Unpopular opinion: people that blame AI don't play on deity

[Ziad]

The question of AI competence is only really an issue for a very small subset players, which is why they don't invest in it too much. While it's easy to say "just make it a toggle", that would be a lot of development time spent frivolously.

It's easy to forget that the average player only plays a video game for a few dozen hours if that. Someone who spends hundreds of hours on one video game will notice all the quirks and balance issues, which can be annoying, but at that point you've pretty much exhausted the entertainment value offered by the product.

It's understandable, if frustrating. Frankly, this is something that the community is better off creating themselves.

The AI+ mod is pretty good, and maybe at some point a Community Patch project will begin for Civ6... most likely when it's feature complete.

First, the game is currently too easy on all levels.

No, it is not, especially without hundreds of hours in experience and the use of exploits like chopping overflow.

 

[AmazonQueen]

I suspect the people who play on diety are the ones who complain about the AI most. They're the ones who have figured out the various challenges presented by the mechanics, care about winning efficiently, and only have the AI left to provide any challenge. Without an effective challenge from the AI speed of victory is the only challenge left.

I don't play on diety, enjoy playing against myself as much as against the other civs, and wouldn't consider myself an expert player. I do occasionally get annoyed when the AI does something particularly silly but most of the time my own decisions limit my success more the AIs inadequacy.

I'm not convinced that people who play on diety are the best ones to judge if the game is too easy on lower levels of difficulty. Too easy for them perhaps but not for the people who play on those levels?

 

[TheDouche]

From reading your posts guys, I understand now that by "better AI" people probably mean a better interacting AI, that will make logical trades and political decisions, good usage of spies and actually pursue their win condition.

I reckon most people don't want a perfectly overflow-abusing AI, that will use its deity advantages to win by turn 150, but it would surely be interesting to see.

I still believe that although it's hard, an AI better at winning than 99% of the players is completely doable (there's not 1000 different decisions every turn as someone said), probably by replicating the playstyle of really experienced players (plenty of those on these forums).

All in all, better does not equal fast-victory for the game experience of most. Let's hope the new patch not only adds diplomacy, but improves the player to AI interactions. Cheers

 

[cvb]

The AI+ mod is pretty good

But whatever you do now the AI will interfere with its random moves. So we need finally the DLL sources or at least an additional special modified DLL, which doesn't move units at all, so that unit movement can be done completely via Lua(?) - just for single player.


Edit:
TheDouche: "better interacting AI, that will make logical trades and political decisions, good usage of spies and actually pursue their win condition"

as well as improved unit movement on tactical & strategic level (and as mentioned already: USE ALL units at all).

 

[Ashbery76]

Because people want to play an A.I that can use the systems well and not massive cheats that make the game unfun because they are not really playing the game.

 

[MooFreaky]

This clearly doesn't apply to everyone, but I think it stands true for most.

I'm not a troll and I am in no way trying to stirr things up, but I feel like the truth should be said.

If the game had a decent AI, 90% of the players couldn't enjoy it. I know it, some of you know it, and guys at firaxis know it aswell. If openAI can stomp top human players at dota2, a game requiering way more complex decisions per minute, it won't be that hard to do it here.
Heck, they could only use neural networks for the first 100 turns to get the lead, cutting complexity massively, and then use some pre-programmed easy to replicate science victory scheme.

This will probably lead at AI being better than 99% of the players, without the need for bonuses. But It Would Suck! Like my last "cultural peaceful" deity game did when Nubia blessed me with 5 pitatis and 3 swordsmen by turn 40. Couple reloads later I still didn't find the way to get out of the situation and still have my wonder. Sigh

I find all the complaining ridiculous. Some people wouldn't buy gs because of AI planes, but I don't think they love the game enough to play it at all. Maybe people should enjoy their overly-easy deity games, and move on. Just my two cents.

You are making a few statements that are incorrect here.
Firstly, regarding the DotA AI. This is something that is getting passed around the internet without people actually telling the whole story. The AI was beating top players at 1on1 games, not team games AND that changed once the players got used to the idea. Once players started getting used to the AI and realised what it was capable they started adjusting and then began winning over the AI substantially. The AI's advantage early on was that it used tactics that the humans weren't used to (because they are tactics that can only be used by someone with precision of an AI, such as allowing yourself to be deliberately damaged then retreating the perfect distance to keep out of range by millimetres with perfect control, then countering with a nuke combo).
That isn't to say the AI isn't amazing, but let's tell the full story here.

