Join me, Anthony Salter (aka Viridian) as I talk about what it's like being a new employee at Stardock and detail what I'm doing to help Elemental be the best strategy game ever.
What we're shooting for and why.
Published on February 18, 2010 By AnthonySalter In Elemental Dev Journals

Hi!  I'm Viridian, also known as Anthony Salter, and I just finished a major revamp of the citybuilding system.  I'd like to talk today about our design goals for citybuilding and how the system currently works.

First, let's look at how our major inspirations, Civilization and Master of Magic, handled cities.  In both games, cities were fairly abstract.  They consisted of a single icon on the map and a screen full of sprites and numbers.



And you would end up with dozens of the things as the game progressed; to the point where you'd probably end up ignoring some because they were too small or unproductive to help you win the game.  (Didn't prevent them from falling into civil disorder and bringing your whole empire to a grinding halt...grr...)

 

As strategy games developed, a genre of incredibly detailed citybuilding games emerged, including the Caesar series, the Settlers series, and the Anno series.

 

 

In these games, everything is simulated practically down to the atomic level.  These are the kinds of games where you need to mine ore to make tools to cut down trees to gather lumber to take to the sawmill to make planks to build new buildings.

Now, there's no doubt this can be fun.  I've enjoyed both the Settlers and the Anno series of games myself.  The only problem is that citybuilding, while important, isn't the only thing you do in Elemental, and thus we can't allow it to dominate the game the way it does in Anno-style games.  (I can hear certain people weeping on the forums already but it's true.)  So what we've tried to do is create a happy medium.

I've spent all this time telling you how citybuilding won't work; it might be a good idea to tell you how it will.

What exactly did we want when we set out to create our citybuilding system?

Well, first, we didn't want city spam.  Thus, we created a system where building a smaller number of larger, older cities is rewarded.

As you probably know, Sovereigns can create cities, thus creating a town hub.  There are five levels a hub can go through - they start as outposts, then upgrade to hamlets, villages, towns, and cities.  At each upgrade point you'll get eight new tiles to build improvements on - and your city will be able to support more efficient improvements that it couldn't before.

Another feature of cities is that they are (mostly) auto-upgrading.  If you expand your city to a village and you have the Housing technology researched, then all your huts will upgrade to houses - instantly, and for free.  Your city needs to be at the proper level and you must have the technology researched in order for this to happen.  Again, I can hear the cries of some forum-goers who think that this will negatively impact the game, but we're facing facts here.  Ninety percent of the time when we get a new housing tech we simply demolish our old houses and build new ones right where the old ones were.  Because of the hard forty-tile limit you can't just throw more out there - non-optimal improvements will literally be a detriment to your city.

Indeed, crafting a good city is going to be a continual series of trade-offs rather than a forever-growing list of improvements.  And as the city grows and the game progresses, you will find yourself continually repurposing your cities rather than building new ones.

An early city in Elemental.

 

Our goal is to strike a balance, so that we aren't overwhelming the player with city management, but we still provide a robust enough experience that you don't just think of your city as numbers and sprites.  When someone attacks your city and your little people start running around screaming, we want you thinking, "Hey!  Stop picking on them!  How 'bout a little FIRE, Scarecrow?!"

 

EDIT:  I originally stated that Sovereigns needed to expend essence to create cities.  This is incorrect; they expend essence to bring the land back to life so the city can be built.  I have fixed the error in the article.


Comments (Page 9)
11 PagesFirst 7 8 9 10 11 
on Feb 22, 2010

I like the general tone of your posts and the general gist of what you are saying. I think you are headed in the right direction.

If I may, I would like to offer constructive criticism and questions on 2 things you mention:


we created a system where building a smaller number of larger, older cities is rewarded.

Good. Smaller numbers of cities means more personality to each. Yet how will the rewards be done exactly?

 

non-optimal improvements will literally be a detriment to your city.

This seems very dangerous if taken literally.

IF you mean that a suboptimally organized city (characterized primarily? by suboptimal improvements within the city) will be detrimental in the sense that your city will fall behind because it will be unable to produce what it needs to produce in the best way (because, for example, you cannot be producing any wombats if your tiles are all full of clunker-producers), then that sounds reasonable, but basically only means that suboptimally organized cities will be suboptimal.

But IF you mean that suboptimal city improvements will drain your resources without providing you with benefits, then I think this is the wrong track to take for sure, because it would introduce no-brainers into the game. More no-brainers = less real polyvalent strategic choices, and less choices = less fun (in a strategy game).

