1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

So apparently, the AI has not been improved

Discussion in 'Civ6 - General Discussions' started by Kruos, Feb 12, 2019.

  1. Spudsie74

    Spudsie74 Chieftain

    Joined:
    Sep 20, 2013
    Messages:
    111
    Gender:
    Male
    Location:
    Lake Charles, LA. USA
    This shows you what a slacker I am...I play at "Emperor," wher the AI gives me little trouble after the first 50 turns (and not much before that). It certainly will not wipe out my Civ...but just may keep me a little more honest. It is almost more about beating the clock, along with making decisions on the fly, rather than an external threat (the AI). For what it worth, I don;t thaink and AI that can't do decent midevil combat with be good with aircraft, or tanks. They might do OK with a nuke, if thee ever could build it. :p
     
  2. SammyKhalifa

    SammyKhalifa Warlord

    Joined:
    Sep 18, 2003
    Messages:
    4,879
    That's nice.
     
  3. PhilBowles

    PhilBowles Warlord

    Joined:
    Nov 20, 2011
    Messages:
    5,007
    Mechanical simplification would be my suggestion - as it is I'm among those who feel that endless make-work decisions with little gameplay impact in Civ V don't do anything to improve the player experience for the human either, though I don't want to go back to stack combat. My suspicion is however that this is in tension with Firaxis' understanding either of the desires of the audience, or at the very least of their own marketing requirements (as new systems require new mechanics). Games like Crusader Kings II (I'd like to use Stellaris as an example, but since the Stellaris AI is abysmal I can't) or Starcraft II (which has notoriously been 'mastered' by AI, albeit with extensive targeted investment and highly artificial situations) succeed in offering immersion and strategic depth in relatively simple systems - as did iterations of Civ that people look back on as having 'better AI'.

    If you're working from the premise that we're using the existing Civ VI engine, I think the best solution is targeting the specific areas where the AI gets bonuses and tightening up its ability to use those. Civ AIs have always had bonuses to help them compete despite complaints with every new iteration that this is an unwelcome new feature - but bonuses only help so long as the AI can actually use them to advance their game plan. The clearest example of the problem is the gold bonuses in Civ V: the AI was given extra gold but practically never spent it and was eager to give it to a human in trade deals, to the extent that the human could actually be advantaged on higher difficulties by 'gold-farming' the AI. As well as the practical gameplay issue, it made it very obvious that the AI had a not-so-hidden bonus, and this is when players tend to complain about bonuses.
     
    ChocolateShake likes this.
  4. Anno

    Anno Chieftain

    Joined:
    Aug 6, 2016
    Messages:
    82
    Location:
    Columbus, Ohio
    I actually think the game isn’t nearly enough of a sandbox. The game would be better for not having victory conditions, or at least not the handful of scripted ones currently in the game. Then the AI wouldn’t have to be focused on “winning”. I (and probably the majority of players) already don’t really focus on winning, so it’s weird that it’s the goal of your opponents.
     
    BackseatTyrant, Xur and wilcoxchar like this.
  5. Uncle_Joe

    Uncle_Joe Chieftain

    Joined:
    Sep 9, 2005
    Messages:
    467
    I tend to agree with the sentiment if not to the same degree. I've felt the same with Civ5 BNW (until the VP mod). There is no real 'threat' from the AIs (at least past the initial rush phase). And it doesn't even have to an actual attack to be a threat. In Civ4, some AIs could steamroll and threaten to win the game unless *I* took some sort of counter-measure or aggression. In Civ6, I don't see that happen.

    In many case, I can just keep a small army and that is enough to prevent any AI attempts (rare anyways) to thwart me. But the AIs usually don't even try. If they would effectively raid trade routes, I'd at least have to provide enough naval presence to protect my shipping...but they don't.

    I would love to see the AIs become a threat again (not only to attack, but to push for the win). Without that, you are indeed correct in that the game is closer to a sandbox or 'god game' rather than a strategy game.
     
  6. Archon_Wing

    Archon_Wing Vote for me or die

    Joined:
    Apr 3, 2005
    Messages:
    4,444
    Gender:
    Male
    Just by looking at past patch notes.
     
    tedhebert likes this.
  7. Midori

    Midori Chieftain

    Joined:
    Sep 14, 2016
    Messages:
    37
    The AI is playing a game that's more difficult and complicated than the types of games that the most sophisticated AI systems in the world have taken on. It took until 2015 for an AI to beat a human at go, and that's still much simpler than any Civ game. I'm actually surprised that the AI can accomplish anything and still play within the rules of the game.
     
    Wizard-Bob and ChocolateShake like this.
  8. ChocolateShake

    ChocolateShake Chieftain

    Joined:
    Dec 28, 2016
    Messages:
    517
    Gender:
    Male
    Out of curiousity, which single player strategy games would you consider not to be sandboxes?

    If we're using TotalBiscuit's measure of a game having a fail state, then both Civ and Oxygen Not Included do.

