Civ VI is SO close to greatness - A call to the developers

[Victoria]

The problem is, the game does not have a minimally competent

When I was young my dad had a patient that was an ex NZ chess champion and grandmaster.
I went to his place from the age of 5-8 most Saturdays and never won a game.
You triggered a memory because on my first meeting he said “we will just play a game and not talk about it. I will only discuss a game with you when you become competent”... I was always to scared to ask after that despite being best at school.

Love your post but I think we disagree on what minimally competent means.

I also think a game where the AI has no memory means they are unlikely to ever be competent in many people’s eyes including the author, and memory is a hard thing to code well for a game like this. While the AI cannot remember seeing you you can abuse it.

 

[Victoria]

War on multiple fronts. Strategic battles with archers around a river and having to strategicely plan and maneuver to overrun their stronghold. It really feels like a tabletop war game in those moments and i can't explain you how much i love it. But it's so rare it's sad.

And that happens in which good AI game with tactical abilities... and please do not say EU again.

I really don't get why people in this forum can't grasp the concept of difficulty options.

Everyone can grasp that, not very nice thing to say. Do you think Firaxis (who have stated they have an AI developer) do not try? Have you ever considered you may be underestimating the difficulty of increasing difficulty.

Memory is literally the main thing besides the main processor a computer/AI has.

You are not a stupid person so why say this sentence? You do yourself no favours.

I have 2 destroyers a battleship and an admiral and so does the AI. Tell me how many possible moves does the AI have to consider to project 2 turns into the future? I am just not sure who is grasping what anymore but I do not envy their AI chap.

 

[Gedemon]

Wrong. I'm asking for harder difficulty options or an option to make the AI more potent in war by option.

That means coding multiple AIs for the same game, one per difficulty level, that would be an enormous task.

 

[Archon_Wing]

You can tell the AI sucks, because the people that defend it can't even find a single reason why it does work, and instead simply attack the motivation/preferences of those that do. Or simply put programming on a pedestal. Yea, well, no. If a cook screws up, he doesn't get to tell the customer "but it's really really hard to cook". An certainly imagine an engineer saying "but bridgebuilding is really hard" when it collapses. These are rationalizations. As harsh as it sounds, nobody cares.

I mean I can only speak from experience, but I could tell the AI was stupid @ prince. People told me to move up to King and Emperor, and so I did. No dice. Immortal, and it's all the same things. And yea Deity too, w/e.

It's one thing to say nobody cares about a tough AI. I can also agree with this. I don't care about speedclearing deity either. This is why a company will not spend oodles of money just to please a hardcore fanbase's desire for some extra masochism. Most of my friends would just rather play on a level where the AI doesn't attack them at all, but guess what; that's already taken care of for them.

What does put people off is not just "muh challenge", but dealing with AI that denounce at the drop of the hat, attack you with 1/10 the military, think warriors match up to infantry, and dance around a city like idiots. And all I hear is rationalization. Oh no, Civ 4 and before had it easy, 1upt. Yea sure, one unit per tile makes the AI give you everything for a work of writing. And all of this is evident to even casual players that write steam reviews. What is apparently true is that all the complaints are not just all people that want to play the hardest civ ever.

The funny thing is that many of the things I described above have been either mitigated or fixed. And there will be many things that will be fixed. And who's helping with the fixes? The people who are bringing them up and complaining! The others that rationalize? They wouldn't do a thing to solve the problem, because the AI being inactive during the modern era is only a bad thing if you think it is. Well, they're about as useful as Workshop of the World. Take that however you will.

 

[Victoria]

because the people that defend it can't even find a single reason why it does work

The AI is utterly utterly useless at naval combat. My pet peeve is I walk between 2 city walls and only one shoots me, the AI always targeting scouts as a priority is stupid and I took advantage of this in particular. You will find people are not saying the AI is any good, we are saying there are reasons it is crap.

