Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Discussions Rules and Guidelines
Report this post
To a great extent, this is down to the same inherent limitation that forces the design solution of giving the AI players advantages as the way to let them create a challenge for HI (human intelligence) players. The game is too complex to let its system AI judge the game state well enough to see what an HI can see easily, that you can no longer possibly lose, and that is the same inability that leaves the player AIs dependent on programs that are too limited to mimic a good game play strategy at all well.
It seems to me that the devs could do better at making the end game more of a challenge if they shifted the AI player programs in the end game, away from the balanced strategy that serves them better in the early and mid-game, and towards a strategy that cones in on the one particular victory condition they are closest to achieving. HIs have little trouble making the transition in the end game, and become ruthlessly focused on the much more limited breadth of effort that is needed to get to that particular victory condition. The transition would certainly involve the same sort of global assessment of the game state that AI struggle with, but it seems to me that an AI that has got the first satellite into space has got to a relatively clear and programmable end point for switching all efforts to doing only what science victory requires.
The other thing the devs could do here is to make the victory conditions easier to attain. The last mile between you being not at all likely to lose anymore, and you being categorically unable to lose anymore, can eat up a lot of turns, and that could be shortened by just making the victory conditions less extreme. The TW series recognizes a lesser victory with easier conditions than the complete victory, so Civ could imitate that and have two or more such levels to keep the real fanatics happy that they are better than the rest of us because they won a more victorious victory level than we did, while we lesser beings settle for the consolation of just ending the game 50 turns earlier
What players can do is, most simply, to just quit when the game gets rather pointless because you can see clearly that you can no longer lose and so no longer have to use strategy to win.
Of course, if you are the sort of obsessive compulsive person willing to put in the hours to get good at this particular game, you are less likely than the general population to find that solution of just abandoning an incomplete game at all acceptable. I am plenty OC enough that I've never done that. What I do instead is to just coast after I reach the point at which neither careful nor daring strategic play is any longer necessary, because I can't possibly lose anymore. I find science victory best for this, the fewest mouse clicks to the victory screen, so I mostly end games with that victory type, no matter how I got to the point of being unable to lose anymore.
1. Turn times would be longer, probably becoming enormous by end game. People would complain. Personally, I would rather have long turn times than stupid AI, but most people would lose their crap and give the game negative reviews. Even if endgame turns were one minute long, I wouldn't mind but everyone else would freak out.
2. People would lose (a lot), and people would complain that it is too hard. We live in a safe space, everyone gets a trophy society, and unfortunately, most peoples response to losing would not be to practice.get better and make better decisions, it would be to quit and cry to the developers and leave bad reviews.
People want to think they are good at strategy games. A lot of people aren't. It is not a good financial strategy to sell a game that reveals how bad a person is at strategy games. You need them to buy DLC's, and they won't when they can't win all the time and feel good about themselves (even if it is fake).
The bad AI is not a bug. It is a feature.
The only way we will ever get a Civ like strategy game that has really good AI is for a developer to make a game of passion and literally not care if it sells one copy.
This is a very good idea. In two of the games I managed to win at Emperor level one or two AI civs got to the moon before me but in both cases I got to mars first which presumably I would not have done if the AI civs had been more focused. (In one case one of the AI civs got to the moon before 1800 !!!).
That is what I have decided to do. I am going to accept that King difficulty is as hard as I want the game to be and stop playing once I am almost certain I am going to win.
Take your proposition 1., that the only, or at least main, reason that the devs didn't develop, and publish with the game, such an AI, is that computers would take too long to make the AI's moves, and this would make sales plummet.
I really doubt that you could build an AI for this game comparable to the AI that has been developed for chess, that is indeed capable of defeating the best HIs. This game is way more complex, systems within systems, while chess is just one system. I am much more confident in claiming that even if you could build such an AI for Civ6, it would require at least as great a project as the effort it took to develop programs that could beat humans at chess. That effort took many very talented people decades to get to a program that could beat grand masters of chess first, then even HI world chess champions more recently. No doubt some people ended up recovering some costs for all that effort towards the end, when actual products you could sell on the market were the final result, but, as with almost all pharmaceutical breakthroughs, the initial effort was a labor of love, a passion product of people and corporate executives who saw no prospect whatever of recovering any of the considerable costs associated with these early stages. Quite at odds with the frequent end-game with many pharmaceuticals, even now that we have marketable chess AI, it's not as if anyone is making fabulous profits off of them.
