September Update Thread

[Vizurok]

Btw, random question, did the prophet assemblingTyphoon also prophesy features of the alleged third expansion or only civs?

As far as I can recall, he envisioned a third full expansion pack with 8 civs. It seems that it's impossible to recover his words but he even suggested a quite logical civ list.

 

[Disgustipated]

I don't think they will drop info on a new expac just yet

I would expect some youtube live video to run for 24 hours taunting us like last time.

Did you play wow classic?

errm yes.

 

[WillowBrook]

Btw, random question, did the prophet assemblingTyphoon also prophesy features of the alleged third expansion or only civs?

Pretty sure it was only civs for xp 3, they had some features of what turned out to be gathering storm, but only possible civs for xp3, and seems like that list was still quite tentative when they got their info.

 

[Alexander's Hetaroi]

As far as I can recall, he envisioned a third full expansion pack with 8 civs. It seems that it's impossible to recover his words but he even suggested a quite logical civ list.

Even so 2 of the Civ slots weren't even mentioned, so they could have been anything.

 

[oSiyeza]

That’s your view, I suspect it is cultural. Humour in England can gain you a lot of respect but appreciate it is tricky to use globally and in written form. Regardless, at most he was being flippant. Not my fight so step out.

WRT the AI, it is not as such, it is a bunch of decision paths that already has shown time and again it is too simplistic.
If a civ has denounced you they will not deal so nicely with you and their ability to accept embassies or friend requests depends on how much they like you so you appear to be making a generalisation.
Yes things are not perfect but they have tried.

I totally agree. They have tried. But I'm not sure to agree with the reason they failed.

Friend request and some straigh forward factors deppend on if they like you or not. Probably cause you can perceive directly the agency of your actions, as a player this way.

The AI reacting in the exact same way if an friend attacks and enemy or an enemy attacks a friend, would be a behaviour very easy to change. It would Increase a lot the complexity of the game, and the approach to diplomacy. Why this has not been done? Precisely because of that.

AI behaviours may be simplistic, but they were introduced by someone with a consistent idea and design philosophy in mind. There are, flaws, and obvious mistakes. But I also see a clear motive in the way it was put together.

The AI is mostly deffensive. It is designed to deffend well from military aggression but also to rarely attack or be opportunistic. This is intentional, so the player is not attacked when in a weak position, or distracted in a different war.

The AI expands science, economy and religion in a steady non overwhelming fashion. Taking its time, being distracted. I think this is intentional, so the player can freely play without much pressure.

It very rarely uses big invasion forces for aggresion or after starting a war, it stops when taking one or two cities, that usually loses inmediately after because it didnt conquer enough territory to hold loyality. I think this is done intentionally, so the player can recover from a hit.

It does not declare surprise wars, or take entire civilizations, it does not betray friends, I dont think it rejects promises, or refuses to aid in emergencies. This is also intentional, so the player would not be frustrated due to the AI being unpredictable or stealing victories from the player.

As an AI researcher myself, to me this is not only intentional. It is consistent with the way war wearines, grievances, casus belli, and the warmonguer mechanics are implemented. The player can ignore emergencies, or decide to become a warmonger, but the AI cannot. The AI behabes like a good boy the player can have fun with.

The entire game is balanced that way. From disasters to dark ages to emergencies. Every system is designed to provide the player a canvas that is mostly decorative. And does not interfere too much, and can decide to ignore. No negative effects, no big drawbacks.

Only the first stages of the game can be frustrating. This is alsi intentional. A player that has invensted many hours in the game should win, to avoid frustration. The entire game is designed to make the game easier as advances. The same happened with the X-Com saga, and is a deliverate choice.

 

[Eagle Pursuit]

Thinking about the mask made me remember an event on Twitter from back in July. Character modeler Matthew Kean tweeted that they were looking for a freelance animator specializing in real-time hair. Shortly afterwards the tweet was deleted.

At the time, I was thinking maybe they were just going to freshen up the leaders. But maybe they are going to put in a leader with really splendid hair or beard.

