What exactly are the advantages of each difficulty?
- Thread starter Shifted
- Start date
Acken
Deity
We don't know exactly because it's not clearly referenced as in civ5. What I know is that Emperor+ has a free settler. And the AI get some bonuses to yields.
I have found evidence of AI starting units here :
<Row Era="ERA_ANCIENT" Unit="UNIT_SETTLER"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_WARRIOR" NotStartTile="true"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_SETTLER" AiOnly="true" MinDifficulty="DIFFICULTY_EMPEROR" DifficultyDelta="0.5"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_WARRIOR" AiOnly="true" MinDifficulty="DIFFICULTY_KING" DifficultyDelta="1" NotStartTile="true"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_BUILDER" AiOnly="true" MinDifficulty="DIFFICULTY_KING" DifficultyDelta="0.5" OnDistrictCreated="true"/>
This means king+ start with warrior + worker and emperor+ with settler but I'm not sure what the difficultydelta means or ondistrictcreated.
There are also some reference to that kind of stuff
<Row>
<ModifierId>HIGH_DIFFICULTY_SCIENCE_SCALING</ModifierId>
<Name>Amount</Name>
<Type>LinearScaleFromDefaultHandicap</Type>
<Value>0</Value>
<Extra>8</Extra>
</Row>
which I don't know what exactly 8 what it reffers to
I have found evidence of AI starting units here :
<Row Era="ERA_ANCIENT" Unit="UNIT_SETTLER"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_WARRIOR" NotStartTile="true"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_SETTLER" AiOnly="true" MinDifficulty="DIFFICULTY_EMPEROR" DifficultyDelta="0.5"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_WARRIOR" AiOnly="true" MinDifficulty="DIFFICULTY_KING" DifficultyDelta="1" NotStartTile="true"/>
<Row Era="ERA_ANCIENT" Unit="UNIT_BUILDER" AiOnly="true" MinDifficulty="DIFFICULTY_KING" DifficultyDelta="0.5" OnDistrictCreated="true"/>
This means king+ start with warrior + worker and emperor+ with settler but I'm not sure what the difficultydelta means or ondistrictcreated.
There are also some reference to that kind of stuff
<Row>
<ModifierId>HIGH_DIFFICULTY_SCIENCE_SCALING</ModifierId>
<Name>Amount</Name>
<Type>LinearScaleFromDefaultHandicap</Type>
<Value>0</Value>
<Extra>8</Extra>
</Row>
which I don't know what exactly 8 what it reffers to
aguliondew
Prince
Yea I found it surprising when I realized that the AI was starting with an extra during my emperor games. That normally happen on diety in civ 5 
. Still a lot of fun.
. Still a lot of fun.
player1 fanatic
Fanatic
Extra setter at Emperor is big red flag for the AI performance, IMHO.
If devs though that 6/8 difficulty requires extra settler for AI to be competitive, compared to olders games where this started at 8/8 difficulty, then AI must be performing much worse then in previous games.
If devs though that 6/8 difficulty requires extra settler for AI to be competitive, compared to olders games where this started at 8/8 difficulty, then AI must be performing much worse then in previous games.
Acken
Deity
It's not very clear what their other bonuses are. The problem I have with the AI right now is that the system it uses is a lot less clear than for civ5.
That said it doesn't perform very well right now but I find it difficult to say if its a balance problem making it hard for it to upgrade/tech or declare wars (warmonger penalty) or just poor coding.
The thing is, the AI has just too much of an unfair early advantage right now. The extra settler, the extra units the extra eurekas... It can come at you with 10units in 20 turns with their strength bonus and oligarchy bonus. And then once you have fought that it becomes rather easy.
I'm sad to say that this is the opposite of what a challenging AI is supposed to be. It should be more fair to the player as in allowing him to do things and at the same time always be a force to be reckoned with.
That said it doesn't perform very well right now but I find it difficult to say if its a balance problem making it hard for it to upgrade/tech or declare wars (warmonger penalty) or just poor coding.
The thing is, the AI has just too much of an unfair early advantage right now. The extra settler, the extra units the extra eurekas... It can come at you with 10units in 20 turns with their strength bonus and oligarchy bonus. And then once you have fought that it becomes rather easy.
I'm sad to say that this is the opposite of what a challenging AI is supposed to be. It should be more fair to the player as in allowing him to do things and at the same time always be a force to be reckoned with.
Sic
Chieftain
- Joined
- Oct 10, 2016
- Messages
- 81
Didn't Firaxis say in one of the latest videos that bonuses work differently in 6? I.e. that they wanted to give a larger bonus early game instead of having bonuses for yields (for instance) that was there throughout the game.
Acken
Deity
If thats the case its a terrible idea.
Why? The early game sets the scene for the rest of the game. The AI needs to be extra good at the early game.
Mr.WorldWide
Smugly Inferior
Because it's resulted in the first 100 turns or so determining the end of the game. AI goes strong at first and if you beat the armies they throw at you the rest of the game the ai can't hold a candle to you. The difficulty should be consistent throughout the game imo
Acken
Deity
Not really. The ai should be a challenge for the whole game and still allow you to have a normal experience in the first 50turns.