 

 

on Feb 22, 2010

Elemental would need a system where the AI has limitations like the player does, without choking the AI so that it can never ever keep up with the player. Meeting the AI and knowing you are dead right from the start is hardly the way to go imo.

I'm a bit confused by this whole post. The Deity difficulty in Civ IV cheats, big time. If it didn't, people like you probably wouldn't find it very difficult at all. In order to create a difficulty setting that would challenge players like you, the Civ IV team had to resort to giving the AI all sorts of advantages. Of course the AI on a cheating difficulty setting won't have the same limitations as the player. But if you bump your difficulty down a few notches to where the AI loses its cheating advantages, then it does become subject to the same limitations as the player.

So it sounds to me like you're suggesting one of the following: either that the toughest AI setting should be as good as the best human players without cheating, or that the AI should never cheat on any difficulty setting... The former is not remotely practical, and the latter would be silly. There is no good reason why Stardock shouldn't include some cheating difficulty levels above their highest non-cheating one; it gives sadistic people the challenge they want and it doesn't affect anyone else (and can't be very hard to implement).

on Feb 22, 2010

@ Onamostikon

The OP clearly stated that available upgrades would be automatically applied. This is to optimize/simplify city development. Essentially its automatically turning all your Longswords into Longswords +1, and later turn them into Longswords +2. Or rather, in the case of cities, its House turns to House +1, turns to House +2.

This is primarily to lend attention to other, less tedius parts of the game. Assume that every time you gain the tech for House +X effieciency, you have to build House +X over your House N, where N = house X-1. Sure, in specific city builders there are different sizes with all House values of X, different pros and cons. In these cases, such as Anno, or (from personal experience) Black and White, all buildings are valid choices due to resources available, ect.

When a building takes the same amount of space, as well as require roughly the same resources, its set to an Auto-Upgrade system. You still have some choice. You can choose where to place houses, and wether its a Commoner's house or a Wealthy house. Commoner's Houses hold large population capacity, while Wealthy houses provide prestige to a city.

 

Its an elegant system that only leads to less "boring moments" and less tedium. You can build a house being confident that it will always exist as the best possible house for its enviroment. Its environment is taken in Two Variables. Tech and City level. No matter what type of Governor, all cities can reach level 5 (it would seem), however with a high Governing level, any city level can grow to a larger city size ... totally dependent on Governance. Perhaps some exceptions, although this seems to be the overall trend.

on Feb 22, 2010

pigeonpigeon,

The AI at deity Civ IV does not 'cheat', it just have very more loose rules to play with than the player has. It pays next to no upkeep for having lots and lots of cities, it pays less hammers for units, it starts with a lot of bonus techs where the player has only two, etc. This makes it so that if the AI has enough room to exland, it will have 15+ cities up in no time and then it will be more advanced than the player when you two make contact. This happens on immortal - one step below deity - as well, and on the lower difficulties too although there you can still smack the Ai around because it will not be all that advanced.

I agree the Ai needs these bonusses because otherwise it would not be very much challenging to play against it. In fact, it is not very hard to program an AI such that it could beat a player every single time, it is hard to make the AI very very powerful but still weak enough that good players have a shot at winning.

What I tried to say was that I would like a solid AI that is limited like the player in terms of development. I liked Civ IV in the sense that the 'masochistic' people like me could have a real challenge on deity. I still lose quite a few games on that level which is cool. What I do not like about that level is that the AI can expand like mad because it pays no upkeep, and when you spend several hours surviving and getting ahead a bit, you find that an AI on another continent had all the space to get 15+ cities where you have 6 or 8 or so, and that that AI is ahead in techs, and that you have no real chance to win any war against that AI. You are then basically screwed unless there is the odd chance you can cheese out a win.

Maybe what I describe is more a map-script issue than an AI issue, but I would love to see balanced gameplay in terms of solid players being able to maybe be slightly behind the AI yet win because they rock and all that, rather than finding myself hopelessly behind because the map decided right from the start the AI can have some slack and have lots of land + no maintenance + peaceful or no neighbors.

I still make very little sense maybe, but I feel the major flaw in Civ IV was that the AI could run away with the game before you ever met them making the game a foregone conclusion when the renaissance era arrives. It did not happen every time, but when it did it always felt like the win was not rightfully with the AI.

In the end all I wanted to say was that I want Elemental to not have that frustration.

on Feb 22, 2010

Thank you. Yes, that all makes sense, and I am for it -- under three conditions. These being

1. that there be no *disadvantage* to a suboptimally organized city other than or in addition to its own suboptimality

2. that there be various ways that a city might be optimized; there should be no "exclusively one single correct optimal" plan. (Of course that doesnt mean that some choices are just not as good as others, but there should not be only ONE way of (at least always) building a city.)

