From Easy to Deity. A new way to code the Ai difficulty?

[stealth_nsk]

It's totally possible to design a game where AI will beat human players completely and this could even be fun. See chess.

However, Civ is not chess. Civ makes its fun by playing on strongest sides of our brain. It requires pattern recognition which no traditional AI algorithms could handle, calculations beyond neural network capabilities and interpreting maps the way no modern algorithm could.

But more importantly, nobody needs really strong AI for civ. AI need to look competent and provide challenge while at the same time role play. So, cheating AI is totally fine as long as player doesn't see it. And that's the biggest problem with Civ5 and Civ6 - their AI is actually better, but 1UPT makes AI actions (at least tactical ones) much more visible, and thus make AI look incompetent. In this context, I believe, using commanders will be able to solve the AI problem without making the game less complex.

 

[Lord Lakely]

Well, if the higher AI difficulties merely involve the AI getting a bunch of free stuff and little else, I'm automatically out of playing higher difficulties. I'd rather play a sandbox game than being scammed out by cheating computer players.

In an ideal world, the AI weights certain values differently based on difficulty, with values being more random and less synergistic on lower difficulties, and more well-rounded and/or specialized on the higher ones.

A militarist on lower difficulties would field an above average army, but would also build a lot of random things it doesn't need, like peaceful wonders and culture buildings.
A militarist on higher difficulties would attempt to field the largest possible army, combined with good economy AND science so that their army gets paid and is on curve with everyone else's.

(this is also why agenda's do not work as a substitute for clear strategies and personalities - sure Harald wants to build "the largest navy", but does that navy help him win? What if the largest navy is just three ships, while everyone else has one or two? I've yet to see a Civ6 game where Norway came close to winning on the basis of boats)

I remember that in order to make the AI feel more challening in Civ6, the devs drastically increased the flavour for building Campuses for every leader. That was a step in the right direction because Science unlocks Wonders and units, and that made the AI less of a pushover.

The problem with that solution is that it wasn't a solution at all - The Campus is not the strongest distrinct in Civ6 - that would be the Commercial Hub and not by a small margin. Food and Production stacking via internal Trade Routes and large Gold reserves from adjacency supercharge any economy - You have more people AND more money to buy things with. It's also why i think the Harbour and Holy Site are the second and third best districts in Civ6 - anything that gives you growth, production and currency for the cost of ONE TILE, is always worth prioritizing. Campuses are good but they don't win by themselves.

Quod Erat Demonstrandum: Multiplayer Civ6 games are centered around building CommHubs and setting up Trade Routes asap and only then focus on either Campuses or Theatre Squares depending on your Civ and leader's strengths.

The essence behind this example is that in order to give the AI bonuses or modifiers on higher difficulties is by understanding what a human player would value. Human players like us know that in order to fund a large army or a bunch of expensive research labs, we need to take care of our RCI first - Food to work more tiles, production to get everything up faster and gold to finance it. In-game currencies like Gold and Faith in Civ6 can also be used to rush out your necessities even faster.

An AI that gets to neglect their RCI because they have a bajillion of built-in advantages doing it for them is going to fall behind eventually, leading to a boring endgame.

 

[Krikkit1]

Well, if the higher AI difficulties merely involve the AI getting a bunch of free stuff and little else, I'm automatically out of playing higher difficulties. I'd rather play a sandbox game than being scammed out by cheating computer players.

In an ideal world, the AI weights certain values differently based on difficulty, with values being more random and less synergistic on lower difficulties, and more well-rounded and/or specialized on the higher ones.

A militarist on lower difficulties would field an above average army, but would also build a lot of random things it doesn't need, like peaceful wonders and culture buildings.
A militarist on higher difficulties would attempt to field the largest possible army, combined with good economy AND science so that their army gets paid and is on curve with everyone else's.

(this is also why agenda's do not work as a substitute for clear strategies and personalities - sure Harald wants to build "the largest navy", but does that navy help him win? What if the largest navy is just three ships, while everyone else has one or two? I've yet to see a Civ6 game where Norway came close to winning on the basis of boats)

I remember that in order to make the AI feel more challening in Civ6, the devs drastically increased the flavour for building Campuses for every leader. That was a step in the right direction because Science unlocks Wonders and units, and that the AI less of a pushover.

