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

Does Civ 6 have the potential to surpass Civ 4?

Discussion in 'Civ6 - General Discussions' started by labellavienna, Oct 25, 2016.

  1. LDiCesare

    LDiCesare Deity

    Joined:
    Dec 22, 2005
    Messages:
    2,612
    Location:
    France
    It depends how long you expect people to play the game.
    If people are happy with just discoevering a new game and then leaving for something else after less than 50 hours of play, then you don't need any AI.
    If they want people to buy expansions, however, they may want to cater to people who like the game, so those who play a lot of it, not to make bad press so they'll be able to sell expansions.
    If they want to sell DLCs, they need to have people play the game for a long time.
    Currently, once you're beyond the exploration/discovery phase, the game offers no challenge, so it's unlikely to keep people interested and have them buy DLCs.

    And anyway, you don't need a "vastly" improved AI to make the game vastly better.
    You need an AI that can win, and that only requires a few changes.
    In my current game, Arabia swarmed a city defended intially just by a lone hoplite corps. They brought 8 mamluks and never attacked the city. Hadthey attacked, they'd have lost maybe 2 mamluks and taken the city. By jsut moving around and pillagin, I killed mamluk after mamluk withmy field cannons and city defense. The only thing they needed to do was "attack city". It should not be a lot of work.
    In the same game, Russia has 2 great writers and 2 great musicians dancing around St PEtersburg but they haven't built a single culture district, so they can't use them. It is a small improvement to actually use one's great people, but it would vastly improve the experience. I could steal ther great works, ot they could threaten to win a culture victory for instance.
    When I am offered a deal, I can sometimes take everything form the AI, sometimes if I refuse to take part of the deal as in, no I don't need that horse, they refuse. As a result, I can't bargain.
    Fix these three things and, even with poor tactics and upgrading but given production and combat bonuses, the AI may be challenging enough for the game to warrant playing more and buying DLCs/expansions.
    Personnally, if they don't fix the AI in some patch quickly, I'll shelf the game the same way Ishelved Civ 5 and will not buy any expansion or DLC but return to Civ IV or EU IV or whatever else I can find..
     
    Evs likes this.
  2. ahcos

    ahcos King

    Joined:
    Mar 19, 2010
    Messages:
    919
    I do believe you're wrong here. The game's not completely bug-free, but still you can see that it had a lot of fine-tuning and beta-testing. So why do you believe something as fundamental as this AI went by seemingly unnoticed?

    Because it's hard to balance "fun AI" with "challenging AI". The designers want the player to win, after all. I'm pretty sure the designers could write a AI that punishes your mistakes (or at least punish them more than right now), but would that be fun for a lot of players? I'm not too sure about that.

    Fine-Tuning an AI so that the bonuses it receives balance out with it's "intelligence" (read: ability to capitalise on mistakes etc.) while keeping it fun to play against, is very VERY hard. It gets harder as there are more and more things you have to take into account. And there are tons of things to take into account for an AI in Civ6.

    Would it be easy to write an algorithm that enables an AI to abuse it's bonuses to achieve a Science-Victory in the early 1500's? Well, yes. Would it be fun? Nope, so you have to restrain it some other way.

    Sure, the stupid AI is (and always was) ricidulous at best, annoying at worst for ALL the Civ games. The question is how you build a game around the fact that you have to balance the AI's power against the players fun. And here's where Civ6 falls short, mainly due to 1UPT. More clever people than me have shown why it is such a big problem for the AI in the past... it's just not an "easy fix" if you take everything into account. I guess they'll release a patch so that the AI isn't as bluntly ******** as it is right now, but bury your hopes for anything remotely close to a "clever AI" in Civ6.
     
  3. rezaf

    rezaf Warlord

    Joined:
    Dec 3, 2003
    Messages:
    179
    Totally agree.

    I also feel making the AI challenging is probably not even on Firaxis' agenda. Before Civ5, starting as low as Warlord, going to the next highest difficulty basically required a rite of passage, only CivFanatics (pardon the pun) had any business playing on the highest ones.
    Civ5 became the game where every casual player can "master" deity - mostly due to AI ineptitude in warfare caused by 1UPT.
    But due to the fact that they can now win the game on the highest difficulty, those players feel like a friggin field marshal for outsmarting the AI by pulling off grand tactical maneuvers such as covering their archers with melee units or using their catapults to bombard cities. Everybody is a winner!
    What incentive is there for Firaxis to change this given to the success of Civ5 and now Civ6?

    Besides, I'm still not convinced the AI CAN be improved to a degree it'll be able to competently handle 1UPT.
    Stack based combat was like clubbing someone with something heavy - the bigger the stack, the heavier the hit.
    Teaching the AI to deal with such combat was like teaching a gorilla to use a stone.
    1UPT combat is, at least theoretically, much more intricate - it's like using an assault rifle.
    The AI is still the same old gorilla though - it only tries to club you with the rifle as if it were a big stone, and the human player stands there smiling ... and pulls the trigger.
     
    ahcos and Roald Amundsen like this.
  4. LDiCesare

    LDiCesare Deity

    Joined:
    Dec 22, 2005
    Messages:
    2,612
    Location:
    France
    Are you serious? I stated a FACT. If you don't believe me, then just play the game and see it for yourself. Or read what everyone else says: Building 5 or 6 archers wins you the game on Deity.
    The AI is never able to take out a city after the ancient era. It only leaks cities when losing a war by giving them in the peace treaty.
    So, I believe this fundamental AI being a piece of **** went unnoticed because it IS totally crap.

