.

 

[legalizefreedom]

"Trying" implies attempting to attain one of the victory conditions as defined by the game. In addition to its usual tactical failures, the AI is *intentionally* designed not try at the strategic level. This is not a respectable choice and players should not respect it.

People can open up Civ 6 then not play Civ 6 (up to and including literally, such as going to do laundry while the game is open). If they're not in MP that's their sole prerogative. If they are in MP and their opponents didn't agree to them doing that they should be kicked. This choice has no bearing on the AI regardless.



What's flawed is the reasoning used to make such a case. An arbitrary amount of "simulation is okay", but not other times. Simulation is okay, until it isn't okay.

Incoherent logic should not inform game design, or most decisions for that matter. And incoherent logic is what is being used to "expect the game to be more of a simulation of human activities".

What do actual human activities in Civ 6 look like? I bet they're different from the AI's activities, even among the players purporting they want to role play. People will "role play" then complain if they lose to the AI attacking them. That's not credible rationale and it shouldn't be treated with any meaningful weight in the game's design. Players making that case are necessarily irrational in that context.



Those are expectations of active dishonesty and self-inconsistency. The game has clear incentives. Historical leaders acted on historical incentives. In a scenario where "historical" AI are ignoring their incentives, the assertion that it is a "historical simulation" is objectively false.



They get no such pass for this. Design the game first, then make the AI play the game. If they think the game would suck if the AI actually plays it, maybe they should rethink their design.



If the AI playing the game optimally isn't fun, the AI isn't what's at fault with the game. The game's design/rules/victory conditions are what's broken in that case, not the AI.

Many of your points are the way you personally see things. Everything you say about simulation, could also be said about 4th wall puzzles, etc.

When you're dealing with things, logic works. When you're dealing with people, feel works.

I'm not defending or advocating the game design of VI, only stating that some folks find entertainment in very different ways within the confines of a game. As evidence, GS launches, I start a game of BERT...

For example: I play starcraft like a TBS. Wait, isn't starcraft all about the twitch? Not for me. There are a vast number of starcraft players that would state as fact I'm playing the game wrong and my desires for the game (like not having game speed tied to difficulty) had no merit, but you know what? They do for me. It seems so shallow to me to simply click/think faster to win.

In Civ, victory conditions are clicking faster to win. I'm just not into fast strategy. So sue me.

 

[TheMeInTeam]

Many of your points are the way you personally see things. Everything you say about simulation, could also be said about 4th wall puzzles, etc.

When you're dealing with things, logic works. When you're dealing with people, feel works.

Civ 6/games in general are not people. I know the UN and other morality fools get confused about that, but we can do better .

I'm not defending or advocating the game design of VI, only stating that some folks find entertainment in very different ways within the confines of a game. As evidence, GS launches, I start a game of BERT...

That is why there are different games, game genres, and non-video game entertainment.

What I'm calling out are statements with logic akin to "I hate all games unless they have dominoes", then in the next sentence saying that chess is that same person's favorite game. That isn't coherent, and it shouldn't be used to inform decisions about how chess should be designed or played. Alphazero shouldn't start putting foreign objects on the chess board.

Wait, isn't starcraft all about the twitch? Not for me.

Not for anybody optimizing to win. The event that started this thread is evidence.

In Civ, victory conditions are clicking faster to win. I'm just not into fast strategy. So sue me.

In StarCraft, if you join games with teammates and don't attempt to destroy opponents' bases you'd be rightfully banned for griefing. What you do on your own in SP doesn't matter, but it still shouldn't influence how StarCraft is designed or units are balanced.

As I said, if the developers don't anticipate that the game is fun, they should make the game differently. There's no law stating that Civ needs 4+ victory conditions, single winners, or any particular VC or rule. Those were all things the devs intentionally put in the game. Why, then, are they openly admitting they don't think the game when played to incentives is fun by instructing their AIs in the spot of a human player not to play it?

Nobody seems to be able to answer for that one.

Yes they are, on a fundamental level: In real history you don't achieve "victory". So we can either drop the victory conditions altogether, or have a purely gamey AI, or something in between. I choose something in between and I dont see anything wrong with it.

What is "wrong" is using incoherent rationale as a basis for how the game should work. You get arbitrary results and a lower quality game.

Maybe reasonable was wrong choice of words on my part. Strictly speaking not every ruler acted what I'd call reasonably. What I meant was believably. Believably as leaders of a country and not as gamers in a game.

It's very hard to create all the right initiatives for the Civ leaders to make them act believably as real life rulers.

Civ AIs that lead civ *are* gamers in a game. They are competing players taking a slot that could also be a human, and operating under the same rules as a human.

The quoted logic is incoherent. Civ 6 rules represent the world in which the Civ 6 leaders "live". When Civ 6 leaders ignore the game rules, they literally *can't* be behaving historically. The concept is an oxymoron. Historical leaders didn't ignore their incentives.

The art is to make the game gamey (for the player) while also keeping immersion and a semblance, or a feeling of "realism".

Quoted feeling of "realism" is objectively fake.