 

[Victoria]

The AI reacting in the exact same way if an friend attacks and enemy or an enemy attacks a friend, would be a behaviour very easy to change. It would Increase a lot the complexity of the game, and the approach to diplomacy. Why this has not been done? Precisely because of that.

It has been done. If you take a city. The size of the city does affect the warmonger/grievance penalty which some people are unaware of but also it affects your allies less and also affects those at war with the same opponent less.

The AI expands science, economy and religion in a steady non overwhelming fashion. Taking its time, being distracted. I think this is intentional,

if you look over each game you will see the AI growth in these areas to be more linear while yours get exponential at a certain stage if doing things in the most efficient way. The AI does not play in the most efficient way because it would put off new players and changing the rules for increased difficulty is a nightmare for developers. You research AI but in that field you can do so without having worked on real development projects where these lessons become learnt. Are you also a developer or more a researcher?

It does not declare surprise wars, or take entire civilizations, it does not betray friends,

It does! .... it does many things you do not see. Many people used to complain about joint wars, that they were a poorly implemented anti human mechanism but getting log of these incidents from people I could show that the AI was squabbling amongst themselves and a joint war often consisted of one AI civ trying to bribe another AI civ on multiple occasions before it agreed to joint war you and even then it made peace 10 turns later quite often. Just because you do not see it, it is happening.
I was messing around with GOTM 70 in the weekend, I went to the second continent in T140 to find 3 civs there, but later I went in T100 to find 5 civs there, they did not mysteriously give up, they were wiped out.
Yes the AI does a lot of things oddly but it is not as easy to sort out these decision trees as you may think. Some of Firaxis decision will be made on selling the game and making it winnable and the majority of people like to get that winning feeling. The issue is fanatics that know how to play soon realise optimal styles of play make winning easy. But fanatics are not the larger fan base. A lot of under 10’s play the game as well, and that’s great but they do have to be catered for in some way.

Your points have some validity but are not to the extremes you mention and it takes time going through diplomacy logs to find out they squabble and hate each other and some of the reasons they often do not run away is because they are not efficient.

 

[leandrombraz]

The AI is mostly deffensive. It is designed to deffend well from military aggression but also to rarely attack or be opportunistic. This is intentional, so the player is not attacked when in a weak position, or distracted in a different war.

I don't think the AI is defensive and they're definitely opportunistic. Try to leave a city undefended where the AI can see. No walls, no Garrisons, no units nearby, with your military strength relatively low. The AI LOVES an undefended city, to the point they will try (and fail) to cross your whole territory just to get to an undefended city, if you have one on the other side, like a moth to the flame. You can see that behavior on the AI tendency to declare war on CS that just got liberated and have no walls or units. Even better, play early game like a innocent pacifist, preferably in high difficulties, just build almost no army, no walls, make no friendships with your neighbors. Sit there believing that the AI won't attack you when you're in a weak position, see what happens. Early game in high difficulties is notorious harder than late game because the AI is aggressive when they have the upper hand, it will exploit your weakness and try to take you down. What make the AI look defensive is that as the game progress it have a hard time doing anything offensive, mostly against the almighty player. The AI isn't defensive, it just fails to be offensive and often have no other option but to sit there, almost static.

That was the joke, yes. Im not sure what was the motivation of the response, though. As it was referred to me and not to what i said. That was not granted at all. Humor, even dark humor is not contended with respect.

Anyway, Regarding the actitude of the AI... It is actually coherent with itself in most cases. It is just very naïve towards, emergencies, grievances, and war. It lacks self interest, and when representing very charismatic and militaristic leaders feels weird.

This may be a design decission. But the fact that they are always altruistic when judging offenses comited to other civs, or overcommit to helping any civ in need. Regardless the deals they may have or how much they like or dislike the civs in question, is what breaks the ilusion for me.

When the AI overcommit to help they aren't being altruistic, they're being competitive. When Gathering Storm was released, the AI wouldn't waste as much gold in emergencies, it was easy to win those without investing too much since there was no real competition from the AI. Firaxis increased their competitiveness on that, so the AI is trying to get that diplo point, not being Altruistic.