    Yes, it's hard to balance fun and challenging.
    But the devs failed at both.
    First of all, people playing lower difficulty levels expect less aggressivity from the AI and barbs but it's still there. That's a basic misunderstanding. Lower difficulties demand that AI and barbs be less aggressive, it doesn't seem to be the case.
    Second, when I play onDeity, I expect a challenging AI. It is not even able to take a city!
    Do the designers want the player to win? I hope they don't. Then again, I love roguelikes. When I play on the hardest difficulty level, I expect to lose AT LEAST every other time. Otherwise, where is the challenge? I'm not asking that low level AI be able to perform. I'm asking that deity level AI should be able to perform at least the most basic tasks. It can't. If there's no challenge even at the highest difficulty level, then there is no point in having difficulty levels at all.

    That's totally irrelevant. I ask the AI be able to do ONE thing, namely attack cities. It's not very VERY hard. It's very very easy. I've coded AI. My job is selling software simulation to the military where units will attacktheir opponent automatically when they have a superiority according to doctrine. I probably know better than most what coding an AI is about.

    Why would the AI have to have bonuses in the first place? They gave bonuses because it's an easy cop-out for not making competitive AI. They give it a starting advance to help it. But it still does nto work because it fails at the most basic tasks such as escorting a settler and taking a city.

    No.
    I never bet Civ IV on deity. I bet Civ VI on deity in less than one week without cheese or even using what I think would be an optimal strategy (read that I'd use in MP).
    There is an order of magnitude between civ IV (hell, even V!) AI, where the AI can capture cities even after walls have been built and the Civ VI AI.

    It's not due to 1UPT. Having 6 stacks or 6 units around a city and attacking with none has nothing to do with stacked combat.
    1UPT makes it harder to code the AI, but the AI could still be somewhat challenging even with it.
     
  5. ahcos

    ahcos King

    Joined:
    Mar 19, 2010
    Messages:
    919
    @LDiCesare

    Well i can't say you're wrong or anything, but i still do believe that you're not taking everything into account here. Yes, the AI is flawed, but maybe (just maybe) it's not that easy to fix it.

    Say the can take a city of yours, but doesn't do so - maybe there's some algorithm that checks the number of cities, and this prevents it from taking the city (just like it would prevent it from building a settler). Maybe it doesn't even want to take and keep the city, and is hindered by the fact that the AI is coded to not raze cities.

    I could go on here, but i think you can see my point.

    It's hard to judge what the underlying problem is in such cases. Sure, often enough it might and will be a badly coded AI, but in other cases the AI might be hindered by things that aren't even related to it's ability to conquer cities, but rather that "conserve" the main aspect of the game: player's fun.
    Civ6 is catered towards a casual audience, just like Civ5 was before. Keeping the AI in check is a central aspect of the game's design. That might lead to really weird choices the AI makes. I'd like to reference Sulla's article at this point: http://www.sullla.com/Civ5/bnwreview.html - the part where Civ5's lead designer Jon Shafer is quoted, talking about Civ5's AI. It's the same underlying issue here.

    Now again, i'm not saying that the AI is NOT flawed, i'm just saying that it's flawed because of things other than just being "bad", and maybe also underlying game design decisions. Therefore, it's not an "easy" fix. You can't just make the AI attack cities without taking everything else into account.

    Civ4's AI was just as stupid, that's the sole reason players beat Deity occasionally (me too with the great help of this very forum, still very proud of that), it's just that the underlying mechanics worked hell of a lot better in Civ4 to cover that up.
     
  6. LDiCesare

    LDiCesare Deity

    Joined:
    Dec 22, 2005
    Messages:
    2,612
    Location:
    France
    Yes, I see your point. I disagree. I've coed games. I've worked for a game company. I've been developing software for years.
    I'm also a customer who has played lots of games and has never seen, honestly, such a poor display of an opponent.
    The AI which wages war and leaves tons of units around the city doing nothing but dying is worse than useless. I mean, seriously, if the units stayed home and just defended, they would cause more damage because they'd heal. The AI currently manages to be worse by doing something than it would be by doing nothing. The ai would be better if it never moved its units!
    There is never any reason not to take a new city. There is no game mechanism making it desirable not to have one more city. Sure, it will not be happy if it's big, but guess what, they were not big cities.
    The AI is coded based on the underlying game mechanisms. You can't separate them.
    Releasing such a flawed ai is just not acceptable.
    Any time the ai has a unit parked in a city, it never attacks. Why? To make it fun for the player? If so, why does the ai even build military units?