Example: while in real life we too had a space race, USA and USSR didn't even consider to go to war to prevent each other from being the first.

What do you think would happen if someone winning the space race meant "game over" - the other country couldn't continue to exist? Not only did USA/USSR consider war and engage in proxy wars in real history, but I expect nukes would fly under this hypothetical.

But in civ the game, not doing that likely results in you losing, so a gamey AI should definitely do it.

Exactly right. In the game, the incentive is to attack. So the player/AI should attack if trying.

Even if technology is already there, the amount of work needed would make the price of the game ridiculous.

You seem to be misunderstanding the proposed solution.

I'm advocating completely changing the rules of the game, not going into some advanced tech AI design. My whole point is that so long as the devs consider the design of the game itself poor, having a good AI isn't even a practical goal.

 

[legalizefreedom]

Civ 6/games in general are not people. I know the UN and other morality fools get confused about that, but we can do better .



That is why there are different games, game genres, and non-video game entertainment.

What I'm calling out are statements with logic akin to "I hate all games unless they have dominoes", then in the next sentence saying that chess is that same person's favorite game. That isn't coherent, and it shouldn't be used to inform decisions about how chess should be designed or played. Alphazero shouldn't start putting foreign objects on the chess board.



Not for anybody optimizing to win. The event that started this thread is evidence.



In StarCraft, if you join games with teammates and don't attempt to destroy opponents' bases you'd be rightfully banned for griefing. What you do on your own in SP doesn't matter, but it still shouldn't influence how StarCraft is designed or units are balanced.

As I said, if the developers don't anticipate that the game is fun, they should make the game differently. There's no law stating that Civ needs 4+ victory conditions, single winners, or any particular VC or rule. Those were all things the devs intentionally put in the game. Why, then, are they openly admitting they don't think the game when played to incentives is fun by instructing their AIs in the spot of a human player not to play it?

Nobody seems to be able to answer for that one.

But the way people consume games as entertainment does lean to the feel.

There is virtually no such thing as a perfect game for an individual, so we play the closest match while tweaking the experience to our preferences.

All the stuff you advocate for is perfectly fine with me. Refining the game is unquestionably good.

Just because you can always "win" with domination, for example, doesn't mean that everyone enjoys playing that way. I would rather play the way I find the most entertaining and leave the victory focus for min/maxers. It's just not the primary driving force behind my play. It's a requirement, but not the reason I play.

And maybe the AI has a bit of that frivolous, impractical spirit...

 

[TheMeInTeam]

But the way people consume games as entertainment does lean to the feel.

There is virtually no such thing as a perfect game for an individual, so we play the closest match while tweaking the experience to our preferences.

"Feel" is still derived from something in tangible reality. The game doesn't just display a picture of rocks and ask people to "feel" like they're having fun playing it. We can at least strongly suspect the "fun" therefore comes from properties of the thing (aka Civ 6). It's what the thing does vs doesn't do that's in question.

If a given person X claims that Y mechanic makes them feel good or bad, we should soundly reject their position if they arbitrarily enjoy or dislike Y without apparent reason.

Nearly every "AI should behave historically/role play" argument carries that fundamental flaw. Players are literally arguing against history in the name of history, and considering it a valid point. Whatever Civ 6 designers ultimately choose, it would be nonsense to reject history in the name of history. That has no control for anticipated experience, nor does it give coherent basis for shaping mechanics.

If AI trying in Civ 6 wouldn't be fun, that's a flaw in the properties of Civ 6. If Civ 6 intends to incentivize things other than military as a primary vehicle to victory it fails. This is not the fault of the AI. It needs to be fixed first, before the AI CAN truly be "fixed".

On that note, the following things are irrelevant to this particular back and forth discussion:

• Just because you can always "win" with domination, for example, doesn't mean that everyone enjoys playing that way.

• I would rather play the way I find the most entertaining and leave the victory focus for min/maxers.

• It's just not the primary driving force behind my play. It's a requirement, but not the reason I play.

Finally...

And maybe the AI has a bit of that frivolous, impractical spirit...

There's no evidence that WE have "spirits" (unless we're talking alcohol, then maybe ). The AI just does what it's been programmed to do. Right now, that is something other than consistently playing Civ 6.

 

[CornPlanter]

What is "wrong" is using incoherent rationale as a basis for how the game should work. You get arbitrary results and a lower quality game.

I don't see anything incoherent about my rationale.

I'm advocating completely changing the rules of the game, not going into some advanced tech AI design.

And I am saying rules as you propose would require advanced tech AI design if they are complex enough to offer both challenge and immersion only by providing AI "realistic" initiatives to work with.

 

[Growth]

This thread started in Jan 2019, and we're about ending 2022 right now, and I'd like to just leave this here:

CICERO: An AI agent that negotiates, persuades, and cooperates with people​

Now be honest, had we asked the question back in 2019 about the table game Diplomacy, how many of you would think
we'd be having human level performances by now? Are we now just one small github team away and a few months from having human a model that
has human level performance in CIV?

 

[AntSou]

I don't think either Diplomacy or Starcraft are more difficult than Civ when it comes to designing a competent AI.

