Whatever they do I hope the AI is substantially upgraded

[Leathaface]

I think making the AI better at claiming victory when it is close to it's chosen victory type should be given priority. Playing on Immortal I sometimes see the AI well ahead in Science and it builds 1 or 2 Spaceship parts but then it just stalls completely. The AI should win a Science victory if it is significantly ahead in Science.

I've also never seen the AI win a Culture victory, i'd like to see that change in GS. If the AI has a bucket load of culture (think Pericles) the only challenge that brings ATM is if you're going for a culture victory, delaying your culture victory enough for another Civ to win a Science victory.

 

[Civrinn]

The devs will probably never improve the AI unfortunately (modders might). However i always find this strange concept thrown around in these forums that "people only want the AI to be upgraded because Deity players need MOAR challenge!", no not at all. The AIs silly behavior is immersion breaking as well. At whatever difficulty I need to feel like the game is trying its best (at its difficulty level), yes feel. Clearly it won't be trying its best because its a lower difficulty. Basically things like "AI can't use aircraft, AI peacing to easy, AI not actually taking cities even though they are defenseless..." those things. An AI upgrade isn't strictly about giving people more of a challenge, it is also about creating a more satisfying game. Winning by itself isn't satisfying, for example I could play chess with a 3 yr old. I would get no satisfaction for beating that 3 year old.

 

[Aristos]

The devs will probably never improve the AI unfortunately (modders might). However i always find this strange concept thrown around in these forums that "people only want the AI to be upgraded because Deity players need MOAR challenge!", no not at all. The AIs silly behavior is immersion breaking as well. At whatever difficulty I need to feel like the game is trying its best (at its difficulty level), yes feel. Clearly it won't be trying its best because its a lower difficulty. Basically things like "AI can't use aircraft, AI peacing to easy, AI not actually taking cities even though they are defenseless..." those things. An AI upgrade isn't strictly about giving people more of a challenge, it is also about creating a more satisfying game. Winning by itself isn't satisfying, for example I could play chess with a 3 yr old. I would get no satisfaction for beating that 3 year old.

EXACTLY. You are hitting the rusted nail in the head, multiple times.

 

[RohirrimElf]

I would have been happy with an air overhaul for the next expansion.

Don’t really understand why the AI declares war on you (often a combined war of 2 AI’s). And then don’t bother to attack you. There are still wars declared where you could just wait and then declare peace without a single shot being fired.

The last war i fought against Scythia was a good one. Scythia’s units hit me like a wave. And each turn i took down 1 or 2 of their units. Untill it stalled and i undertook a counterattack. Taking one of their cities. But at least it was a good attempt. To many wars are just phony wars. Or like a bombard coming towards your city all alone. Bumping against an outdated melee unit on its way. You focus it down with a ranged unit in 2 turns. And that is it. (I play on immortal and deity)

 

[MarigoldRan]

For AI to be good at war late-game, we need stacks of doom back.

 

[Leathaface]

I find the AI is more dangerous when they attack with Cavalry, since Cavalry has a lot of movements points. If you have a city close to the AI and it's obviously the first city the AI wants to take, they could send Knights your way and capture the city pretty quickly.

I've also had Sumeria capture my city closest to him with Knights, then head West straight to my capital instead of taking a city to the south of the first city. The AI is capable of making wise decisions, but there if there are 2 or 3 cities the AI can attack the AI can split the troops up and not aim their offense at a specific city.

 

[Trav'ling Canuck]

For AI to be good at war late-game, we need stacks of doom back.

The AI in Civ 5 used air and AA units well enough to be dangerous at late game war. It was actually weaker at war earlier in the game, before it's heavy production bonuses had the chance to really grow into a superior force and before it's relative competence at using air planes kicked in.

 

[Alexadamz]

My last game at Emperor I took Meroe from Nubia with me War Carts. Nubia has her Pitati Archers ready but didnt attack back. They could range attack the city and never done it. At least would be more of threat, as their other city was close and after depleting my city points they could strike fast with a horseman or a chariot. That is just stupid. They never done anything and after recovering my 3 War Carts HP I just killed them all and took their last city.

For real, I can't understand what is the dev's idea dor the AI and the game difficulties. It angers me and is less immersive the gameplay that just give starting advantages to the AI at higher difficulties, so they can start ahead and be more challenging. For me, the real challenge would be the AI playing better, after starting from the same place as we do. Maybe they could get more yields from each population or bonus from their palace, so they could have things faster than we do. Having an extra settler and more troops from start is a real shady idea, that free the devs from really making AI smarter at play.