    I understand the target audience is to be large, but that is no excuse for giving such a pitiful ai.
    The flaws are not that big, just a few things where it should be attacking and is not. This should never have passed QA imo. Sure, you have to play a few hours to see it, but that's what QA is for. Make sure that end game exists, make tests such as reloading a game or a scenario which is near to end game and checking the behavior of :"In 10 turns, that city should fall".
    Players don't want to lose, but they want either a sim game or a challenge. In the former case, they can by disabling barbarians. In the second case, they simply cannot.
     
  7. ahcos

    ahcos King

    Joined:
    Mar 19, 2010
    Messages:
    919
    Well yes i do see your point, it's just that i doubt that something like that just went by unnoticed, when everything else about the game seems pretty polished. Am i dead wrong here? I've only watched streams so far, mainly because i don't have much time at the moment, but also because i don't want to play Civ5 3.0 with the horrible, horrible no-fun-to-be-had warfare 1UPT brought to the franchise.

    So, i kinda doubt that this just went by without being noticed. I just don't believe it. They must've realised that it's not that simple to fix, otherwise they'd have done it... or is there something i'm missing here?

    On a side note, Civ6 doesn't seem to be much of a challenge either, much like Civ5. If you want some challenge, maybe the franchise just isn't worth playing anymore, idk.
     
  8. LDiCesare

    LDiCesare Deity

    Joined:
    Dec 22, 2005
    Messages:
    2,612
    Location:
    France
    It is something you won't notice until you play a few hours. The ai will take cities before walls are up. The barbs will also be quite annoying.
    So it needs some work to find out. You have to be losing a mid-late game war (or battle) for it to appear. I am not sure that's been seen.
    There are other gross errors. For instance, 2 out of 3 or 3out of 4 starts on an inland sea map don't dhow most rivers. That's gross, but you have to run a peculiar game and actually play it before you realise you were not unlucky to start without river but there's a display bug. Still, it means that half of one of the 5 provided maps is unplayable and this went through QA.

    I think the QA and polish all went on visuals and sound, but not on gameplay. Some people found lots of 'TODO' in the script files meaning that a lot of things are unfinished, and these are gameplay or UI related.
    The team likely had a deadline and were short, so they cut a few features (maybe renaming cities, probably build queues) and left the ai in an unfinished state.
    But overall, I think most of these problems can be addressed in patches. And I would also say let them have some time to do the patching. Probably, many code changes were made at the last minute, causing regressions but these regressions could not be fixed because of the deadline. Patches don't have a deadline, so they can work with less pressure and deliver something neat, hopefully.
     
  9. UWHabs

    UWHabs Deity

    Joined:
    Oct 10, 2008
    Messages:
    3,999
    Location:
    Toronto
    The fact that I've put in more hours in civ 6 than in Beyond Earth speaks to what potential it has. Still a ways to go, certainly, but if we're also talking of the lifetime of the game, through the usual 2 expansions and multiple patches, there's still a lot of time to improve things.

    My hope is that in the next couple weeks, we'll get the first patch to fix up the broken elements (stuff like selling units, major balance issues, fixing the worst AI mistakes), then sometime maybe early in the new year we get the first patch to really tighten up the AI and really make it a challenge.
     
  10. GhostSalsa

    GhostSalsa Emperor

    Joined:
    Jul 7, 2012
    Messages:
    1,010
    Location:
    Austin, TX
    If we're making 1-upt, incompetent AI invasions, and overpowered archer focus fire, the biggest issue of VI, then we should get back to the only other solution envisioned for V besides the impossible "smarter AI": innovate the terms of invasion to move battles to the open field.

    This is what warfare in VI needs to be interesting, beyond the here-lamented hex puzzle aspect (though I think that part is still very fun in V if you engage in offensive wars while lagging or equal in tech): It needs to stop being about units getting creamed near cities and narrow terrain.

    It needs a solution like Endless Legend's expanded battles, though not that one, because in practice it turns out to be too detached from game flow. But a new something

    Such as:

    The AI uses its armies to encircle your empire and pillage cautiously, essentially an empire-wide siege

    Or:

    There is no city attack and no more city capture, except for capital capture. Instead cities are just pillage-able tiles like any other, and the invasion behaves as army v army. Cities (aside from capital) do not actually trade hands until peace deal, so both invader and defender need peace terms to resume use of the cities. Human defenders can be immensely threatened by AI carpets in a very short time, but still able to recover progress quickly if they repel the army (pillaged city tiles self-repair for original owner, minus population).

    Or such thing. 1-upt isn't bad for civ, it just creates a lot of problems that need to be solved. VI didn't focus on the problems enough before release, and war hugely sucks rn.

    Edit: Oh yeah, AND:

    Archers should only work defensively, like anti-aircraft do, zapping the first unit to attack a nearby friendly unit. This will make more interesting specialization and promotions (anti-archer army or anti-melee army), but still work with 1upt before stacking options come online.
     
    Last edited: Oct 29, 2016

Share This Page