Interesting info from the gamespot preview

[Felis Renidens]

People who spend enough hours to become pro are more likely to get on forums dedicated to the game and agree on things about it. The question is how the AI can be configured to satisfy pro players without annoying more casual players and without giving them lots of bonuses and effectively balancing the game on 'emperor' as the level where AI and player are the same.

And the people who spend that much time on the game are also more likely to multiplay or use mods. From your signature I guess that I don't need to tell you that...

 

[Lucius_]

From gameplay perspective, there's no difference between giving AI "skills" or just increasing its bonuses. "Competent" (which actually means doing things which mimic human behavior to human observer) AI is just an immersion thing, which most (although not all as you correctly pointed out) players enjoy.

I disagree with this. Take happiness in CiV for example. The AI has such a huge happiness bonus that they can simply ignore that feature of the game entirely if they wish. I don't think the AI should have entire game systems not affect them. There is a huge difference between giving the AI the skills to deal with happiness and huge happiness bonuses.

Personally, I'm on the less bonuses, better AI side of this argument. By giving the AI such huge bonuses, at Immortal/Diety level, the player is essentially starting in a hole and has to dig out of, then once ahead around mid game, the game becomes a cake walk. It makes the for a weird difficulty curve over the course of a game, imo.

 

[Acken]

I dont get the point youre trying to pass here.

I already said that the game being casual friendly on lower levels is a given. Nobodys arguing it should be pro level on prince....

 

[Amrunril]

From gameplay perspective, there's no difference between giving AI "skills" or just increasing its bonuses. "Competent" (which actually means doing things which mimic human behavior to human observer) AI is just an immersion thing, which most (although not all as you correctly pointed out) players enjoy.

Personally, I like "competent" AI myself, but the amount of importance some players put for it is funny to me.

I think there is a substantial difference between improving the AI and increasing its bonuses. The more numerical bonuses you add, the less the AI acts as an actual opponent playing the same game. The AI gains the ability to execute strategies that are literally impossible for a human player, and the human player is forced to adapt both to these possibilities and to the expectation that the AI will play badly. Additionally, the player gains the ability to exploit AI bonuses through mechanics such as trade routes and espionage. It isn't feasible to have the AI receive no bonuses, but getting closer to that goal would dramatically improve single player civ as a strategy game.

You should be able to win most of the time when playing at 'prince' (same level as AI) even if you are not a 'pro' as long as you don't make serious mistakes. That's how I see it. Victory at higher levels should also be possible if you know what you are doing.

If the AI is playing like a pro player that won't happen.

Some things like unit shuffling shouldn't happen (they should fortify instead) but you don't need a pro player to notice such things.

If you have an AI that works well on high difficulty levels, its easy to go back and adjust bonuses to make low and mid levels the right difficulty. It's much harder to make a weak AI competitive on high difficulties, as you can see by the way Deity bonuses in Civ V (which some advanced players still consider too easy) warp game pacing and strategy. If the AI ever reached a point where it could consistently beat mid level players with no bonuses, I could see an argument for toning down its "skill" on low-mid difficulty levels, but I don't see this being an issue any time soon. After all, Prince difficulty in Civ V isn't playing with the same bonuses as the AI. The AI plays on Chieftan, and all the bonuses it receives on high difficulty levels are on top of that.

 

[stealth_nsk]

I disagree with this. Take happiness in CiV for example. The AI has such a huge happiness bonus that they can simply ignore that feature of the game entirely if they wish. I don't think the AI should have entire game systems not affect them. There is a huge difference between giving the AI the skills to deal with happiness and huge happiness bonuses.

But AI can't wish

The algorithms are designed so it's expansion isn't too pressing even on highest difficulty levels. The happiness for AI is just set for "too high" to not interfere with it as real cap for AI have no relations to happiness.

 

[Lucius_]

But AI can't wish

The algorithms are designed so it's expansion isn't too pressing even on highest difficulty levels. The happiness for AI is just set for "too high" to not interfere with it as real cap for AI have no relations to happiness.

Ok they can't "wish", but surely you knew what I meant? If not, here you go. India will tend to build very few cities. It CAN build more because it has plenty of happiness to do so, but doesn't because it's programmed not to.

However, Alexander, Washington, Catherine, Bismark, these guys not only can, but often do settle new cities all the way into the atomic era, as long there is still space somewhere in the world where a city will fit.