Starcraft doesn't have "infinite possibilities" any more than Chess does. It's a 1v1 with 3 factions and with a limited pool of optimal plays. It's a competitive game exactly because of its limited possibilities. Same goes for games like League of Legends, which give the impression of limitlessness due to its character pool, but that mainly involves optimizing specific strategies.

Diplomacy is a game that involves a lot of a Prisoner's Dilemma kind of thinking, isn't it? Don't have time to watch that right now.
---

Civilization games are on a board with multiple players where the test of good AI is not only being able to accomplish and optimise basic strategies and tactical plays (I sincerely hope this improves in Civ 7), but also, and probably more difficult, to design an AI in such a way that it feels like a Character, rather than a Human Player attempting to achieve victory by any means.

In other words, an intelligent AI in Civ is one that can be "selectively dumb". E.g. A leader with a naval orientation whose personality is directed towards naval raiding requires that to be its focus rather than seeking the single most optimal strategy to win... which would probably involve focusing on land units (unless playing archipelago maps).

Frankly, AI in Civ needs to be nowhere near that level of optimization. THIS IS EXACTLY WHY IT IS SO FRUSTRATING. There's so many times in Civ 6 where simpler solutions would be better handled by current AI capabilities, yet the game seems often designed without this taken this into consideration.

At other times, like tactical play during warfare, or strategic decision making regarding district placement, the game doesn't seem to be that far away from an "aceptable" level. That's all the game needs.

Atm, the Civ AI just kinda royally sucks at the very mechanics which distinguishes Civ 6 from previous titles (District Placement, 1UPT tactical play, Builder and tile improvement management).

 

[aieeegrunt]

The apologia for the terrible AI in Civ6 is pretty sad

You can use any of the AI mods on the workshop and get an immediate noticeable improvement

 

[legalizefreedom]

Starcraft doesn't have "infinite possibilities" any more than Chess does. It's a 1v1 with 3 factions and with a limited pool of optimal plays. It's a competitive game exactly because of its limited possibilities. Same goes for games like League of Legends, which give the impression of limitlessness due to its character pool, but that mainly involves optimizing specific strategies.

I believe that comes from the fact it is an RTS. Units can move in any direction, stop on any spot, etc. If you zoom out far enough I suppose you could boil it down to chess moves, but that's a stretch.

The apologia for the terrible AI in Civ6 is pretty sad

You can use any of the AI mods on the workshop and get an immediate noticeable improvement

The AI in Civ is not bad because they aren't smart enough to make it good. It's bad, because in general, people don't like to get their faces smashed in.

 

[pokiehl]

The AI in Civ is not bad because they aren't smart enough to make it good. It's bad, because in general, people don't like to get their faces smashed in.

I agree it's insulting and not a good argument to imply the developers literally are unable to make better AI due to skill. However, purposefully dumbing down the AI to make it easier for people seems like nonsense to me.

First of all, that's what difficulty settings are for. Second of all, the AI literally is unable to interact with core aspects of the game's systems. That sort of flaw goes beyond what you're implying.

Ultimately I think the answer to why the AI is bad comes down to a choice of how to allocate the scarce resources of time and money across the development of a very complex game--not to make the game easier on purpose or because they literally can't fix it.

 

[legalizefreedom]

I agree it's insulting and not a good argument to imply the developers literally are unable to make better AI due to skill. However, purposefully dumbing down the AI to make it easier for people seems like nonsense to me.

First of all, that's what difficulty settings are for. Second of all, the AI literally is unable to interact with core aspects of the game's systems. That sort of flaw goes beyond what you're implying.

Ultimately I think the answer to why the AI is bad comes down to a choice of how to allocate the scarce resources of time and money across the development of a very complex game--not to make the game easier on purpose or because they literally can't fix it.

Yeah, I think it is the combination. We're not going to make AI that frustrates the average player, so we're not going to break the bank creating one.

 

[ThunderLizard2]

Well simply because it is very hard to make AIs that are good at any complicated system. It was not "easy" to make one good at Starcraft. The work has been done over many years by potentially 1000s of people costing billions of dollars, and I suspect they want to keep the knowledge somewhat confidential for commercial reasons. I doubt Firaxis has that capability.

Also note that these AI systems typically use neural networks so there is a problem with transparency and control. A neural network might be exceptional at making decisions but you don't know why it makes them and you might not necesssarily be able to tone them down - who wants to play against a supercomputer and lose everytime? Whereas the CIV AI is at least deterministically controlled and explainable.

You're missing the point entirely. You don't over complicate the game and then throw up your hands that it's too complicate for the AI. If a feature is too complicated for the AI then take it out!. You should watch the lectures by Soren Johnson and Sid Meier on designing AI for Civ games.

 

[AntSou]

If a feature is too complicated for the AI then take it out!.

This! (though keep in mind you're replying to a nearly four year old comment)

There's so many times in Civ 6 where this exact thought pops to my mind. Just recently we got Lincoln, which is an example of that issue.

In Civ 6 the question seems always framed from the point of view: "will this be fun for the human player?". Whether the AI can do anything with it doesn't seem to be a priority.