 

[bbbt]

And I guess I can assume this is an insult.

Insult? Not intended. It's a joke about the use of the term 'SJW' which has strong 'internet subcultural' associations in it's usage. I.e. I don't think I've ever seen someone use it in these forums despite extended argument threads about why certain leaders were included.

At the time, I was thinking maybe they were just going to freshen up the leaders. But maybe they are going to put in a leader with really splendid hair or beard.

Hmm, this sort of splendid hair or beard perhaps: https://www.google.com/search?q=hammurabi&source=lnms&tbm=isch

I assume we are talking expansion, there's been no indication of any sort of DLC pack in their updates?

 

[Haig]

Gathering Storm was announced on November, but when did the steam activity began?

 

[djvandyke]

I wish these patches weren’t announced until they are nearly ready.
I inevitably become reluctant to start a new game.

 

[Eagle Pursuit]

Gathering Storm was announced on November, but when did the steam activity began?

Mid-September.

If that PDreview build is "Product Development Review", and I saw a few devs mention meetings on Twitter, it could be that they are finalizing details before 2KQA begins its work.

 

[steveg700]

It does! .... it does many things you do not see. Many people used to complain about joint wars, that they were a poorly implemented anti human mechanism but getting log of these incidents from people I could show that the AI was squabbling amongst themselves and a joint war often consisted of one AI civ trying to bribe another AI civ on multiple occasions before it agreed to joint war you and even then it made peace 10 turns later quite often. Just because you do not see it, it is happening.

I don't think the AI is defensive and they're definitely opportunistic.

I'll third these sentiments. I think the AI is very territorial, and I've certainly seen them use surprise wars and also wipe another civ out. Then again, I always play on large maps with lots of civ's.

The weaknesses, derived from personal experience, seem to be:

• Open Borders and Delegations seem to have more weight in the war/peace decision than their small diplo modifier would suggest. I've inadvertently scum-saved myself out of wars a few times now by offering an open borders agreement a couple turns before the AI would have. These two moves fast-track friendship a little too quickly. I would almost rather a civ have a can on friendship declarations that increases through progress in the civics tree until eventually unlimited, and then the same thing happens after alliances hit the table.
• They accept losses to loyalty with way too much aplomb. You can pick an empire to pieces in a way a human would see as a clear and present threat.
• The diplomacy bonus for an alliance and embassy seem unlikely to be countered by any amount of grievances, agenda violations, or other negative behavior. Once you're in, you're in. Of course, for the duration of an alliance, negative modifiers don't have any effect anyway, given that alliances are unbreakable. We need some kind of out, such as grievance accumulation or broken promise, whereby we can exit an alliance prematurely. And then the AI needs to actually do it when their ally acts against their interests.

 

[The Kingmaker]

Mid-September.

If that PDreview build is "Product Development Review", and I saw a few devs mention meetings on Twitter, it could be that they are finalizing details before 2KQA begins its work.

I like the way you think.

 

[_hero_]

I wish these patches weren’t announced until they are nearly ready.
I inevitably become reluctant to start a new game.

It's a no win proposition for developers. If they just announce something and release it at the same time, people complain that they wish they had known in advance so they could have scheduled some free time to try it out. If they announce early but with an exact date and something comes up in the last QA check that forces a delay, people are furious. Now they announce a little early without a specific date and yet people again complain.

This is why we can't have nice things.

 

[iammaxhailme]

It's a no win proposition for developers. If they just announce something and release it at the same time, people complain that they wish they had known in advance so they could have scheduled some free time to try it out. If they announce early but with an exact date and something comes up in the last QA check that forces a delay, people are furious. Now they announce a little early without a specific date and yet people again complain.

This is why we can't have nice things.