The problem with that solution is that it wasn't a solution at all - The Campus is not the strongest distrinct in Civ6 - that would be the Commercial Hub and not by a small margin. Food and Production stacking via internal Trade Routes and large Gold reserves from adjacency supercharge any economy - You have more people AND more money to buy things with. It's also why i think the Harbour and Holy Site are the second and third best districts in Civ6 - anything that gives you growth, production and currency for the cost of ONE TILE, is always worth prioritizing. Campuses are good but they don't win by themselves.

Quod Erat Demonstrandum: Multiplayer Civ6 games are centered around building CommHubs and setting up Trade Routes asap and only then focus on either Campuses or Theatre Squares depending on your Civ and leader's strengths.

The essence behind this example is that in order to give the AI bonuses or modifiers on higher difficulties is by understanding what a human player would value. Human players like us know that in order to fund a large army or a bunch of expensive research labs, we need to take care of our RCI first - Food to work more tiles, production to get everything up faster and gold to finance it. In-game currencies like Gold and Faith in Civ6 can also be used to rush out your necessities even faster.

An AI that gets to neglect their RCI because they have a bajillion of built-in advantages doing it for them is going to fall behind eventually, leading to a boring endgame.

That is one advantage of the 3 act…a Player (Human or AI) can’t fall behind too much since they automatically “catch up” in both tech and civics every age

 

[Core Imposter]

That is one advantage of the 3 act…a Player (Human or AI) can’t fall behind too much since they automatically “catch up” in both tech and civics every age

Doesn't that meet the definition of a kludge?

 

[Core Imposter]

I think it's intensely tin foil to think firaxis makes bad computer opponents so you buy the next game.

Well, I never said that but your comment, "People play games until they get tired and a new game gives them a fresh experience.", I am tired of "fresh" new experiences that are just new collections of names and graphics. Maybe there will be some gameplay in Civ 7. I really haven't heard anything about gameplay really.

 

[gdr_willter]

Well, I never said that but your comment, "People play games until they get tired and a new game gives them a fresh experience.", I am tired of "fresh" new experiences that are just new collections of names and graphics. Maybe there will be some gameplay in Civ 7. I really haven't heard anything about gameplay really.

Well I'll not blame you but I heard kinda enough gameplay things.

 

[GoneFishingX]

I am tired of "fresh" new experiences that are just new collections of names and graphics. Maybe there will be some gameplay in Civ 7.

I mean civ 7 is fresh beyond names and graphics. They totally revamped the core gameplay to the point where tons of people are unhappy. They make pretty solid changes with each iteration and people get annoyed. Civ 3, 4, 5, and 6 are very different games, but this time they might have taken a step too far in their desire to try new things.

Well, I never said that

You did say that, I quoted you in my response. Here it is again.

It wasn't good design, either, for American manufacturers to engineer and build refrigerators and freezers, many of which are still in service 40 years later. The planned obsolescence model in the game industry today is all about not building the necessary artificial intelligence that would allow for a game to repayable and become a staple of a generation's entertainment regimen.

I can't think of anyway to interpret that statement other than you saying they make the computer opponents bad so that way you buy the next game. That's a deeply strange and conspiritorial opinion to me.

 

[Core Imposter]

I am old enough to remember when concerns about planned obsolescence in manufacturing industries was claimed to be a tin foil hat conspiracy theory so if you want to put that label on me, I really do not mind at all. I understand why game companies run their businesses the way they do.

I have one opinion of why game companies do not have good AI in their games, you have your several. Let's leave it at that.

 

[Krikkit1]

Doesn't that meet the definition of a kludge?

No simply a rubber banding to prevent science (or culture) from being too overwhelming of an advantage (land still seems to be the snowball of choice)

 

[Nikolai II]

The game Phantom Brigade (a turn based mech combat game) let's you customize the difficulty by tweaking specific aspects of the game. See below:

I've played the game and I really like that level of customization in the difficulty setting because it let's you tweak the aspect that you need help with or that you want to make more challenging while leaving other aspects of the game the same.

Civ7 could do something similar. You could have sliders for unit strength, unit cost, science output, culture output, influence output, gold output, upgrade cost from towns to cities, build cost for buildings and wonders, aggression of independent peoples, mastery techs on or off etc... This would let the player choose what aspects of the game to make easier or harder. I think this would make difficulty levels a lot more interesting. Maybe a newbie player struggles keeping up with the AI in science, they could play a normal game but give themselves a science boost. Maybe another player struggles with combat so they play a normal game but just make unit strength higher to make combat easier for them.

Civ as a franchise is not very into detailed customization like this on game setup, not to this degree. I doubt we'll see it in vanilla.