    Not every game is for everyone, Civ doesn't stand out to me in terms of single player 4x in terms of difficulty. And a game being difficult often means it offers a different kind of enjoyment than games designed to be challenging. If you want to feel like you can lose regularly, I think going for multiplayer is the best option. I haven't seen any single player games where the AI would really stand out as a serious challenge. Instead it's usually given bonuses and the players who play at those difficulty levels have to figure out the puzzle of how to catch up to them. That's the best way to look at getting a challenge out of a single player game in my opinion.

    I played SpaceChem years ago and I still haven't finished it due to how difficult some of the puzzles can get (I got to the last level then decided to take a break). Then of course you can also optimize your solutions further, so there's a lot of challenge to be had there. Puzzle games are in my opinion the most challenging single player games around.

    As someone quite familiar with AI, neural networks, etc., in the end it's just the result of human planning and coding. The amount of time it would take to make an AI that could truly challenge players in a 4x game isn't worth the effort at the moment. Determined players will always have far more time to plan out their own plays than a game AI would ever be given, and not only that, they can also consult with other players to find any weaknesses the AI has in no time. Yes the AI could be better but the reward just isn't there. And when someone in the industry feels there really would be a reward for that kind of investment, whether an indie developer or a big studio, they'll definitely go for it.

    It'd be great if Civ could satisfy both optimal play and a relaxing gameplay style, but that would require a lot of effort for something that most players probably aren't interested in. Plenty of people, I'd guess most of the civ audience, enjoy putting a plan together and following through with it while getting some surprises along the way, and Civ is great for that.
     
    Last edited: Feb 13, 2019
    Wizard-Bob likes this.
  9. Gorbles

    Gorbles Load Balanced

    Joined:
    Nov 24, 2014
    Messages:
    2,220
    Location:
    UK
    Fair enough! :)

    I definitely would prefer the latter over the former. I mean, personal preference aside, it's increasingly hard to market games that don't bring something "new", because new things are still what sell. Improved AI (for a particular margin of improved) just . . . doesn't. I don't know how to solve that. Civilisation is still a massive seller, so despite concerns here, and concerns highlighted by you about Firaxis' understanding of their audience, they're obviously still doing right by a majority. Even with the current (competitive) 4x offerings (Paradox being a big one, but it seems to me there's a definite increase in interest even at the indie level for turn-based strategic gameplay).

    It's hard to discuss this for me without repeatedly referring to the business side of things. It's the part I like least about software development in general, but urgh, I don't know. Aside from being completely idealistic about things (which, funnily enough, a lot of folks used to assume of me for simply having a positive opinion of developers), it's one of those Things that have a real impact, but just isn't really discussed or even acknowledged. Perhaps understandably so, because it can easily become a kind of catch-all "well that won't happen because of business". But hey. I'd take that over "the developers are content with a product I think is subpar" (not you, generalisation there).
     
    ChocolateShake likes this.
  10. Aristos

    Aristos Lightseeker

    Joined:
    Dec 11, 2001
    Messages:
    3,575
    Gender:
    Male
    Location:
    Deep inside...
    Funny that just today Strategy Gamer published an article comparing civ 4 to civ 6, and guess what the focus of the author is?

    https://www.strategygamer.com/articles/civilization-4-longevity/

    The title says so much...

     
    Last edited: Feb 13, 2019
  11. steveg700

    steveg700 Warlord

    Joined:
    Feb 9, 2012
    Messages:
    2,932
    I think you'll find at least one or two posts in this thread that describes how many people play Civ merely as an exercise in efficiency. There's no assumption that they will actually lose, even on deity, so it just becomes either a matter of beating a personal best, or screwing around and playing with subsystems to see what fascinating new quirk can be uncovered.

    And that's good enough for them, which is all good and well. What's less palatable is the notion is that everyone else should stop persisting with the unreasonable expectation that the game's developers have the time and resources to develop their game. Or persist, but be consigned to a subforum (which is a passive-aggressive way of declaring a topic of discussion off-limits in the main forums).
     
    Civrinn and Aristos like this.
  12. ChocolateShake

    ChocolateShake Chieftain

    Joined:
    Dec 28, 2016
    Messages:
    517
    Gender:
    Male
    It's an interesting article but I think he should mention the difficulty levels involved, he says he played against the standard Civ 6 AI, I would think of that as prince?

    He seems to play the same way I do, going light on military until it's needed, and I've certainly had the AI in Civ 6, as of Rise and Fall, succeed at attacks when I go too far (I tend to play from emperor to deity). When I made units cheaper and stretched out the eras, I get to see huge armies and navies. And if I want to make it more challenging, I can also play with immersive eras to force the AI to get all the relevant techs to stay up to date militarily.

    The AI definitely has an easier time just throwing together stacks of doom. No need to worry about terrain and positioning reduces the tactical complexity a lot. And cities actually had to be protected rather than getting a free strike. I don't like the development of cities that can defend themselves from Civ 5.

