Share this project


Share this project

Lead a barbarian kingdom to glory during the final days of the Roman Empire. An empire builder by Jon Shafer, designer of Civ 5.
Lead a barbarian kingdom to glory during the final days of the Roman Empire. An empire builder by Jon Shafer, designer of Civ 5.
3,009 backers pledged $106,283 to help bring this project to life.

Building an Empire Builder

Empire builders (aka "4X turn-based strategy games") are a beloved genre. Unfortunately, they're also somewhat of a rare breed, particularly when compared with the deep yearly lineup of first-person shooters and RPGs.

The problem with 4X titles is that they're not easy to build. Challenges hide behind every corner - not just on the design side but also with the tech and art.

Today, we'll delve into the obstacles developers of these games must face, along with why the end result is worth all of that hard work. And maybe if we're lucky, this article might help motivate someone out there to go and create one of their own!


Designing a 4X Game

Unique gameplay is what sets empire builders apart from all other strategy titles, and a handful of attributes are especially important in this recipe.

The most crucial of these is that you shape the world, not the designer. Players start with a single colonist, city, or planet, and from this lone seed, grow an entire civilization. Everything is shaped by your hand. This level of ownership is empowering, and sadly, very rare in gaming. This is why the deceptively-humble Minecraft can become one of the business' greatest success stories.

The bad news for 4X designers is that an incredibly free-form experience of this sort is fragile. Rather than laying out a clear, controlled path for players, you must instead trust completely in the web of mechanics you've woven. Is it too easy to build a massive empire? Too hard? Is a very niche, unassuming strategy actually so powerful that it's pointless to do anything else? A single flaw in pacing or balance can bring down the entire game.

The risks might be great, but the payoff for successfully walking this tightrope are unmatched - there's just nothing as satisfying as successfully forging a mighty world empire! What are players building this empire on? A map, of course.

But in a 4X game not just any map will do. No, we're talking about a random map. This amazing feature is practically unique to our genre, which is kind of a shame as it's one of the best in all of gaming.



One benefit random maps provide is a sense of discovery. What's out there? Who knows! It's different each time. One of your primary tasks is to venture off into the unknown and find out. Exploring and experiencing new environments is a major reason why some people play other genres, such as RPGs. Well, 4X games have the added advantage of this simply being just one of many excellent bullet points!

The second advantage is the need for players to adapt their strategies to the circumstances. If a single-player game ships with only six maps it's eventually going to be "solved," and boiled down into a small set of optimal strategies, ala tic-tac-toe. But when you have no clue what obstacles, resources and opponents are out there you can never guarantee you have the perfect plan.

These two elements combined are the reason why 4X games are unmatched in terms of replayability. Not only are there are an infinite number of worlds to explore, but every time you play you'll be faced with new challenges and have new opportunities to take advantage of.

So what are the challenges associated with random maps? Well, the biggest one from a design perspective is ensuring your worlds are not just fun to play on but also believable. If you're developing a 4X title based on history and your continents look like big squares then you have a problem! A game doesn't need to be realistic, but it does need to at least be believable and roughly match with players' expectations. Design aside, procedural worlds also have major technical implications, but we'll get to that in the next section...

The final perk 4X games feature that we'll talk about is the value offered by overlapping systems. Wait, isn't this an important feature in pretty much every game? That is indeed true, but the formula used in empire builders is unique. That addictive "one more turn" feeling comes from always having something just around the corner to look forward to. As one aspect of the game is winding down, another steps in to take its place, and by the time you finally check the clock you realize that there's no way you're going to get a full night's sleep!

As with most aspects of a 4X game, proper pacing and balance are the designer's biggest hurdle. It's not easy to craft a ruleset where the pacing isn't fixed but still gives the player decisions and rewards at nice, regular intervals. This is one of the reasons why iteration is so important - your first one, two or even ten attempts will miss the mark. But maybe with the eleventh everything will finally fall into place!



Architecting a 4X Game

When people think of technically-demanding games, rarely are 4X titles the first that come to mind. But most hardware has been developed with one goal: drawing large, highly-detailed 3D models. And how many of those do you see in empire builders? It's hard to say, but the leaders in the recent Civ games are the only example I can come up with!

What 4X games do typically have though are lots small objects. And by lots, I mean LOTS. Between the every building in every city, each individual character in the units, hundreds, maybe even thousands of trees... it adds up quickly, and if you're not careful this can bring someone's computer to its knees.

Making a game's art two-dimensional (ala ATG!) can be a huge help, because then you're only drawing one graphic per object, instead of needing to render dozens or even hundreds of tiny sub-pieces for each. However, if you decide to go 3D you're probably going to have to build your own completely-custom suite of technology in order to maximize performance, as middleware engines aren't designed to solve the problems present in 4X games.

Drawing all of those objects is certainly difficult, but managing them might be even worse. We've already talked about the design issues associated with random maps, but the technical challenges are actually far more daunting, which is why so few games include them. Cramming everything you need into the limited processing power and storage space available is already tough. And on top of that you have no idea what the world is actually going to look like? Yikes! For most teams, this is a problem where the available solutions are simply too expensive in time, loss of graphical fidelity or both.

