Djospe

Chieftain
Joined
Aug 26, 2012
Messages
8
Dear Civ players,

This is my first post ever on civfanatics, although I have been playing the series extensively since Civ 3 and although I'm a relativity good player (I beat deity in most games I play). I had to post this idea because I think that it could vastly improve the Civ gaming experience. I would like to see if you guys agree with me and if we could maybe sent a petition to Firaxis.

Civ 6 has tons of different systems, mechanics, scenarios, civilizations, customization options, and more. Standard game settings feature enough room for each player to create a lot of cities, features a lot of civilizations for trading, religious conversion, military conquest, etc. That's all good and nice - in theory - but in reality, I think that this is the root cause of many problems we are all aware of:
  1. The Artificial Intelligence is super weak, often making extremely poor choices in all aspects of the game - city placement, units management, trading... If less content would be created, more time could be dedicated to the AI, and there would be less mechanics and situations for which the AI needs to be programmed. This would surely lead to a drastically improved AI, which is the main hurdle for serious players.
  2. An average game takes too much time, and the feeling is especially slow in mid and late game. Why is the feeling especially slow is mid and late game? Well, that's when all of the systems and mechanics kick in, that's when we've build/conquered all of the cities that we could, and that's when we're mostly just waiting for the game to end, but every turn takes a lot of time because we have tons of things to manage. With less content (less game mechanics, less civilizations per map, less tiles per civilization, less turns per game), the mid and late game would be much faster. Sure, we can change the game speed, map size, sea levels, etc. when customizing the game, but this creates imbalances and further confuses the already confused AI.
  3. There is no truly meaningful, convenient, ranked multiplayer. If the game would take less time and if there would be less civilizations per map, then fewer players would leave the game and the turns wouldn't have to be fully simultaneous. With less content, maybe a meaningful, ranked multiplayer games could be introduced in-game?
Some examples of the game systems that could surely be either removed, automated, or reduced (less units/slots/events but more impact per unit/slot/event) are: gossip, art buildings theming, espionage, religion, world congress, trading routes, trading with other civs, citizen/tile management, natural disasters, etc. Also, there could be less citizens per city, less districts per city, less buildings per district, less technologies, less policies, less wonders - less everything, but more impact of everything, reducing micromanagement, increasing fun.

I think that this would really create a smoother and more interesting mid and late game, while also allowing for a much better AI, a much better multiplayer, and an overall better experience of the game. Do you guys agree with me?
 
If less content would be created, more time could be dedicated to the AI

I think its more like if less content would be created less developers were hired.
The AI might still profit from it since less mechanics mean less mistakes. But less mechanics does not mean a better game in general. I would rather have some mechanics (diplomacy, world congress, warfare, maybe even victory conditions) reworked than removed. In fact I would even like more stuff to keep the late game interesting.
If you want to play without (or nearly without) leavers you can always look for a nice MP community.
 
Sounds like a boring game that would be reduced almost exclusively to fighting other civs. Hard pass.
 
I'm on board with fewer and better game mechanics. But throwing out citizen management and trade routes would basically be throwing out core elements of the Civ game that have been around since 1.

Also, IMO Civ 6 has way fewer buildings than any civ game in the past. If anything, Civ 6 feels pretty stripped down in this regard (although, of course, districts have kind of replaced buildings for the most part).
 
Sounds like a boring game that would be reduced almost exclusively to fighting other civs. Hard pass.
Yeah. I'm more or less an Imperialist, but I like to build up my infrastructure before I conquer my enemies.
 
If less content would be created, more time could be dedicated to the AI, and there would be less mechanics and situations for which the AI needs to be programmed. This would surely lead to a drastically improved AI, which is the main hurdle for serious players.
Unfortunately it doesn't quite work this way. If it was a matter of putting in some man hours, firaxis would likely do it.