The people who are bringing them up and complaining! The others that rationalize? Well, they're about as useful as Workshop of the World.

we bring things up all the time complaining about them. I often complain about how bad the AI is but there is reason for it. If Firaxis has not got the message about a bad AI then lord help us all, I think they are sick of hearing about it. There comes a time of desensitisation where they just ignore it because it is always there. We complain about many things and you call us useless? Cmon, be fair rather than inflammatory. A fighter that one shots destroyer fleets, global voting and dip victory, so so many things could be better.

with AI that denounce at the drop of the hat,

There is reason for denouncement, if you are not within their target land, have sent them an envoy and no other civ is bribing them and you are not attacking their mates or have stacked up too many grievances they will not denounce you. Oh maybe you broke a promise also. One of the few things that works against warmongers is denouncement because the AI will stop trading anything decent with you. Denouncement has a cost and is a threat... just because you have not checked the logs does not mean there is no reason. Joint wars was a good example, everyone whinged so I looked into it an in every case I found, one civ paid another civ to joint war.

 

[Archon_Wing]

The AI is utterly utterly useless at naval combat. My pet peeve is I walk between 2 city walls and only one shoots me, the AI always targeting scouts as a priority is stupid and I took advantage of this in particular. You will find people are not saying the AI is any good, we are saying there are reasons it is crap.

I'm not sure that "it is crap" is a universally held opinion.

I actually don't have a problem if people think it's sufficient at all. It's just there are many bizzare explanations like "civ 4 was simpler" that get tossed around everywhere or somehow ignoring modders have done it repeatedly makes much sense. I'm not saying you're not allowed to like the ai or the game, or whatever but some reasoning is just plainly out the window.

we bring things up all the time complaining about them. I often complain about how bad the AI is but there is reason for it. If Firaxis has not got the message about a bad AI then lord help us all, I think they are sick of hearing about it. There comes a time of desensitisation where they just ignore it because it is always there. We complain about many things and you call us useless? Cmon, be fair rather than inflammatory.

I am really not sure how you read it like that. I never said most of the forum doesn't complain.

I am saying the people that did complain made things better, while others that are doing the opposite didn't help. Clearly there are many in the first catgory; not even sure how you'd associate with the 2nd.

Also I'm referring to not just this forum but the whole discussion in general that I've witnessed on here, Reddit, and on Steam, so to localize any malice here would really not be my thing.I really hope you don't think I post on this forum because I secretly (or not so secretly) hate everyone.Now maybe political discussion yes, but about a game? no.

rubbish, There is reason for denouncement, if you are not within their target land, have sent them an envoy and no other civ is bribing them and you are not attacking their mates or have stacked up too many grievances they will not denounce you. One of the few things that works against warmongers is denouncement because the AI will stop trading anything decent with you.

You missed my point. I'm saying diplomacy and the rest of the AI actions have been improved.... And here's the plot twist. They have been making the AI better so it's clear where they stand on the matter.

The funny thing is that many of the things I described above have been either mitigated or fixed. And there will be many things that will be fixed.

^
Therefore keeping this issue to light does help. Without user input, it is much harder to fix these things. Because your players have played the game much more than you have, after all.

 

[Mahi]

To be fair once you've beaten the main army in a MP game it's just steamrolling from there on. The same applies to SP, just even worse since the AI handles first encounters badly. There's just no way of coming back once your army is wiped out. The flaw lies in how the game denies a player for any comeback. Being a human or computer player.

 

[Boris Gudenuf]

To be fair once you've beaten the main army in a MP game it's just steamrolling from there on. The same applies to SP, just even worse since the AI handles first encounters badly. There's just no way of coming back once your army is wiped out. The flaw lies in how the game denies a player for any comeback. Being a human or computer player.

This, I think, is less an "AI Problem" then it is a problem with the One Path Linearity of the game. To not steadily progress is to Lose. Period. In what I consider one of the most egregious instances, in the game (if not Dido) to lose your one and only original Capital is to lose the game. What makes this 'way past annoying for me is that you'd be hard pressed to find any Civ in the game that ever had an actual capital that didn't lose or change that capital at least once in their history. It is completely artificial mechanism that serves only one purpose: it makes defining a Domination Victory very easy.