    I much preferred Civ 3's combat to 4's actually. My impression of the AI's combat ability in Civ 6 is that it's decent, but could use more bonuses to help it get around the challenge of the human's ability to exploit ranged attacks, placement and terrain.
     
    Last edited: Feb 13, 2019
  13. steveg700

    steveg700 Warlord

    Joined:
    Feb 9, 2012
    Messages:
    2,932
    True, the preservation of Civ's traditional victory conditions does Civ no great service in terms of encouraging a variety of strategies. I would think that it might be better served with the "grand strategy" approach of just setting up a bunch of smaller, focused goals or "quests" to accomplish.
     
    Casworon likes this.
  14. Casworon

    Casworon Chieftain

    Joined:
    Aug 9, 2014
    Messages:
    171
    I also strongly agree. A victory point method would work much better i think. Where there would be numerous ways to build up victory points throughout the game. Allowing you to focus on different elements of the game at different time periods. Perhaps building a wonder at one point. Capturing a neighbors capital another. Or focusing on scientific discovery later. All going into a victory point pool that will then win the game
     
  15. Gorbles

    Gorbles Load Balanced

    Joined:
    Nov 24, 2014
    Messages:
    2,220
    Location:
    UK
    Literally nobody is saying this :)

    That suggestion seems to be a rather uncharitable rewording of "there isn't an infinite amount of time to develop a video game to the extent it will satisfy all players". Which still isn't an argument made, but certainly one I was driving at. If you have a counterargument, please do make it instead of doing whatever this is. It's been a solid discussion so far.
     
  16. Anno

    Anno Chieftain

    Joined:
    Aug 6, 2016
    Messages:
    82
    Location:
    Columbus, Ohio
    I think it’d be best to just remove the concept of victory. Or at least make it some sort of incidental thing. Allowing the player to set their own goals and play to them is the way to go, imo. Support it with a ton of ingame achievements to give people some direction.

    I don’t really care for Paradox games as much as Civ but I think they have that concept in a much better place than Civ. At least for my tastes.
     
    Aristos, Gorbles and ChocolateShake like this.
  17. ChocolateShake

    ChocolateShake Chieftain

    Joined:
    Dec 28, 2016
    Messages:
    517
    Gender:
    Male
    It's an interesting idea, I was just thinking the other day, maybe there could be some sort of happiness victory, this could be a way to encourage a different style of play. Setting my own goals sounds like fun, it's basically what I do already, and I enjoy role-playing or setting limits for myself in other ways. I think the devs are going in the right direction by adding in more victory types.

    Modifying each victory type to have a few different flavors would be great too.
     
    Gorbles likes this.
  18. steveg700

    steveg700 Warlord

    Joined:
    Feb 9, 2012
    Messages:
    2,932
    This discussion does not revolve around your personal notions of what constitutes a solid discussion. Literally nobody is saying any of that, of course, that's just a reasonable (if not charitable) way of interpreting your current demeanor.

    The counterargument is the manifest one: literally nobody is suggesting that a video game be developed to the extent that it will satisfy all players. You weren't saying that, but that's what was nestled behind the tactfully-placed smileys. The expectations that aren't being met by various posters aren't lacking in points of commonality, and many of the common points are not unreasonable. I anticipate that any response will likely be in the way of putting me to work to ply you with specifics, but I've already supplied the one stated at the beginning of this thread. If they just focus on something specific like the use of airpower, that would be some gesture that expresed concerns over AI competence was of importance to them. If they can't improve everything, just improve something that most people agree is a problem.

    If the issue is that it isn't worth the time to improve existing systems, rather that developers can only focus on the bottom line of deploying marketable new features or expanding to new platforms, then that seems to be the position of a business not performing a regular series of passes on quality-of-life enhancements. If quality is not a high priority, not all consumers will shrug and give you a pass.
     
    Last edited: Feb 13, 2019
    Aristos and Civrinn like this.
  19. leif erikson

    leif erikson Game of the Month Fanatic Administrator Supporter GOTM Staff

    Joined:
    Feb 2, 2003
    Messages:
    25,608
    Location:
    Plymouth, MA
  20. Gorbles

    Gorbles Load Balanced

    Joined:
    Nov 24, 2014
    Messages:
    2,220
    Location:
    UK
    Correct. But likewise, if change is to be made, this "pressure" I was talking about earlier in the thread would need to be leveraged in such a way that it actually makes an impact. Placing it at Firaxis' feet isn't the best way, from my perspective.

    You mentioned air power. Other people have mentioned air power. Why haven't Firaxis addressed air power? It's a question that has no answer. Unless someone here, working at Firaxis, willingly broke NDA, gave us the precise answer, we wouldn't know. We can't know. So why not debate what we do know? You say not all consumers will give the responsible parties a pass for not polishing a premier product (and I agree). Why should I give these constant unhelpful attacks on the developer a pass? I've argued why they're not helpful. I could argue that they polarise debate further, preventing constructive discussion of the actual issues that are annoying you.
     

Share This Page