Right now its a crappy boxer with a one shot gun. Totally unfair but if you dodge the bullet youll put it to sleep no problem.
The comparison between civ5 and civ6 is easy. Even more so with mods and civ6 does high difficulties completely backward.
But this is a problem with the mid- and late-game AI. That doesn't mean that we need to nerf the early-game AI.
player1 fanatic
Fanatic
Even in older games, AI was more front-loaded and bigger threat early.
This is why Prince wonders strategies never worked on higher difficulties.
Ideally, AI should be as much "back-loaded" as possible. To give more challenge as time passes not other way around.
Only time this could happen was with runaway civs, that consume their neighbours, and that just happens from time to time.
This is why Prince wonders strategies never worked on higher difficulties.
Ideally, AI should be as much "back-loaded" as possible. To give more challenge as time passes not other way around.
Only time this could happen was with runaway civs, that consume their neighbours, and that just happens from time to time.
The problem with back-loading the AI advantages is that the early game is most crucial. A strong early game leads to a snowballing effect, whereas a weak early game can be very difficult to recover from. I'd obviously prefer a strong AI all the way through, but if I had to choose were to emphasize the cheaty advantages, I'd always pick early game.
Acken
Deity
Oh the bonuses can start early alright. All im saying is that the current system is awful. Getting rushed on t20 by 10 warriors is stupid. And then its followed by a boring mid late game.
player1 fanatic
Fanatic
Actually, I think it would be better for game flow, if AI had raising bonuses over eras, to simulate late game advantage.
Early advantage is just that. An early advantage. After you catch up, all challenge is gone, and game becomes too easy.
Tomice
Passionate Smart-Ass
That's exactly my opinion. Why don't they learn from the successful total conversion mods?
Free settlers just aren't fun.
- Joined
- Jan 12, 2004
- Messages
- 1,373
The AI doesnt know what to do with its 10 warriors at the moment
Acken
Deity
Thankfully otherwise youd just restart. And thats exactly the problem with front load bonuses.
Amrunril
Emperor
- Joined
- Feb 7, 2015
- Messages
- 1,239
I strongly agree that giving the AI most of its bonuses in the form of a head-start is a mistake. This method does succeed in making the game challenging, but it also makes competition with the AI a highly binary affair. You spend most of the game far behind the AI, struggling to survive and catch up. Then, you reach a tipping point where you pull ahead and never look back. This brief tipping point is really the only time where you're competing on a semi-even footing with the AI, and it's, not coincidentally, the most satisfying point in the game. I often found myself choosing a difficulty level based not on how much of a challenge I wanted but on what stage of the game I wanted to be the most exciting.
What's more, strong starting bonuses dramatically warp a number of game dynamics, even beyond these competitive effects. The early eras of the game are proportionally shorter than they should be. Early game focused human civs can't pull ahead in the same fashion they could in an even start, but neither can mid/late game focused AIs catch up with them later on. Early wonders and religion (and now, presumably, great people) require massive commitment just to be part of the competition. However much or however little challenge starting bonuses present, they make it impossible for the game to proceed as it would between players with default starting positions.
Shifting most of the AI's bonuses to yield bonuses over time would restore the game's intended dynamics, making the game play out much more as a competition between equals. Numerical bonuses would, given proper balance and level choice, allow the AI to progress through the game at roughly the same rate as a human player, albeit through raw numbers instead of skillful play. The AI would be less overbearing in the early game, but it would also be a more persistent threat, with greater ability to recover from early setbacks. Instead of a single tipping point, the game would proceed as an extended back and forth, based on civ strengths, tech paths and periods of strong or weak play. AI bonuses over time would, in my view, make for a better game than massive starting bonuses in every way imaginable, and I'm highly disappointed by Firaxis' decision to double down on this latter path.
What's more, strong starting bonuses dramatically warp a number of game dynamics, even beyond these competitive effects. The early eras of the game are proportionally shorter than they should be. Early game focused human civs can't pull ahead in the same fashion they could in an even start, but neither can mid/late game focused AIs catch up with them later on. Early wonders and religion (and now, presumably, great people) require massive commitment just to be part of the competition. However much or however little challenge starting bonuses present, they make it impossible for the game to proceed as it would between players with default starting positions.
Shifting most of the AI's bonuses to yield bonuses over time would restore the game's intended dynamics, making the game play out much more as a competition between equals. Numerical bonuses would, given proper balance and level choice, allow the AI to progress through the game at roughly the same rate as a human player, albeit through raw numbers instead of skillful play. The AI would be less overbearing in the early game, but it would also be a more persistent threat, with greater ability to recover from early setbacks. Instead of a single tipping point, the game would proceed as an extended back and forth, based on civ strengths, tech paths and periods of strong or weak play. AI bonuses over time would, in my view, make for a better game than massive starting bonuses in every way imaginable, and I'm highly disappointed by Firaxis' decision to double down on this latter path.