On the specifically AI Front, I got corrected (correction: I got whacked down like the target in a Whack-a-Mole game) when I argued that AI shouldn't be that hard and some real programmers weighed in. Two illustrative examples come to mind.
First, for the past 30 years they have been talking about real "Artificial Intelligence" and about how it was just around the corner. Well, it turns out that it's not a corner, it's a Moebius Strip and you can keep turning rounded corners forever. Basically, the simplest things that "Live" Intelligence does are very hard to program, and having a program 'teach' itself by trial and error works only up to a point, and that point is far more simple than any Civ game.
Second, compared to Civ, Chess is a really simple game. All "units" or pieces have absolutely fixed paths, there is a small number of possible "Tiles", and almost no variation among the tiles (some have uncrossable edges, some don't - that's it) and yet it took years and a lot of high-caliber people to come up with a program that didn't make really idiot mistakes. And the Master Chess Programs which are now able to beat even the best human opponents took the resources to develop, and require computational power, that is simply not likely to be available for a commercial computer game.

I agree the AI could be better: I am not so sure that focusing on the AI will result in a game that is both playable and affordable. As @Victoria and others have pointed out here, many of the "problems" with the AI are in fact not AI problems at all, they are designed in to give 'artificial' limits on what the human player can do: pick a fight with your hapless AI opponent, and Never Trade Again, or at least, not for an Era or so. - Cheap alternative to spending the time and resources to develop a 'competent' tactical AI, even if the in-game result can drive you to occasionally beat your head against the keyboard in frustration. . .

 

[Trav'ling Canuck]

To be fair once you've beaten the main army in a MP game it's just steamrolling from there on. The same applies to SP, just even worse since the AI handles first encounters badly. There's just no way of coming back once your army is wiped out. The flaw lies in how the game denies a player for any comeback. Being a human or computer player.

Agreed. Which is why I disagree with the premise of the OP that Civ 6 is anywhere close to greatness. The economic system is less interesting and less balanced than previous versions of Civ, and it's not any better than any other 4x game at dealing with the snowballing impact of winning early wars.

As long as the game is designed with a 4x lens, perhaps it'll never be more than a prolonged way to reward victory for actions taken in the first 50 turns, as each subsequent decision has a smaller and smaller impact on the outcome of the game. A game where winning wars is beneficial but not determinant of the final outcome would be more interesting to me, within the context of the grand historical scope embraced by Civ, but is contrary to the nature of 4x where more and bigger is always better.

 

[Gorbles]

@Steve Eric Jordan - Steam reviews are not the universal rule of what is right or wrong with a game, and shouldn't be treated as such. Note: not saying they're not useful. But you're acting like this is some huge obvious thing that Firaxis haven't considered (or 2K haven't applied the pressure for, assuming a typical publisher / developer breakdown).

You can tell the AI sucks, because the people that defend it can't even find a single reason why it does work, and instead simply attack the motivation/preferences of those that do. Or simply put programming on a pedestal. Yea, well, no. If a cook screws up, he doesn't get to tell the customer "but it's really really hard to cook". An certainly imagine an engineer saying "but bridgebuilding is really hard" when it collapses. These are rationalizations. As harsh as it sounds, nobody cares.

You've made this a tad too easy, inbetween all the arguments that nobody made, but uh the AI works because each AI faction does something each turn and the game doesn't implode whilst doing so.

On the surface of it, this sounds dismissive. I'm being serious. We're not discussing whether or not the AI works. We're discussing whether or not the AI works for an increasingly wide range of people in this thread who are to varying degrees speaking for the experiences of a bunch of other people on other sites and in other communities, across differing demographics. If we're debating simply "does the AI work", the answer is a tedious but resounding "yes".

If you don't want to care about the why, then there's not much point to debating it. AI is complex. Deal with the nuance involved, or risk making yourself look silly. As most people do when invoking allegories involving completely different disciplines like cooking (conveniently around someone who's entire immediate family is in the restaurant business, yours truly)

 

[DJ_Tanner]

People continue to project their opinion as some vastly held belief of all players, particularly around here where you have a vocal and hardcore base. This group here, are NOT the main playerbase, they are certainly a very active playerbase, but what happens around here is not indicative of the main base who are far, far, far more casual in their gameplay.