 

[isau]

It would be easier to accept Civ 6s AI if we didn't have the Civ 5 Vox Populi AI to compare it to. VP smoked me a couple of times. I still remember the first surprise war declaration from the Vikings and the subsequent fear factor.

I will grant tho that at least in Civ 6 the x combat is slightly better than 5. Archer cheese in unmodded 5 really hurt that game.

 

[darko82]

It would be easier to accept Civ 6s AI if we didn't have the Civ 5 Vox Populi AI to compare it to. VP smoked me a couple of times. I still remember the first surprise war declaration from the Vikings and the subsequent fear factor.

I will grant tho that at least in Civ 6 the x combat is slightly better than 5. Archer cheese in unmodded 5 really hurt that game.

If we did not have Vox Populi, I/we would compare it to Civ IV, which was more challenging on high difficulty (especially with the AI mod). The AI was still pretty poor but at least it posed threat and had the ability to destroy you or win the game. Those bonuses they got were stupidly insane, however.

If people here accept the poor AI and buy the game, the developer won't do anything about it. They've sold games with poor AI for over 25 years, so they think it's normal and acceptable.

Some people even defend the AI, justify it, or they say they don't mind or care. Maybe those who work for Firaxis? The worst thing is that Civ VII will have the same problem because of people who accept it.

 

[Gorbles]

If we did not have Vox Populi, I/we would compare it to Civ IV, which was more challenging on high difficulty (especially with the AI mod). The AI was still pretty poor but at least it posed threat and had the ability to destroy you or win the game. Those bonuses they got were stupidly insane, however.

If people here accept the poor AI and buy the game, the developer won't do anything about it. They've sold games with poor AI for over 25 years, so they think it's normal and acceptable.

Some people even defend the AI, justify it, or they say they don't mind or care. Maybe those who work for Firaxis? The worst thing is that Civ VII will have the same problem because of people who accept it.

A very elaborate conspiracy, considering the very simple and obvious counterargument: AI is hard. You don't need to invent Firaxian defenders of the AI that are for some reason allowed to post publicly, defending said AI. I'm pretty sure we've even been over the modder arguments in this thread already.

Nobody says you have to accept anything, nor that you can't criticise it. But if your only response to reasonable counterarguments is "the game is bad because people defend it" with a side of "perhaps it's Firaxis making these comments", then you don't look very reasonable in response.

 

[darko82]

AI is hard.

Hard or not, Vox Populi shows it can be done. It can be improved. But who cares if the game is selling well...

Nobody says you have to accept anything, nor that you can't criticize it. But if your only response to reasonable counterarguments is "the game is bad because people defend it" with a side of "perhaps it's Firaxis making these comments", then you don't look very reasonable in response.

I never said what you are trying to imply. Not "the game" but "the AI", which should be number one priority for them to deal with. Unfortunately, it's not. Besides, I just wanted to emphasize my disbelief by such response. Let's end here.

 

[pino2000]

A few people want better AI. Most people want some bling bling cosmetics via a new civ (happens to be the one where they live). From a sales perspective this is kinda given.

The perfect amusement ought to accomodate diverse kinds of players - calling for Prince to be legitimately hard invalidates the point of having changed trouble levels. As a reasonable, moderately aged person, the amusements that baffle me most are the ones troublesome even at lower trouble settings (taking a gander at you, God of War). I play Civ on King and that is the ideal trouble for me - not totally unimportant but rather not disappointing either. For what reason would you need to deny me of my good times?

 

[Gorbles]

Hard or not, Vox Populi shows it can be done. It can be improved. But who cares if the game is selling well...

I never said what you are trying to imply. Not "the game" but "the AI", which should be number one priority for them to deal with. Unfortunately, it's not. Besides, I just wanted to emphasize my disbelief by such response. Let's end here.

Of course it can be improved. But making a game means you have X time to dedicate to Y features. It's easy to say "AI should be the number one priority", but you have no idea what would be sacrificed for such a thing.

 

[RohirrimElf]

I dont hope all the time and energy (of the single lonely AI programmer) goes towards implementing the new features for the next expansion. And not adressing the current AI, which is still far behind of what would be acceptible for an AI for a civ game.

With all the bonusses the AI could keep up and even win on the 2 highest difficulties on a maco level. With balancing i think good improvements will be made for the expansion. I’m looking forward to new content. But facing the AI in combat is so one sided. When you could face the AI in battle with half of their forces. And often not lose any or just a single unit during war.