This is an issue because first the player cannot settle new cities and remain happy the way the AI can. It is a strategy the human player cannot do. Second, after a certain point, it's simply not beneficial to found new cities, thus the AI is wasting precious turns building settlers late in the game causing them to fall further behind because they are obsessed with owning that oil tile on that 3 hex arctic island on the other side of the globe from their own empire. Improved AI would fix both of these issues, which are both somewhat caused by giving the AI bonuses.

 

[stealth_nsk]

Ok they can't "wish", but surely you knew what I meant? If not, here you go. India will tend to build very few cities. It CAN build more because it has plenty of happiness to do so, but doesn't because it's programmed not to.

Yes, that's what I'm talking about.

This is an issue because first the player cannot settle new cities and remain happy the way the AI can. It is a strategy the human player cannot do. Second, after a certain point, it's simply not beneficial to found new cities, thus the AI is wasting precious turns building settlers late in the game causing them to fall further behind because they are obsessed with owning that oil tile on that 3 hex arctic island on the other side of the globe from their own empire. Improved AI would fix both of these issues, which are both somewhat caused by giving the AI bonuses.

That's actually great, great example. AI is programmed to settle at given rate even at stages where it's not considered worth for the player. But that's the case where "incompetent" AI is actually better for the game! It lets player to settle more or less enough cities early in the game and if gaps are left, they are filled by AI later in game, so player could conquer them to puppets if needed.

Of course the reasons why such "incompetent" AI is better is because Civ5 mechanics with puppets, global happiness, tall focus, etc. is far from perfect. But you never could have perfect gameplay, so masking it with AI is quite acceptable approach.

 

[KrikkitTwo]

The general tactical weakness of the AI(due to move and shoot and poor promotion use) means war (particularly naval war, or the use of level4+ units) acts as an exploit...in the sense that the AI is so weak there that it ruins immersion.

Now I could see where some AI improvements could be cost ineffective. But that is true of everything asked for.

The problem of casual players losing on a "no bonus" level (which civ has never had the AI has always had bonuses to make up for its incompetence)...is fine.

If Deity is the first level where the AI has no penalties, and the human has no bonuses (those are 2 different things with different concepts attached) that would be great...average players could play on average levels (change the names of the levels so that average players don't feel like they are losers)

However, I don't think that will ever be a problem (at least until robots take all our jobs and make civ games to keep us placated)

 

[redwings1340]

There's a pretty big difference between a smart AI and an AI with bonuses.

Take chess as an example. A smart AI is one that can make the right moves to beat you. If you play as white, and the AI plays as black, and you two start with the same setup, the chess bot will likely lose to you on lower difficulties, and beat you on higher difficulties, because an AI on high difficulty is really good at chess. Its not getting bonuses, just playing better than you.

If I were trying to design a chess game as an amateur programmer, I would not be able to create a good AI, so if I wanted my chess program to have a good AI, I might have to adopt the civ model. Pretend I'm good enough at AI I can program it to move pieces, but it can't look ahead, so all I tell it to move randomly, but always take an opponents piece when the piece is in range. On higher levels, I might need to make rooks and bishops queens in order to get my AI to be a challenge for humans to beat. You can have better strategy, but the game is unfair, so its up to you to exploit my AI to lure them in to stupid traps to win.

Clearly, the first AI difficulty setting makes for a better game. However, Civilization is a far, far more complicated game than chess (especially for an AI), and being able to get the AI to play the game at all, let alone in a coherent way, is a decent accomplishment in itself. The goal of creating the AI is to make it so that they need as few queens as possible to defeat a human player on the highest difficulty settings, but that's unrealistic with such a complicated game, so we settle for AI bonuses instead of intelligence.

Hopefully Civ VI doesn't need quite as many bonuses for the AI to be competitive, though. That would make it a more fun game, and keep the immersion a bit better.

 

[stealth_nsk]

The general tactical weakness of the AI(due to move and shoot and poor promotion use) means war (particularly naval war, or the use of level4+ units) acts as an exploit...in the sense that the AI is so weak there that it ruins immersion.

I feel what defending city against the AI could be exploit, although it's a problem with too strong cities. Fixes should be there at the first place.

Exploit is some action which greatly increases your effectiveness against logic. AI tactical weakness is not an exploit, is something everyone face and the game is balanced around it.