3. I would like to see automatic upgrades have different options for their resolution. Example: Your House+1 gets upgraded automatically at no cost once the tech (and/or whatever other prereqs need be met) gets available IF you have also invested in Quick Building (or whatever you want to call something, be it a Governor or be it a Tech, which gets you that bonus), otherwise, they upgrade automatically AT COST X (cost might be reduced by other factors), and they take Y turns to upgrade depending on how good your Governor and/or synergetic techs and/or special buildings also present might be (example: advanced forges with a very good Governor or however Boogie called them gets you free immediate better swordsmen once prereqs are met, advanced carpentry with very good Governor and special Groovy Woodshack building gets you free immediate House+1s when prereqs are met, otherwise, costs time and money). I just want there to be DIFFERENT WAYS for things to work out depending on what STYLE of Builder you are. Note that ALL of the options I list here require no micromanagement, they merely make the automatic upgrades better/quicker/cheaper.

What say you?

on Feb 22, 2010

leeboy26
My one hope is that there is an option that allows you to change the growth/building turn speeds of settlements in bigger games. For some reason games developers are very unimaginative when it comes to difficulty options. I'd prefer more than just 'easy' 'medium' and 'hard', give me custom options, damnit! (SoaSE was on the right track). I want to play a huge map that takes me real-time hours to build cities up, it helps emotional investment as well.

Couldn't agree more. Massive Epic Games FTW!!!

on Feb 23, 2010

The AI at deity Civ IV does not 'cheat', it just have very more loose rules to play with than the player has. It pays next to no upkeep for having lots and lots of cities, it pays less hammers for units, it starts with a lot of bonus techs where the player has only two, etc.

This is also known as cheating. Sure, it doesn't type in porntipsguzzardo (points to whoever gets it), maybe it doesn't have unlimited LoS, but the very fact that it pays less for everything and starts with massive bonuses is cheating. Rather than rely on the AI using good strategy and tactics to out-play the human player, it relies on endowing the AI with bonuses of all sorts to compensate for its lacking performance in those areas.

You could call it a handicap, but it really is just cheating. The AI is not playing by the same rules you are in an attempt to level the playing field (because you are a better strategist) or to provide a masochistic experience to those who enjoy such things

For example, Stardock prides itself on its ability to program good AIs without resorting to making them cheat. Take GC2 for example - the AI through the Tough difficulty setting plays by exactly the same rules as the player does, and plays quite well. Bump it up above Tough, though, and it starts to bend the rules - it pays less, builds faster, and so on. That is what "cheating AI" means. 

on Feb 23, 2010

Well you and I have basically different perspectives on cheating. You make it sound like starting with an advantage is cheating while in go-tournaments lower ranked players start with extra tokens on the board to make up for lesser skill, and in golf solid players and rookies have different handicaps since the rookies get a discount on the number of hits they have to count, therefore giving the rookies a chance against better players.

This is evening the playing field in my eyes, whereas cheating would be granting the AI the ability to create units out of thin air, being able to create 2 units in one single city in one turn, getting techs for free later in the game just because the AI desperately needs it, etc. This is something the AI does not do, they still play by the rules, just the rules are far far less harsh for them as they are for the player.

on Feb 23, 2010

When the AI is playing the game by a different set of rules then everybody else, the AI is cheating. That's the way the term is always used when it comes to AI, since the AI can't cheat by using things like hacks. You're describing the same thing in something like a tournamnet, where it's just called a handicap instead. Same thing, different terminology.

The test of an AI is how challenging it can be while playing under the same rules as the player.

on Feb 23, 2010

Well in that case the AI does cheat. In that case I also think cheating is not a bad thing since the players will figure out what the AI is going for and adapt their plans while the AI has no such ability. The AI needs some sort of trick to keep up with the player and to make up for the lack of 'vision'that a player has.

on Feb 23, 2010

When the AI is good to begin with, cheating helps it be better, yes. But a bad AI given cheating help is still bad.

AoW2 suffered from that. The AI wasn't very smart about how it would do things like forming up stacks. In the early game, its cheating advantage and the general fact that it started off with more terrain then the player made things challenging. You had to compete against superior forces to gain ground.

But the AI's fatal flaw is that it liked to use mixed stacks. Some high tier, some low tier units combined. Anybody good at AoW2 knows that due to the stack limit, top tier units are ALWAYS better for offense once you can build them. A hero and 7 top tier units will destroy any stack that isn't comprised the same way, and with the proper use of spells will destroy several weaker stacks at the same time.