Second no-one is saying they want an AI that will absolutely destroy a human player. Merely, that they want an AI that is actually capable of basic functions like taking a city. Right now the AI will shuffle units over and over, to the point that it is incredibly hard for it to even take a well placed city. If you have 2 units defending a human player will almost never lose a city even against overwhelming numbers.
It's about an AI that doesn't need such massive buffs, but can play more intelligently.

I suspect the people who play on diety are the ones who complain about the AI most. They're the ones who have figured out the various challenges presented by the mechanics, care about winning efficiently, and only have the AI left to provide any challenge. Without an effective challenge from the AI speed of victory is the only challenge left.

I don't play on diety, enjoy playing against myself as much as against the other civs, and wouldn't consider myself an expert player. I do occasionally get annoyed when the AI does something particularly silly but most of the time my own decisions limit my success more the AIs inadequacy.

I'm not convinced that people who play on diety are the best ones to judge if the game is too easy on lower levels of difficulty. Too easy for them perhaps but not for the people who play on those levels?

Actually we are. Because players don't want the AI to cheat, they want an AI that scales on skill not front loaded advantages. On deity you lose many aspects of the game because there is no way to match the frontloaded benefits of a computer. Even if you do an optimal build it's not even worth rushing the early wonders, or taking a risk on a cultural or resource based opening because you know that you will always be attacked so you need to go military early. That is boring and that is the problem. So it's like the same game on prince as it is on deity, where you can play with a wide variety of strategies because the AI will be smarter and more responsive, rather than just having extra cities and units thrown at it early.
No-one is suggesting that the lower difficulty players should have a harder time of it, just improved competence at actually functioning.

(there's not 1000 different decisions every turn as someone said), probably by replicating the playstyle of really experienced players (plenty of those on these forums).

You'd be surprised. Actually I think there would be more (probably well more). Each unit has 19 possible positions it can end a turn (on flat terrain, and including remaining stationary), and when you factor in future pathing, potential positions of enemy units etc there would be hundreds or thousands of permutations just from where to effectively place units.
What units to build, what structures, is it better to maintain the current building or switch, rush finish now or later or never, managing tiles being used, purchase of new tiles and the cost/benefit of that etc. There are things that we just play through without a thought, but they are decisions that the AI currently isn't even close to answering. And linking these decisions together into a cohesive strategy where one decision influences the others means that the complexity of those decisions increases further.

The AI in DotA isn't capable of the long term planning on this level, it is far more reactionary to its current situation. That isn't to say it can't be, just that you need to compare apples with apples.

 

[SupremacyKing2]

The game is all just too damn passive. Just passive. Either that, or the game needs to just become a city building sandbox (in which case, can I have my money back please).

The single player game does feel that way. In a lot of ways, civ6 is a sandbox game with the human player in the middle and barbarians, city states and AI civs just there to provide some interactive obstacles.And virtually all of the resistance to the player happens in the early game. Barbarians prevent the player from just spamming unescorted settlers. City states restrict the player from settling new cities in certain areas. The AI civs force the player to invest in some military instead of just focusing on building and founding new cities right away. The key is to focus early on your military. If you neglect your military in the early game, you can face a serious challenge. I've had the AI rush me in the early game and conquer my capital because I had not focused enough on my military defense. But there is little to no resistance after the human player has overcome those initial challenge and has established a secure empire of around 6-7 cities. The external threat is gone. The barbarian threat is mostly pacified. City states do not pose any serious military threat to you. And if you have a strong military, the other AI civs will be rather passive and certainly won't pose a serious threat to you either. If they do declare war on you, they will shuffle units around but won't come close to taking any of your cities ever again. And there is no internal threat since the game lacks any mechanism for representing rebellions or civil wars. Changing governments or policy cards never creates any problems with your people. At this point, you enter the very boring phase of building your empire unopposed for the next 100 turns until you win. It's boring and passive because it's just picking the next logical building, district or wonder based on your victory goal. And there is no challenge or obstacle. Just pick the next item in the city's build queue, wait and repeat. If you are going for a culture win, get a Great Artists/Musician etc every so often and pop a great work, rince and repeat until you eventually win. If you are going for a science victory, beeline for the space district tech (rocketry I believe), build the space projects and wait to win. The only alternative to this is to go for a domination victory since then you at least get some excitement with moving units and attacking cities. But there is a snowball effect where as you conquer territory, you become ever increasingly more powerful and you will steamroll the AI easily.