Wrong. I'm asking for harder difficulty options or an option to make the AI more potent in war by option. Nothing would change for casuals hopping in on Prince. I really don't get why people in this forum can't grasp the concept of difficulty options. Everyone can choose the one that fits him best.

I don't think anyone is, or has ever been against having options for harder difficulties or options given it was just a switch here or there to flip. The issue arises if those harder difficulties come at the expense of other items. I don't get why people on this forum can't grasp the concept that resources are limited and there is a cost-benefit to adding anything. To make a "fully potent" AI would require other items don't get addressed, and that would be an issue. Again the number of people that see this is as a deeply needed feature, while very vocal, are not the core playerbase, the core players want new features/systems and upgrades to those currently in the game over having a more rough and tumble AI. If they have the ability to "fix" the AI after that, then by all means, no one is going to be upset by that, but the focus is certianly not, nor will it be anytime soon, on the variety of AI levels.

Yea, well, no. If a cook screws up, he doesn't get to tell the customer "but it's really really hard to cook". An certainly imagine an engineer saying "but bridgebuilding is really hard" when it collapses. These are rationalizations. As harsh as it sounds, nobody cares.

But the thing is the AI works, it may not work to how you want it to, but yes it most certainly does work. It may not do exactly what YOU want it to, but it is in no way "broken" or inoperable. This isn't a cook screwing up, this is you walking into a McDonald's and being upset that the McRib you ordered wasn't a perfect mid-rare, or yelling at an engineer who built a perfectly structurally sound bridge that it is awful since it doesn't also include a pedestrian walkway.

 

[Acken]

It's not just the AI though. Balance is terrible and many systemms half baked.

 

[Victoria]

sorry, nothing personal.

Yup, one of those people that says “no offence” as they offend you. Got it.

 

[Aristos]

That means coding multiple AIs for the same game, one per difficulty level, that would be an enormous task.

Incorrect. See how Gazebo and Ilteroi solved that problem easily and efficiently for VP.

 

[Casualty of war]

The AI used to only send you a "Go away" note if your troops were right next to their borders; now they will notice if you are 1 hex away. Improvement. They still only detect Unit and not how powerful a Unit, but babysteps.

The AI is moderately skilled at making immediate decisions and gets progressively worse at planning those decisions out. The AI used to continue working on a Wonder while you had their city under siege; now they will sometimes abandon that project to build some defensive Units or Walls. Improvement.

Like chess, your skill at Civ is primarily based on how many moves you can think ahead. Some people can plan out their entire game from their starting dirt, and they finish Deity games in 150 Turns. Some people can map out an Era pretty well. The AI can maybe manage a dozen Turns, enough to move a force into striking position. But this is only after they hit programmed thresholds like doubling your military strength or accumulating enough Grievances, and once your skill at thinking ahead outdistances the AI the game becomes easy.

There are things that the AI trounces most players at. If they have an embassy in your capital, they check every Turn to see what you are up to so think about how much you want that 25 Gold. The AI always knows where you are in the rankings against themselves and the other AI, and be afraid if your military strength compares poorly because they will come for a visit. The AI gets those diplomatic messages too.

An AI could be designed that would be the Deep Blue of Civ, but then only Bobby Fischer could play.

 

[Pietato]

I think you guys think in the wrong direction. Giving the AI free units is not the way to go here. The AI has everything it needs. Production, science, etc. All it needs now is using that to build more units and, when it uses them, commit to their goal. (Killing someone, defending, whatever)
Giving free units to the AI in the start or every now and then would change nothing and is just a weird, unrealistic workaround that doesn't tackle the real problem at all, i don't get how 3 people here think this is a good idea.

Well no duh, genius. We all know a superior AI is better to the lame free units and bonuses the AI gets. We all just exist in this place called reality...

 

[DJ_Tanner]

Lord knows, I'm one of them . But if I had to boil down all of Civ VI's problems into a single sentence, it would be that the Design Team is much better at meeting Marketing Goals than designing a historically-based game. They have a game that sells, in a 'franchise' that is one of the most successful of all time. This particular iteration of the game sucks on many levels, but it does not matter as long as it looks good, Sells, sounds good, Sells, gives the impression of being a historical game to people who don't know history (i.e., the population of the USA), and Sells.