So the player churns out one stack that the AI simply cannot defeat despite it's advantages, because the AI is simply playing the game wrong. All the cheating in the world (and at the highest difficulties it cheats substantially) and all th extra units it gets can't make up for its inability to make a decent stack.

Eventually you break the AI's main force, and then it retreats into defensive mode at every city it owns, stops being able to attack you, and you enter the mopup phase.

A good AI on the other hand would need less cheating, because it's better at using the advantages it has.

on Feb 23, 2010

3. I would like to see automatic upgrades have different options for their resolution. Example: Your House+1 gets upgraded automatically at no cost once the tech (and/or whatever other prereqs need be met) gets available IF you have also invested in Quick Building (or whatever you want to call something, be it a Governor or be it a Tech, which gets you that bonus), otherwise, they upgrade automatically AT COST X (cost might be reduced by other factors), and they take Y turns to upgrade depending on how good your Governor and/or synergetic techs and/or special buildings also present might be (example: advanced forges with a very good Governor or however Boogie called them gets you free immediate better swordsmen once prereqs are met, advanced carpentry with very good Governor and special Groovy Woodshack building gets you free immediate House+1s when prereqs are met, otherwise, costs time and money). I just want there to be DIFFERENT WAYS for things to work out depending on what STYLE of Builder you are. Note that ALL of the options I list here require no micromanagement, they merely make the automatic upgrades better/quicker/cheaper.

 

I think taking any time to upgrade improvements is going to bog down gameplay. This is a time of powerful wizardry and fanastic ability. Why wouldn't it only take a month or week to build a larger house? It only takes a week at most to do it in present day Washington state, and don't even have giants.

I like the simplicity of the current, and probably permanent, system we are using. I takes a type of management out of the game which is important for multiplayer, the main reason this game is better than gal civ, as opposed to a prequel.

 

              

 

on Feb 23, 2010

seanw3

////3. I would like to see automatic upgrades have different options for their resolution. Example: Your House+1 gets upgraded automatically at no cost once the tech (and/or whatever other prereqs need be met) gets available IF you have also invested in Quick Building (or whatever you want to call something, be it a Governor or be it a Tech, which gets you that bonus), otherwise, they upgrade automatically AT COST X (cost might be reduced by other factors), and they take Y turns to upgrade depending on how good your Governor and/or synergetic techs and/or special buildings also present might be (example: advanced forges with a very good Governor or however Boogie called them gets you free immediate better swordsmen once prereqs are met, advanced carpentry with very good Governor and special Groovy Woodshack building gets you free immediate House+1s when prereqs are met, otherwise, costs time and money). I just want there to be DIFFERENT WAYS for things to work out depending on what STYLE of Builder you are. Note that ALL of the options I list here require no micromanagement, they merely make the automatic upgrades better/quicker/cheaper.///
 

I think taking any time to upgrade improvements is going to bog down gameplay. This is a time of powerful wizardry and fanastic ability. Why wouldn't it only take a month or week to build a larger house? It only takes a week at most to do it in present day Washington state, and don't even have giants.

I like the simplicity of the current, and probably permanent, system we are using. I takes a type of management out of the game which is important for multiplayer, the main reason this game is better than gal civ, as opposed to a prequel.
       
 

 

See, indirectly this is my point. According to BoogiePac, at least my take on it, is that you Will NOT have to re-upgrade your improvements, as it will do so automatically. Of course, only higher level cities have access to the High-End buildings, therefore it takes a combination of upgrading your Overall city to a new level, as well as research in civilization techs, in order to get the best building capacity. Of course, once you achieve that, the buildings will upgrade on their own.

I believe that is the intent of the post, and I think this was essentially what you were saying??? gah, well at least once I start playing the beta I'll know what we currently have.

on Feb 24, 2010

It would be a mighty challenge to downtune the impact of improvements auto-popping up with techs. If one AI goes after tech-victory and another after questing, I feel like I should be able to keep up in power if I focus on my military might without even trying to keep up in techs or quests. Then again the only way for me to leverage my military advantage is to go conquer things... So the teching AI is going to get it first, and it should be able to put up a fight even though all my energy went towards creating a solid military and the AI teched a lot.

The same goes for the questing AI. If it will put up a good fight then I am very much pleased. What would be very bad is where the bonus an improvement gives will very much rocket propel your people forward to the point where you can barely be brought back in terms of winning the game.

on Feb 24, 2010

@Shurdus:

The teching AI would have better but fewer soldiers with more impressive equipment.

The questing AI would have more heroes with better gear but fewer troops.

Who would win would depend on who did better with their strategy.

11 PagesFirst 7 8 9 10 11