Let me summarize this way. There are 4 basic components to a civ game:
- Exploration: active but ends before the mid game once you've sent scouts all over your continent.
- Expansion: active but also ends before the mid game once you've settled cities in your space around your capital (maybe 6-7 cities).
- Conquest: active but ends very early in the early game unless you are going for a domination victory.
- Building: Very passive and lasts the whole game. If you are going for a science or culture victory, it is most likely the only part of the game that you will be doing after you finish the other 3.

So basically, the early to mid game involves 3 active phases and the mid to late game involves only 1 passive phase.

 

[Xur]

Watch a reasont twitch stream where the devs landed a settler on an island with 1 defender... it was surrounded by barbarians... the barbs attacked with ~2... then backed off even though it was an easy kill for them with that many barbs... things like this annoys me... it's super simple.

 

[PhilBowles]

No, it is not, especially without hundreds of hours in experience and the use of exploits like chopping overflow.

I've played 400 hours in Civ VI. I didn't even know exactly how the chopping overflow exploit worked until watching let's plays a few weeks ago, and I've never used it. Almost all of that has been spent playing Deity, since I learned very early that the game was unplayably easy on any other difficulty and even on Deity at release I won without understanding mechanics like Amenities and by essentially random-walking through the tech and culture trees.

The difficulty has increased a lot since then, at one point to the point that I did consider going down to Immortal, but it didn't take much tweaking to continue winning on Deity playing fair. See my last post - at this point I can leave it well past turn 200 before finally pushing for a victory condition, and still get there (I may have to start earlier for culture or domination - but science has been the default victory condition since Civ 1 so I rarely even try for the others, especially as I find them rather tedious in Civ VI). Even then I don't need any kind of planned or efficient tech order - after getting to Rocketry I had to go back to the start of both of the other tech paths I needed in that game.

There's something to be said to allowing flexibility in how you play, but playing almost blindfolded should not win games on the highest difficulty.

Watch a reasont twitch stream where the devs landed a settler on an island with 1 defender... it was surrounded by barbarians... the barbs attacked with ~2... then backed off even though it was an easy kill for them with that many barbs... things like this annoys me... it's super simple.

Barbarians in Civ VI won't take cities either, even though they'll attack them - possibly the same restriction in the coding applies. They physically can't take cities - I've had cities knocked to 0 health by barbarians but the barbarians remain outside.

I still believe that although it's hard, an AI better at winning than 99% of the players is completely doable (there's not 1000 different decisions every turn as someone said), probably by replicating the playstyle of really experienced players (plenty of those on these forums).

I don't know if this is possible as I don't know how AI coding and training works, but it would be ideal if they could set a threshold for each skill level (Prince should beat players of an average skill level perhaps 40% of the time, King 60% etc., rising to 100% for Deity) - this would require identifying the average skill level of players and repeatedly testing against either average players or an AI found to perform comparably well.

Because people want to play an A.I that can use the systems well and not massive cheats that make the game unfun because they are not really playing the game.

The AI isn't a player. It isn't really playing the game and isn't capable of cheating. Giving AI bonuses on different difficulty levels has been the way Civ games have worked since Civ I and the game's far too complex for anything else. What matters is that the AI's base level of competence needs to enable it to use those bonuses to increase its chances of victory - this is the most basic feat the AI needs to be able to accomplish. Not "I have extra Warriors, so I'll throw them all at the human in the first 20 turns and end up losing all these free bonuses I could use later on".

 

[Elhoim]

Like my last "cultural peaceful" deity game did when Nubia blessed me with 5 pitatis and 3 swordsmen by turn 40. Couple reloads later I still didn't find the way to get out of the situation and still have my wonder.

This is the whole point of the OP. Basically, the AI "beat him" so he goes on a rant on why it's unnecessary, that all we just need is "ego pandering" and if you don't get everything you want the game sucks.

Also, the problem there is the implementation of difficulty by Firaxis, by front-loading the AI bonuses, creating an annoying game start, by making your only choice "try to survive", rendering the rest of the systems moot, and then if you do you just won.