There's a pretty big difference between a smart AI and an AI with bonuses. Take chess as an example. A smart AI is one that can make the right moves to beat you. If you play as white, and the AI plays as black, and you two start with the same setup, the chess bot will likely lose to you on lower difficulties, and beat you on higher difficulties, because an AI on high difficulty is really good at chess. Its not getting bonuses, just playing better than you.

Chess are bad example as:

1. There are no stats to give bonuses to.

2. The goal of chess program is to teach playing against human opponent. Civilization have no such goal.

 

[Lucius_]

The goal of chess program is to teach playing against human opponent. Civilization have no such goal.

According to whom? O.o

Sure that could be a byproduct of a good Chess AI. There may be specific chess AI's whose purpose is to teach chess; however, the chess app on my phone, I'm pretty confident it's goal is to be at least a somewhat fun and challenging chess game in hopes I'll buy the paid version or accidentally click an ad.

 

[unpossible251]

Better AI would be preferred over a computer player that just gets a lot of buffs. I think it goes without saying. I doubt the programmers ever find intelligent tactics amongst the AI and say "hmmm. Better program that out"

I'd like to see intelligent advisers, if I could. The advisers for CivRev could actually give you some real good insight into what you should do. The CiV advisers? Not so much. Mister Science could state how many beakers you were pullinin per turn, but couldnt put it in any context. Am I accelerating? Are rivals gaining on me?

I'd forgive Civ6 for a lot of tactical blunders from buff-upgraded enemies if my AI advisers could give me some genuinely good advice about how to run a war or an economy.

 

[Felis Renidens]

Only fool runs an empire without amenities

 

[stealth_nsk]

Better AI would be preferred over a computer player that just gets a lot of buffs. I think it goes without saying. I doubt the programmers ever find intelligent tactics amongst the AI and say "hmmm. Better program that out"

Sure, but it's a question of priorities. The amount of work to implement "move and shoot" feature is much bigger than implementing large game system like religion and balancing it out. With limited developer resources they have to make a choice and the choice looked obvious.

 

[KrikkitTwo]

Better AI would be preferred over a computer player that just gets a lot of buffs. I think it goes without saying. I doubt the programmers ever find intelligent tactics amongst the AI and say "hmmm. Better program that out"

I'd like to see intelligent advisers, if I could. The advisers for CivRev could actually give you some real good insight into what you should do. The CiV advisers? Not so much. Mister Science could state how many beakers you were pullinin per turn, but couldnt put it in any context. Am I accelerating? Are rivals gaining on me?

I'd forgive Civ6 for a lot of tactical blunders from buff-upgraded enemies if my AI advisers could give me some genuinely good advice about how to run a war or an economy.

If your AI advisers could give good advice, then the AI players would have good advice on how to play as well.

 

[IronfighterXXX]

Personally, I like "competent" AI myself, but the amount of importance some players put for it is funny to me.

Yeah, it really 'funny' that people want competent AI in a strategy game.

The amount of work to implement "move and shoot" feature is much bigger than implementing large game system like religion and balancing it out. With limited developer resources they have to make a choice and the choice looked obvious.

What are you talking about, do you realise that even this simple mod made it?: http://forums.civfanatics.com/showthread.php?t=521955

Firaxis hasn't implemented this after few years just because of lack of listening to feedback and laziness.

 

[darko82]

Firaxis hasn't implemented this after few years just because of lack of listening to feedback and laziness.

Are they so shameless?

 

[stealth_nsk]

What are you talking about, do you realise that even this simple mod made it?: http://forums.civfanatics.com/showthread.php?t=521955

Firaxis hasn't implemented this after few years just because of lack of listening to feedback and laziness.

4MB dll file without any media inside, so you really think it's "simple" mod? Do you know, how many hours was spent here total (including testing)?

 

[True_Candyman]

4MB dll file without any media inside, so you really think it's "simple" mod? Do you know, how many hours was spent here total (including testing)?

It's replaces an existing file, so I would have thought the vast majority of that is unchanged copy pasta.

 

[IronfighterXXX]

4MB dll file without any media inside, so you really think it's "simple" mod? Do you know, how many hours was spent here total (including testing)?

Well, I don't know, maybe indeed Ninakoru is a computer genius and Firaxis don't have competent programmers to do it,

but in this case they just should kindly ask him for permission to use his work and I suppose he wouldn't refuse