 

[Lazy sweeper]

When has this . . . ever happened before? What evidence do we have between low-end AI difficulty has any impact on volume of sales (which, historically, as a Chieftan enjoyer, has been pretty trivial)?

This seems pretty core to your OP and I simply don't get it. Everyone was young once. Everyone hated losing games once. Some folks still do, nomatter how old they get.

Nintendo had Civ Revolution, super easy. Nintendo core philosophy revolves around making easy games (now, back to Zelda 1 it was the opposite, but there was scarcity back then, and it was not artificial) Nintendo ditched the DS for the Switch, because of mobile gaming. Smartphones were more powerful then their DS and young gamers all had a smartphones, and less and less felt the need to buy a DS. So the Switch was born. They are the only portable console makers on the market and the reason for that is that they kept their eyesight well fixed onto young demography, which by definition is the audience which has more spare time to play games.
I don't see anything irrational in stressing this.

Civ 1, 2 and 3, were mostly seen as computer games, not console games. Yes there were Civ ports for the Snes but they never gained much traction. Without a mouse it was painful. Civ Revolution was the first attempt to make a Civ game for the console market. And it worked. The simplification was replicated in Civ VI.
Civ VI was basically a Civ Revolution 2.
Civ VII is too compicated for a Switch. Nintendo fans want an easier, more arcade gaming.
Firaxis should make two versions of Civ. One for the Consoles, and One for the Mac/Pc. As it used to be.
We Mac/Pc users will get the same hybridization of the consoles for this reason. It makes sense from a business perspective.
But I guess consoles sales would quadruple if they had an exclusive version, easier to learn all the quirks.
I gifted my nephew a Game Gear with Sonic, and she loves it more than anything she ever had.
Those simpler consoles has something the newer generation are missing completely...

And yes we of the older gen are not much different in the end... I still enjoy playing black and white tetris on my gameboy, or D&d...
but we don't buy new games a lot... my last console was the N64...

 

[Gorbles]

Nintendo had Civ Revolution, super easy. Nintendo core philosophy revolves around making easy games (now, back to Zelda 1 it was the opposite, but there was scarcity back then, and it was not artificial) Nintendo ditched the DS for the Switch, because of mobile gaming. Smartphones were more powerful then their DS and young gamers all had a smartphones, and less and less felt the need to buy a DS. So the Switch was born. They are the only portable console makers on the market and the reason for that is that they kept their eyesight well fixed onto young demography, which by definition is the audience which has more spare time to play games.
I don't see anything irrational in stressing this.

Civ 1, 2 and 3, were mostly seen as computer games, not console games. Yes there were Civ ports for the Snes but they never gained much traction. Without a mouse it was painful. Civ Revolution was the first attempt to make a Civ game for the console market. And it worked. The simplification was replicated in Civ VI.
Civ VI was basically a Civ Revolution 2.
Civ VII is too compicated for a Switch. Nintendo fans want an easier, more arcade gaming.
Firaxis should make two versions of Civ. One for the Consoles, and One for the Mac/Pc. As it used to be.
We Mac/Pc users will get the same hybridization of the consoles for this reason. It makes sense from a business perspective.
But I guess consoles sales would quadruple if they had an exclusive version, easier to learn all the quirks.
I gifted my nephew a Game Gear with Sonic, and she loves it more than anything she ever had.
Those simpler consoles has something the newer generation are missing completely...

Didn't VI make the turn processing a lot, lot faster than in V? And you're saying they did this not by improving how the game worked between iterations, but by simplifying it?

I'm just not connecting the dots I guess. Speaking as someone who played Pokemon Red as a kid, who has a Switch now, but has also been PC gaming since around the same time as I played Red as a kid.

Also, doesn't VII have a Switch version confirmed?

 

[Lazy sweeper]

Didn't VI make the turn processing a lot, lot faster than in V? And you're saying they did this not by improving how the game worked between iterations, but by simplifying it?

I'm just not connecting the dots I guess. Speaking as someone who played Pokemon Red as a kid, who has a Switch now, but has also been PC gaming since around the same time as I played Red as a kid.

Also, doesn't VII have a Switch version confirmed?

Yes it does, it had also Vi I guess??? But I don't know the numbers of sales... if it met the expectations or what...

I play Civ Rev on the Ds and a playthrough can be less than one hour long...
I have no idea what kind of experience could be Civ VI or VII on a Switch...
I play DS on the fly, on a trip, on a train, I can't imagine playing a full fledged Civ on a portable console...
I mean I would not be interested I think, I'd choose a faster experience game.