 

[iammaxhailme]

I don't think the AI is actually any better in diety, they just have numerical boosts (i.e. cheat) but aren't actually SMARTER. It's the same dumb AI as on prince, just starting with more settlers + building things faster

 

[Victoria]

Google spent $10 billion on AlphaGo, if you give Firaxis that money they will may up with a fantastic AI. Unfortunately not every PC have a fast GPU for neural networks, and therefore only the highest-level computer owners can enjoy that AI.

Yes!, as I have said elesewhere I am sure there are timeouts in the game to stop it 'thinking' too long, or/and the decision trees are dumbed down on purpose both for speed and to alliow a little 6 year old to play on settler without being eaten alibe by Pitati's, then having pitati nightmares for ever after.

Trying to stay off here over Christmas etc. But with all respect to the OP, this is a silly argument.

Agreed, Dota and LOL are quite different games and have huge incomes.

I suspect the people who play on diety are the ones who complain about the AI most.

I suspect you are right, however one needs to be careful about what this means. For example I play on deity accept the AI is what it is for a variety of reasons including money and young players but I still complain. The reason is the statistics are so skewed on deity to try and provide a challenge.

What this all boils down to me is not the polarised view of a billion$ AI but once source code is released and people are willing to spend time without being paid there will very likely be a more challenging game and hopefully deity will not need to be so skewed. This has be proven with Civ V, I think civ VI is much more complex with a lot more choice but hopefully people are up to the task.
Playing as the aztecs and being rolled by Poland early on deity is what I want to see.
I think Firaxis has OP and bad Civs in the game on purpose to allow more peoplke to win on deity and ... wait for it.... to give deity people more of a challenge by playing a crap civ against all OP civs... try it on a small pangea map without restarting, England vs Aztec, Sumeria, Nubia, Rome, Macedon, Korea and Scythia, it is a good challenge and winnable but not easy

 

[steveg700]

This clearly doesn't apply to everyone, but I think it stands true for most.

I'm not a troll and I am in no way trying to stirr things up, but I feel like the truth should be said.

If the game had a decent AI, 90% of the players couldn't enjoy it. I know it, some of you know it, and guys at firaxis know it aswell. If openAI can stomp top human players at dota2, a game requiering way more complex decisions per minute, it won't be that hard to do it here.
Heck, they could only use neural networks for the first 100 turns to get the lead, cutting complexity massively, and then use some pre-programmed easy to replicate science victory scheme.

This will probably lead at AI being better than 99% of the players, without the need for bonuses. But It Would Suck! Like my last "cultural peaceful" deity game did when Nubia blessed me with 5 pitatis and 3 swordsmen by turn 40. Couple reloads later I still didn't find the way to get out of the situation and still have my wonder. Sigh

I find all the complaining ridiculous. Some people wouldn't buy gs because of AI planes, but I don't think they love the game enough to play it at all. Maybe people should enjoy their overly-easy deity games, and move on. Just my two cents.

Okay....

So what?

I'm playing on Immortal or Emperor or Prince or whatever, and I complain that the AI is a poopshow. How is my complaint invalidated? Is there some sloppy baked-in premise that Deity is the default mode that the game should be played in?

Might it be possible that the prodigious bonuses the AI gets on higher difficulty levels would in fact *not* provide a satisfying experience for players seeking a strategic challeng? Or even just a more variable, dynamic, engrossing experience from game to game?

If I'm playing on Deity, is it supposed to suddenly become rewarding that all civ's are trying to build every they have the capacity to build, heedless of their overall victory course, or even whether their cities are falling left and right? Is it going to be rewarding that it doesn't use combined arms, and just marches a wall of knights up to my cities? Or that it marches a wall of catapults that never fire at anything? Or, for some reason, a lone catapult?

As for positing that a player could never win a game against a competent AI---well, for Deity, it should indeed be nigh-impossible to win. If you're routinely beating a game on its utmost difficulty level, that's not so much a testament to skill as a testament to exploitability.

Most people don't expect a highly-sophisticated AI. They just look at other games they're playing by studios that are no more well-endowed than Firaxis (often less so) and see their results and their ongoing efforts.

 

[steveg700]

Complaints about the AI almost always center around its decision making. At launch the AI would almost always have situations in my games where it could take one of my cities but didn't - just like filling a balloon with anticipation and then deciding to let it slowly fizzle out. Just in another thread a poster mentioned how the AI could have taken a city and had a big lead but instead kept offering very favorable peace deals. Even on settler that would be very strange - the most important thing is the AI act in a believable way. I don't want to feel like the AI is pulling its punches or that it doesn't know how to correctly play all aspects of the game consistently.