This is my biggest issue with the game do. Played civ 6 for over 600 hours. Loving building up my empire and lissten to soundtrack of the leaders and narrator. i would be above 2000 hours if civ 6 provided me with a challenge.

 

[Gorbles]

If we assume that there is a dedicated AI programmer, and all their time is already being spent on this game, then you would need additional resources to make the AI better in the same timespan for release. Those resources would come from elsewhere.

It's all a matter of perceived importance and priorities. And Firaxis, or rather, 2K, will have a very good idea of how far down the stack AI is, regardless of how important it is to the community of above-average, well, Fanatics

 

[Kaan Boztepe]

they should first finish adding new mechanics , civs , leaders , units etc. before they start worrying about the poor AI performance. Considering that there is a good chance that there is a 3rd expansion , working on improving AI before everything is in place , doesnt make sense. Although once the game is finished in terms of addons etc i also would like improvements on AI, just not a priority right now.

 

[Civrinn]

For AI to be good at war late-game, we need stacks of doom back.

This is absolutely false. *points to Civ 5 Vox Populi*

they should first finish adding new mechanics , civs , leaders , units etc. before they start worrying about the poor AI performance. Considering that there is a good chance that there is a 3rd expansion , working on improving AI before everything is in place , doesnt make sense. Although once the game is finished in terms of addons etc i also would like improvements on AI, just not a priority right now.

What is the purpose of building more mechanics in a game that the AI can't actually play in the first place? It should be priority right now, considering a decent AI should have came with vanilla.

EDIT: At a bare minimum they should allow modders to go in and fix what they cant if they simply would rather sell us new civs/mechanics.

 

[Lord Lakely]

I definitely hope the AI is fixed on high difficulty levels. I'm currently trying out Emperor (after having played King for the past 2 years) and I hate the advantages the AI gets to increase the difficulty curve of the game. Like, it's more challenging, which I appreciate, but the increased difficulty comes from a whopping 40% advantage in production and a free Settler, not because the AI plays more like a human player would. The AIs actually don't play differently, at all.

I hate playing Civ 6 at the higher difficultiesbecause of this. I don't want to lose the game just because the AI gets some built-in advantages that I, as a human player, will never be able to achieve myself, it's so *frustratingly* unfair and also, lazy programming.

Other games (such as Age of Empires 2 HD) have different AI codes for every difficulty level, which scales their competency for each difficulty. In AoE2 specifically, the AI starts with precisely the same resources as the player, but on Hardest plays similar to how a pro human player would (this level of programming is achieved by pain-stakingly typing in the generic AI code in by hand, but good lord the results are so worth it.). I wish Firaxis had programmed their AI like that, because the current lack thereof makes any difficulty above King completely UNAPPEALING to me.

 

[Gorbles]

What is the purpose of building more mechanics in a game that the AI can't actually play in the first place? It should be priority right now, considering a decent AI should have came with vanilla.

EDIT: At a bare minimum they should allow modders to go in and fix what they cant if they simply would rather sell us new civs/mechanics.

As I mentioned earlier, priorities. The AI is not challenging if you know how to exploit it. A lot of people on here know how to do that, but I really have to stress this but none of us are your average Civ 6 player. Most of us aren't even your average Civ franchise player. And I mean that in a good way!

As for the modding aspect, a lot of people make noise about the DLL, but people don't seem to vocalise (in threads in this subforum at least) exactly what is lacking from the existing modding potential of Civ VI, which is going to be a bit different from what was possible in CiV. Maybe getting that extended is easier than Firaxis (or more likely, 2K), giving out even more source code. I don't know, but it's worth a shot. I've modded other games apart from Civilisation, and I've always tried to get the official tools improved (and when there's no other option, write stuff myself).

Other games (such as Age of Empires 2 HD) have different AI codes for every difficulty level, which scales their competency for each difficulty. In AoE2 specifically, the AI starts with precisely the same resources as the player, but on Hardest plays similar to how a pro human player would (this level of programming is achieved by pain-stakingly typing in the generic AI code in by hand, but good lord the results are so worth it.). I wish Firaxis had programmed their AI like that, because the current lack thereof makes any difficulty above King completely UNAPPEALING to me.

Typing these kinds of strategies by hand is no longer a feasible affair. AoE 2 is an old game. Civilisation 5 and 6 are both far more complicated than it. You'd be asking for a lot of handwritten code, in an age when games are barely allowed the development time required to get them to the market.