Civ6 isn't as popular as chess, but I can't imagine it doesn't have its passionate fans, and that those people do not include any number of folks who struck it rich in Silicon Valley and now have billions on hand that they don't know what to do with. If some sort of Deep Blue for Civ were possible, why haven't we seen one of these billionaires finance it? It's not as if we don't see one of these people launch any number of fool vanity projects, or passion projects if you will, all the time. And if this breakthrough in Civ6 AI science were possible, analogous to chess AI, if you could get the AI to beat the best HI in a fair fight, why not get it down to where chess AI is today, and get the program to the point that it doesn't take forever to decide on its next moves?
Your point 2. is even less controversially refuted by the facts. This is such a complex game that people do lose, all the time, at Settler. Follow this forum if you won't take my word for it, because about 50% of threads are people complaining about exactly that, losing to the AI at lower difficulty, or having a hard time beating it at higher. The other 50% of threads are people boasting that the AI is just too damned easy for their galaxy-sized brains to beat. Maybe some simpler games have one relatively narrow path to victory, and I imagine that fans of that sort of game wouldn't tolerate that path being too obscure and them losing too often (or at all), so the devs put in flashing neon signs to guide these gamers on the right path. Civ6 fans in contrast are gluttons for punishment. We play a categorically open-ended game that takes hundreds of hours to become acquainted with all its systems, spurred on and instructed by failure at every turn, until we get reasonably competent. We either love losing -- or at least almost losing -- or we stop playing Civ and seek out some straightforward game that petty much guarantees a more predictable and secure path to a win.
Even if it were possible to develop and field an AI capable of beating an HI in a fair fight, it would be child's play to put in easier difficulty levels for even the least experienced HIs to beat this AI. Settler level, in the game as it stands, falls far short in what it gives the HI, than what Deity gives the AI. All that Settler does now is to give the HI some piddling combat advantages. Of course beginners struggle to beat the AI at Settler with those puny advantages, because they get none of the huge starting bonuses, and ongoing yield bonuses, that Deity gives the AI, and tghis is a complex open-ended game, with systems within systems to master. If Civ had AI capable of taking on HI on a level playing field, the devs could very easily just give the HI on the lowest difficulty level, bonuses similar to what the AI now gets on Deity. That would solve any problem of the game being too hard to beat tanking its popularity. Of course its present fan base would still be addicted to losing (more accurately -- almost losing!), to the point that players would seek whatever new difficulty in this new version made us almost lose as much as possible.
Maybe a "cumulative victory condition" could be something like if the player is 3/4 of the way to two or more victory conditions and ahead of anyone else by 25% they can declare it over.
I think if a developer can make a 'grand campaign' style game with good AI they will actually be really successful - look at Dark Souls. People want more challenge, they want to feel like they have achieved something instead of feeling like they're being catered to.
... not everyone is like this, though. After a long, hard day at work the last thing I want to do to relax is try to wrestle with a game that repeatedly kicks my butt. It's why I try to avoid games like Dark Souls, Hollow Knight, etc. If it has a reputation for being difficult, I try to avoid it.
There's a reason why I typically play Civ on the lowest difficulty settings. And I appreciate that the developer has options like that in their games. Not everyone wants a mind-numbing challenge, especially when their brains are already numb. They just want to sit back and chill.
I hear you dude, everyone has different interests, but I feel like more and more people are bored of playing easy games. There should be a market for both types of games, casual games and then more challenging ones as well, but I totally respect your point of view.
Problem is that Dark Souls is an asymetric game where you fight a lot of enemies while civilization is a symetric game where you fight opponents (human ou bots) that has the same start and the same goal than you. Difficulty is pretty different by essence because it's easier to balance the difficulty of enemies that have a gameplay different than yours than enemies that have the same gameplay than yours.
But even with this, how does Dark Souls emulate difficulty and difficulty rise ? When you reach NG+, NG+2, NG+3 ? By raising enemy's AI ? no. By giving them hit points and damage bonus, or by rising their number ^^; the same way civ does with its difficulty level. That's the only proper way at this time. Maybe in 15 years we'll have learning AIs way more flexible and that can be run is a domestic computer, but this time hasn't come yet.
I wasn't comparing this to Dark Souls, just saying that many players are looking for more challenging experiences than the ones the big companies generally put out these days.
The post before mine by thpf77 mentioned that devs are inclined towards making games easy to suit the average gamers. I would suggest that this definition has changed, average gamers are not all interested in easy games any more, since many have been playing video games for a long time at this point, the average age is no longer a teenager, like when I was growing up.
You make a good point about AI, it is definitely going to figure heavily in the future of gaming, look at what Chat GPT can do, I have even seen a mod for Bannerlord that uses the app to write dialogue for all the npcs, based on the various factors governing their lives it creates realistic responses.
But that's only a small part of what machine learning is capable of. I find it interesting that no major developers as yet have really embraced the new tech though, kind of shows how 'set in it's ways' the industry is.