In order to have good AI, you need
  1. Systems conducive to non human players. This usually means systems where you can define objectives and solutions that can be efficiently computed.
  2. A development team that understands how to play the game at a sufficiently high level
  3. People who can link those two things together into functioning AI code.
If you don't satisfy those you will never create a "challenging AI."
Look at something like districts. Under the current rules, there is almost no chance you will ever create an AI that can optimally plan cities because its computationally too expensive. So then you need heuristics/"shortcuts" to achieve a "good enough" solution. Because firaxis is also making the rules, they can tweak the system a little to make it a lot easier to achieve "good enough."
The other problem is that players play according to the meta; the AI was programmed without that knowledge. Okay, you say, they can adjust it. What difficultly level do you think the devs play on? The highest I've seen is most honorable Carl the QA lead playing emperor, and the other devs have all referred to him as very skilled compared to them. So who is going to make an Ai that plays the game extremely well, if the developers are only moderately skilled at their own game?

I am in no way knocking them here. It takes a certain mentality to play civ on deity, few players are at that level, and its much more important that a dev is good at game design than playing deity. And going back to point 1, is it more important that the mechanics are fun for players, or conducive to the AI? That's a value judgement that is hard to draw the line on.

An example of adapting the game to the Ai off the top of my head is a current effort by the devs of Stellaris - the AI sucks at managing certain parts of its economy, so they are making some changes that reduce micro, but also make building up colonies more deterministic. That translates to "easy to program an AI build order for." I don't know a lot about civ AI, but I suspect there are a few systems with that sort of problem making it hard to program well even for someone who knows what they are doing.
 
Unfortunately it doesn't quite work this way. If it was a matter of putting in some man hours, firaxis would likely do it.

In order to have good AI, you need
  1. Systems conducive to non human players. This usually means systems where you can define objectives and solutions that can be efficiently computed.
  2. A development team that understands how to play the game at a sufficiently high level
  3. People who can link those two things together into functioning AI code.
If you don't satisfy those you will never create a "challenging AI."
Look at something like districts. Under the current rules, there is almost no chance you will ever create an AI that can optimally plan cities because its computationally too expensive. So then you need heuristics/"shortcuts" to achieve a "good enough" solution. Because firaxis is also making the rules, they can tweak the system a little to make it a lot easier to achieve "good enough."
The other problem is that players play according to the meta; the AI was programmed without that knowledge. Okay, you say, they can adjust it. What difficultly level do you think the devs play on? The highest I've seen is most honorable Carl the QA lead playing emperor, and the other devs have all referred to him as very skilled compared to them. So who is going to make an Ai that plays the game extremely well, if the developers are only moderately skilled at their own game?

I am in no way knocking them here. It takes a certain mentality to play civ on deity, few players are at that level, and its much more important that a dev is good at game design than playing deity. And going back to point 1, is it more important that the mechanics are fun for players, or conducive to the AI? That's a value judgement that is hard to draw the line on.

An example of adapting the game to the Ai off the top of my head is a current effort by the devs of Stellaris - the AI sucks at managing certain parts of its economy, so they are making some changes that reduce micro, but also make building up colonies more deterministic. That translates to "easy to program an AI build order for." I don't know a lot about civ AI, but I suspect there are a few systems with that sort of problem making it hard to program well even for someone who knows what they are doing.

Just to offer a counterpoint, I'm okay with an AI that is sort of drunkenly stumbling around, because I don't play Civ for the same reasons I played a hypercompetitive game like DOTA. But sometimes that drunken AI seems to fall into a ditch and can't get itself out. For example, I have very rarely seen an AI mount a successful offensive campaign against another AI that has gotten to walls. The AI in that case seems to be asked to utilize a highly complex system (1UPT combat) in a somewhat optimal way.
 
You'd have to play Civilization 5 + Vox populi to experience an AI that is acctually very capable of taking over other civs:
Spoiler the Zulus currently taking over continent nr 2 :
upload_2021-1-28_16-10-24.png


And they almost got me look at that map:
Spoiler Help the Zulus are coming :
upload_2021-1-28_16-12-40.png
 
  • Like
Reactions: PiR
Top Bottom