It seems to be they usually announce things 2-3 weeks before they are ready, which is fine, but I just wish they would formally put a more specific ETA on these things. "Soon" has sometimes meant a few days in Civ 6... but sometimes over a month

 

[Zaarin]

Thinking about the mask made me remember an event on Twitter from back in July. Character modeler Matthew Kean tweeted that they were looking for a freelance animator specializing in real-time hair. Shortly afterwards the tweet was deleted.

At the time, I was thinking maybe they were just going to freshen up the leaders. But maybe they are going to put in a leader with really splendid hair or beard.

"Rapunzel also leads Germany in Sid Meier's Civilization Vi..."

 

[oSiyeza]

It has been done. If you take a city. The size of the city does affect the warmonger/grievance penalty which some people are unaware of but also it affects your allies less and also affects those at war with the same opponent less.

It may well be that taking cities have small differences in war, or alliance conditions. Does it take into account, liked, disliked, denounced, trading and other diplomacy status? I really think it does not. If it has, the effect surely is not noticeable, even in those extremes.

It is fairly weird, that when taking city states, or sending AID, these effects certainly do not working at all, if they exist.

The AI does not play in the most efficient way because it would put off new players and changing the rules for increased difficulty is a nightmare for developers. You research AI but in that field you can do so without having worked on real development projects where these lessons become learnt. Are you also a developer or more a researcher?

My point was in some way, showing that a lot of the complains people have. Are not so much due to incompetence, but a conscious choice. For better or worse.

I'm not suggesting, all these are negative factors. In particularly, I like the economic, science, culture and the religion approach enemy civs have. I think is fine tuned, the way it is. I was pointing that it was certainly a part of the design approach. And in some aspects, it is not the best decision.

I like a lot of the thing the game does. Others not so much.

It does! .... it does many things you do not see. Many people used to complain about joint wars, that they were a poorly implemented anti human mechanism but getting log of these incidents from people I could show that the AI was squabbling amongst themselves and a joint war often consisted of one AI civ trying to bribe another AI civ on multiple occasions before it agreed to joint war you and even then it made peace 10 turns later quite often. Just because you do not see it, it is happening.
y

I may have not fully analyzed all what happened in every game I played. However, I can point that I never found myself in the receiving end of a Surprise war. Or backstabed from a civ I had no grievances against. I can say that I never seen the AI doing this. In this end, I'm am unsure If this is good or a bad thing. It is probably good mostly, since most of the time an unpredictable attack would be quite frustrating. However, in many cases I find myself being the only civ, with military agency in the status of the game. And I think different civs should have different behavior in this regard, as a part of their hidden agenda.

An also never in my games an AI wiped any civilization. I know it is possible, but I would guess very specific conditions need to happen, with a civ falling way behind, in early ages, and when there are no grievances against other major civs. I also have never seen a domination Victory from the AI go past conquering a single capital. Most of the time, no civ gets to conquer any capital. Though I guess is probably possible. As far as I know, the AI designer of the game, (though this is based in an old video) has never seen an AI domination victory in hundreds of AI battle royal monitored plays. I also never have seen an AI having significant grievances from warmongering.

Granted, I play mostly on a couple of maps and sizes. And only completed or almost completed 20 games of so. That means, roughly I have faced 150 AI civs. But is still disappointing.

Also it is truth that while the AI uses the strength of the player to value if they are an easy target and will in some cases take unprotected cities. The overall behavior of the AI I would say still quite innocent in this regard. It factors grievances more than it should, and expansion opportunities less than it should. I don't think they take into account the military strength in the specific area. Another component I do dislike, is the abrupt difference from early to middle ages to later ones. Modern ages were ages flooded with military conflicts in reality, where war technology suffered the biggest explosion in history, something that the military units in civ reflect. But is something that the AI rarely puts to use. due its reluctance to be at war in late ages.

 

[pokiehl]

The AI constantly surprise wars me and takes out other civs. I play on King.

 

[Cerilis]

I noticed that every (work)day since the announcement, the civ6 twitter highlighted one part of the balance updates from the announcement video. There only is governments and Great Admirals left now. Patch/livestream on thursday?