I think it is also an important point that CIv has never set out to be a histocially-based game. Civ is a game with history as a setting, not the basis, and that is a big difference. It lets the game be more causal, much more about the fun experience over the trying to ensure any level of hyper-accuracy. It is and has been always a game first, and you can see that with things like TSL being an afterthought not a standard.

 

[acluewithout]

Can't find a post about difficulty in your signature, mate.

My mistake. Signature corrected. Link here.

 

[DJ_Tanner]

Of course there is such games, i always love to talk about Advance Wars here. It had muuuch better war AI like 20 years ago. And the game being more complex is not an argument for units running around uselessly instead of attacking and the AI simply refusing to build units in a f-ing war. That's a complete seperate part that doesn't have anything to do with the other complex aspects.
And even if there wasn't, how does it matter if this happens in different games with "tactical abilities?" I want it to happen in CIV, there is no reason for it to be done by another game first.

First Civ's war AI is better than that of Advance Wars. It may not be superior in a strictly combat setting, but I always found Advance Wars AI to really struggle with keeping up a healthy economy, boosting its diplomatic ties, and growing its cities during a time of war.

You can't remove the simplicity of the game from the conversation, it comparing apples to an orange grove. Advance Wars had one goal fight and total destruction as the only viable outcome, with Civ there is so much more in the decision tree. Even in a strictly war setting there is more to think about. War in Civ is not, nor should it be, an all or nothing game. Wars can be won or lost without the taking of a single city, there are costs and benefits from many facets in a war that are not shown in the total destruction of the opponent, and if the AI took that approach where it threw all caution to the wind and died on every hill the game would suffer a whole new host of problems. There are plenty of issues to discuss and areas for improvement, but pretending its is as simple as dropping in a 20 year old game's AI is not doing much of anything.

 

[Boris Gudenuf]

I think it is also an important point that CIv has never set out to be a histocially-based game. Civ is a game with history as a setting, not the basis, and that is a big difference. It lets the game be more causal, much more about the fun experience over the trying to ensure any level of hyper-accuracy. It is and has been always a game first, and you can see that with things like TSL being an afterthought not a standard.

On the contrary, Civ has always used History as its data base and source material for just about everything it does. It is Historical Fantasy in its implementation, but the Fantastic elements are, themselves, historical in basis: historical 'Leaders' that are in-game Immortal, Civilizations that keep the same Unique characteristics for 6000 years, but they are Historically based Civs and historically based characters (even if, like Gilgamesh, the history is nebulous at best).
And, as a historian with about 7,000 hours of game time in Civ V and Civ VI, I can tell you that every time they get the history wrong, the game suffers.
That does not, by any stretch, mean the game has to be 'hyper-accurate' (whatever that means - I assume you mean sacrificing everything else in the game for historical accuracy?) - a game that is unplayable or no fun to play is not a game, it's a Punishment, meant to be sold to and used by the Masochistic - which is too small a market to sustain a computer game company, even if you include certain sports fans.
But, when the game misrepresents military history, you get Units that don't perform the way they should (Anti-Cav), or Units that are never used at all (Siege Towers), and when they misrepresent or just don't understand Historical Geography we get coastal cities that don't appear to be worth it, rivers lacking their historical importance to Trade and transportation, Tundra areas that magically support major cities in the Classical and Medieval Eras - which is almost as close to Pure Fantasy as an Immortal Governor . . .
And so on and on: I've been posting on this stuff for years on both Civ V and Civ VI forums, and have accumulated a fair number of positive comments and Likes, which I assume means Somebody Out There agrees with some of what I've written.

Of course it is a game first. More specifically, it is a Commercial Game first, or it is nothing (unless one of you out there has a few dozen millions of Dollars/Euros/Tugheriks to finance a Non-Commercial version that plays exactly the way you want it to!). But it has always been a Commercial game (franchise) based firmly on Historical Events, albeit (the reason I'm on these Forums) all too frequently based on misinterpreted, misapplied, or misunderstood Historical Events and processes.