Well, to some extent the AI will always pull its punches. Anyone who has ever played the tabletop game Diplomacy or perhaps even just Risk knows that if all players start off on equal footing in terms of resources, a lone player will generally not be able to fend off two or more players determined to wipe them out.

The competent AI would be constantly creating extremely squeeze plays where a civ is doomed by simple virtue of a starting location that sandwiches them between two other civ's. When that's a human player, it comes across as "unfun".

 

[Cuneiform]

As has been touched on by others, I can forgive the incompetence of the AI if it just used all the features of the game.

The downgrading of Forts in my opinion hasn't helped the AI, as they've always been better on defence than attack.

But if they could just build an Air Force and USE IT then it might make the late game more of a challenge. If this is the case I wouldn't for example object to the AI getting a free Fighter Unit, when completing a Aerodrome on higher difficulty levels.

 

[AmazonQueen]

As has been touched on by others, I can forgive the incompetence of the AI if it just used all the features of the game.

The downgrading of Forts in my opinion hasn't helped the AI, as they've always been better on defence than attack.

But if they could just build an Air Force and USE IT then it might make the late game more of a challenge. If this is the case I wouldn't for example object to the AI getting a free Fighter Unit, when completing a Aerodrome on higher difficulty levels.

Thats what I'd want from the AI at all levels of difficulty. A competent AI that can use all the features available to players and can prioritize objectives and follow through on them..
I don't think we are ever going to get an AI that performs radically differently at different levels of difficulty. AI cheats will probably always be the main source of increased difficulty

 

[BarbarianHunter]

AIs love conquering city states instead of using their bonuses, and do so right at the start - a waste of resources that is only likely to slow them, especially on higher difficulties where the extra settlers they get make an extra early city less valuable.

The problem with the AI conquering city states is actually worse than that. Not only do they often conquer CS's they themselves have envoys with, but they turn themselves into an envoy and science (ie. Int'l Space Station) generating machine for the human player that liberates them.

 

[JBConquests]

I quit playing civ a few years ago basically because the AI was to stupid. I get irritated/frustrated with AI's that win because they are outrageously cheating. My dream is to see a version of Civ that on Deity beats me (and most everyone else without cheating) badly such that I would have to turn down the skill level in order to compete. I know the difficulties in authoring an AI as I am an experienced software engineer but I also know that the AI could be better than it is. I may never see my dream achieved but it could be better. All they have to do is commit more resources to the AI and it can get better if they chose to make the investment. The company has no interest in making such an investment is the problem.

I just started playing and learning EU4 (Europa Universalis 4) recently and the AI is truly good. The game is cheating on my behalf so I don't lose so badly... It will be fun to see how good the AI is once I learn the game better.

I wish Civ's AI was as good as EU4.

P.S. I won't play civ again until I see some kind of news that they significantly beefed up the AI.

 

[LuckyAC]

That's not an opinion, that's a conjecture.

But my conjecture is exactly the opposite - almost ALL of the people complaining about the AI are playing on deity. People who don't play on Deity are generally not very serious players and are happy with more of a sandbox experience rather than a competitive game, or they are just truly bad players - in either case, bad AI is much less of a problem. Plus if they had found it too easy, they could have moved up to Deity.

Although even for a non-Deity player, the BALANCE of how bad the AI is could be a problem - the AI is so bad at war if it doesn't rush you in the first few turns that one of the main elements of the game is essentially removed.

 

[UWHabs]

This is the whole point of the OP. Basically, the AI "beat him" so he goes on a rant on why it's unnecessary, that all we just need is "ego pandering" and if you don't get everything you want the game sucks.

Also, the problem there is the implementation of difficulty by Firaxis, by front-loading the AI bonuses, creating an annoying game start, by making your only choice "try to survive", rendering the rest of the systems moot, and then if you do you just won.

This is an issue too. I play at immortal, because I just don't want to face an AI that starts with 3 settlers at Deity. I would rather play against an AI that got more general bonuses through the game instead of them up front. Like, I would much rather play an AI that got a free monument/granary/builder for every city, a new settler at the start of every era, and other bonuses throughout the game, if we at least started the game with just one settler each. It's hard enough with the current bonuses, where most AI have their 3rd city down before my 2nd settler is even built, and who can come at me with 6 warriors before I can get a 3rd or 4th military unit out. I can't imagine what it would be like at Deity.