Something like this. A poor Ai is provided...

CivWonders by Cryptomnesic Softworks

cryptomnesic-softworks.itch.io

 

[Eyswein]

I think the biggest issue historically has been no AI scaling by age. In Civ 6, lots of bonuses to start, but once you catch up the AI is easy. I'd like to see Civ 7 scale by age (and throughout each age) so it is always tough (or easy on lower levels). I don't mind AI cheating, just make it consistent. I'm also hoping Deity equivalent is too hard for me to beat.

 

[Core Imposter]

I think the biggest issue historically has been no AI scaling by age. In Civ 6, lots of bonuses to start, but once you catch up the AI is easy. I'd like to see Civ 7 scale by age (and throughout each age) so it is always tough (or easy on lower levels). I don't mind AI cheating, just make it consistent. I'm also hoping Deity equivalent is too hard for me to beat.

With different eras the AI can be provided with new bonuses at each to provide a more challenging experience through the entire game.

 

[Acken]

At the very least you want your AI to appear competent. It should look as if it is playing the game.

When it comes to providing a challenge, better balance of the different options typically also improve AI player strength. Scripted AIs never learn the meta, it is using an a priori set of rules and decision weights made by a developer that typically never gets updated. In the meantime your players slowly learn to understand the good and the bad options because Civ has always been rather poorly balanced with some options and strategies being systematically much better than others.
Improving game balance so that choices are closer with each other is half the battle in providing an AI challenge before having to rely too much on cheats and I always have felt sad how little work is typically done there by Firaxis.

It also improves the human user experience by allowing more interesting choices for their own civ so it's a double win.

 

[Kev1916]

Considering how hard they try to peddle dlcs and the effort to reduce size and meaningful decision's , its unlikely that they want to drive away supercasuals by making the game a challenge to play.

 

[Taefin]

If the AI just starts its game-winning project in the first turn it is available, and invests as much in it as it has readily available, this would go a long way to surpassing VI and Millenia, where I found I could be trailing so far with no risk of actually losing.

Removing strategics from unit upgrades, and auto upgrading on each era will hopefully make for a more competitive late game AI, and if ideology driven conquest is a legacy path, and the AI is being hard coded to address it, hopefully also much less passive then. I suspect the real danger will be an AI conquering its neighbor and getting thus obtaining an economy to threaten a few different victory types, vs AI militarily threatening the player, but would love to be surprised.

Aside from Old World, Civ7 might have unique the advantage of avoiding mechanisms like armies from humankind or Millenium that completely removed the threat of the AI, who created weaker armies than the player by using subpar units, or where the battle map allowed simple tactics to result in zero player losses.

 

[Lazy sweeper]

If the AI just starts its game-winning project in the first turn it is available, and invests as much in it as it has readily available, this would go a long way to surpassing VI and Millenia, where I found I could be trailing so far with no risk of actually losing.

Removing strategics from unit upgrades, and auto upgrading on each era will hopefully make for a more competitive late game AI, and if ideology driven conquest is a legacy path, and the AI is being hard coded to address it, hopefully also much less passive then. I suspect the real danger will be an AI conquering its neighbor and getting thus obtaining an economy to threaten a few different victory types, vs AI militarily threatening the player, but would love to be surprised.

Aside from Old World, Civ7 might have unique the advantage of avoiding mechanisms like armies from humankind or Millenium that completely removed the threat of the AI, who created weaker armies than the player by using subpar units, or where the battle map allowed simple tactics to result in zero player losses.

Optional patches with clear indications of their scope and purposes could help identifying dangers, although I had a hard time interpreting your concerns, some of your's could be interpreted as advantages?

 

[Lazy sweeper]

I think the biggest issue historically has been no AI scaling by age. In Civ 6, lots of bonuses to start, but once you catch up the AI is easy. I'd like to see Civ 7 scale by age (and throughout each age) so it is always tough (or easy on lower levels). I don't mind AI cheating, just make it consistent. I'm also hoping Deity equivalent is too hard for me to beat.

There could be an option, or more, that could be enabled by default on higher difficulties that makes Ai better.
Or players worst... for example, just make fog of war 1 tile only to within borders or any units. Or even better, clean vision only one tile ahead if any other tile is not a plain, desert, tundra or grassland.
Making all other tiles potential ambush...
Make successful ambush get a +10 combat surprise factor... if a unit is injured add also +5 panic... and so on... all of this and more could and SHOULD be optional... if not by menu and settings, via optional patches.