Another demand placed on the programming staff, often overlooked, is the need to make systems modular and allow for rapid and extensive iteration. There's no three-step guide to designing a strategy game, and you're going to make a ton of mistakes along the way. Many features will need to be completely retooled and possibly ripped out altogether. If your code isn't written with this in mind, it can get messy, quick.



Beautifying a 4X Game

The big artistic challenge in every strategy game is balancing clarity with style. Interesting, difficult decisions are the bread and butter of this genre, and for those to exist your situation needs to be clear and the options available to you even clearer.

Assets typically look more impressive when viewed against a blank background and much larger than they'll actually appear in the game. Artists can fall into the trap of making their work as detailed and realistic as possible simply because they can. A realistic unit over a realistic improvement over a realistic resource over realistic terrain will typically give you a jumbled, unreadable mess.

Instead, units should pop off of the terrain. Forests must embrace their role as a canvas for the objects which rest upon them. The interface needs to make it clear what you can and can't interact with, rather than trying to compete with the rest of the art. Each piece of art might not be terribly impressive when looked at individually - but when combined they will transform into something truly beautiful.

The same lessons that apply when designing a 4X game also hold true for the art side. The strength of your game is determined by how well all of the small pieces come together to become one. There is truly no better example of how the whole can be greater than the sum of its parts!

- Jon


If you’d like to discuss this topic further (or anything else related to ATG!) be sure to stop by the official Conifer Games forum, and become a member of our growing community!


Only backers can post comments. Log In
    1. Adam Solo on

      You're providing great updates Jon. I've been reading them all and having lots of fun. Lots of good insights.

      Will be waiting for that beta invitation ;)

      Good luck man.

      PS: Civ5 may not have been what we would have hoped for at the release time, but I still think you did a great job, especially on the (bold) decision to go with 1UPT (which clearly will mature in time) and the introduction of city states (a great and fun mechanic). The social policies were also a good feature.

    2. Jon Shafer Creator on

      Thanks everyone!

      @Jørgen Granseth

      I decided to move away from 1UPT as I didn't feel it suited warfare of this era. Instead, the focus is mainly on supply. I might give 1UPT another crack one day, but it'll have to be appropriate for the game in question. Who knows, maybe a WW1 game is in my future!

      As for limiting stacks, there is a mechanic in place right now where the more units that are stacked in a tile, the less supply is available. I have no idea if it'll work in the long run, but that's something we'll have to see in practice!

      - Jon

    3. MadMatt on

      Excellent stuff, please keep going with this behind the scenes look into 4X game development. This is what really got me to back the project (there's more stuff at the Conifer forums, FYI, guys).

    4. Missing avatar

      Jorgensager on

      Hey, I just wanted to say thanks for these updates. They're really interesting (there's no such thing as "too long", by the way!) and made me want to back the project! The game looks great as well, of course, but sharing your game design ideas on this level feels like more than what we'd get if the game were funded otherwise!

      I'm a bit sad to see you're leaving 1UPT for ATG as I love the tactical aspect it brought to Civ5, but I understand the AI argument, and it probably makes a better game if the computer plays it well, hehe.. Will you limit stacks somehow, or will they naturally be rather limited by supply? Also kudos for going 2D and single player to focus on gameplay... it's definitely the right choice when resources are limited! (... and speaking of limited resources, that sounds like a good choice as well [/wordplay on game design]>! ^_^ )

    5. Gregory Curtis on

      Thanks for the great update. I really appreciate your view on the graphic elements of these types of games. The balance of clarity vs. style is something that I feel a few lose track of during the development process. Look forward to more updates.

    6. Missing avatar

      Stephen T. Bacon on

      Agree with @Alan Murphy - it's nice to see your thought process for the design behind ATG.
      re: Camel Case - I'm guilty of m_codeTomfoolery (I mean, I'm "old school"). Luckily Intellisense wasn't invented until the Renaissance, I believe, so I'm safe for the ATG time period :)

    7. Xenophon on

      Awesome update! Really glad I backed this project.

    8. Kevin Konkler on

      Hahahaha, I'm a Java developer and after I commented, I did some research on C# practices and they are different than I expected. Sorry to wrongly bust your balls!

    9. Jon Shafer Creator on

      Thanks guys!

      @Kevin Konkler:

      Because camelCase is obsolete, and must be purged. ;) Actually, no, we do, it's just that those are class member variables, which are capitalized to differentiate them from locals. Intellisense means you don't need to bother with the whole m_ppixyz tomfoolery. It is a shame though, as my fondest programming memory was looking through the Civ 4 codebase and lo' and behold, my eyes rested their gaze upon the best variable name ever: m_iDirty. Yes, yes you are.

      - Jon

    10. Kevin Konkler on

      Thanks for the awesome updates! I really appreciate the inside look.

      Also, why are you using PascalCase for your variables instead of camelCase? O.O

    11. Missing avatar

      saluk on

      These in-depth discussions are really building my faith in the game.

    12. Missing avatar

      Raymond Lee

      Just wanted to pop in and say that these updates and behind the scenes look into the building of a 4x game have been incredible. Whenever a new update comes in, I have to set it aside for a more in-depth read later :). Thanks!