When Civilization 6 launches this October, gaming's longest-running 4X strategy series will receive its most significant update since it switched to hex-grid world maps. Lead Designer Ed Beach was kind enough to walk us through several of the incoming changes -- most notably "unstacking" cities by spreading districts across multiple tiles on the world map -- when he appeared on our E3 stage last week.
But because Civilization is such a dense experience, we decided to sit down with Senior Producer Dennis Shirk to speak in greater depth about the new city building mechanics, political agenda system, and cultural advancement rewards.
GameSpot: Let's start with the big change: cities that extend to multiple tiles. Why the decision to unstack cities, and how will this impact the game?
Shirk: What I'd wanted to do is change up the way the landscape basically affects your game. In Civ 5, you settled cities based on the resources that were in the area, but you didn't care necessarily about much else. Everything is built in the city center: buildings, wonders -- everything goes there. It's not a very interesting decision, you just build it. That's the only choice you have: to build it, [or] not to build it. By unstacking the cities, what [we're] able to do is make you think about where you're putting down your settlement.
If you want a city to be a science city, you know you're going to build a campus district. Campuses get adjacency bonuses, for example, from mountains and jungle. So if you find that perfect spot and you put down your city that has a lot of mountains and jungle, you're going like, "I'm specializing this city to be science-generating." That's also where you're going to build your library and your research lab or your university. And that's across the board for all the districts -- they all have strengths and weaknesses based on where your city's located.
They also take up a tile. So every tile they're taking up, you're not going to be building a farm there or a mine. So you have to balance this out. You can't build all the districts there, you can't build all the wonders, because you've got to be able to keep your people fed. You still have to do those basic things to keep your civilization moving. So as you're putting down cities and districts in different places, you have to specialize. Do you want more culture out of a city? Do you want more production? So you have to play the map. And that's the coolest thing, figuring out this puzzle of the map.
Fresh water is a lot more important now than it was before. You can't just plop a city down in the middle of the desert and expect to do well with it because it will never grow. It doesn't have enough fresh water access. There are some new concepts. There are two things that you need to do to have a city thrive: You need its people to be happy -- so they have to have enough amenities coming in, like luxury resources -- and you need the housing. It's just a concept called housing, but it's basically your population cap. So if you don't have fresh water access out of the gate, you're going to be trapped at one population because you have to have water.
So are maps still procedurally generated?
Did that present any design problems? Because, in theory, you could end up with a map that's terribly inhospitable and makes the game either unplayable or just un-fun.
Oh, definitely. After we put the system in the game, the first time we fired up an archipelago map, we're like, "Nope, not going to work." [Laughs] We had to make changes and adjustments to that. We found that players were restarting a lot if they weren't getting that perfect mix of mountains right away, because mountains are really powerful now for stuff like that. So we've had to retune the way the map generates and staggers stuff out.
But the district system [has] gone through many iterations, and it got to that sweet spot. We've made it so that, for example, not having mountains when you first come in can be offset by the amount of rivers that you have. We always have these balances in place now, where having the different challenges of different starting locations makes the game much more interesting. So you're not always going to have that same strategy when you come back in. I think that would be death for a game like Civilization, where it allows you to play the exact same way every time and you get to get away with it. That's not very interesting.
So you want to force dynamism on players, in a way; make them react.
Right, because you're not going to come back and play it anymore if you don't have those kinds of things. Something that [lead designer] Ed Beach had developed back on the Brave New World expansion when he was doing a lot of the AI work was a mayhem level. This is something that happens in the background; it's how they tune the game. You want this constant level of mayhem, kind of like the real world, where you never have quite this perfect world going on.
So if you're playing your perfect build or strategy, heading towards that culture victory, something will most likely happen somewhere that may take your attention off it for a little bit. Whether your ally is at war with somebody else and you have to make the decisions if you're going to help them or not, or it's happening directly to you. There's going to be something going on all the time.
How do you decide what's an appropriate level of mayhem?
We obviously do a lot of watching what human players do, because you want the AI player to be as challenging to play as if you're playing against a human. We've also got many systems at the office that's literally just playing itself all day long, and then the AI guys are just reading the logs and watching. We never really had that on Civ 5. We'd always have to execute games manually. And this way it's just constantly running, constantly collecting data all the time.
That is slightly terrifying. It sounds like Skynet.
No, it's not quite teaching itself, it's strictly data that's being fed to another guy. That would be a little scary if we were like, "OK, he's going to play, and you run an algorithm so that it gets better every single time, and learns from itself." Yeah, Civilization is the start of Skynet, and then it all goes south. [Laughs.]
Exactly! It's interesting to hear that you guys design mayhem into the game. It almost begs philosophical questions about the nature of mankind. If you're making a game that is literally called Civilization and you feel there's an inherent necessity for chaos and conflict, I don't know what that says about humanity. I'm guessing it wasn't intended as a commentary.
No, not at all. They're not even injecting mayhem. It's about controlling what the AI decides they want to do. So you have the knobs that you turn, and the mayhem level that they watch is just based on how the AI decides to play, how crazy they get, and making sure that's tuned to that perfect spot. You want that little bit of mayhem because it makes for interesting gameplay. In terms of real life, you don't want that at all. But real life may not be the most interesting game to play all the time.
What about players simply who want to perfect their own little corner of the world? Can they treat Civ 6 like a world building game and not so much like a conquest game?
They can. Well, first off, aside from what you can do in-game, there's always going to be that world of modding: people that design specific scenarios, specific ways to play. But just as in Civ 5, if you want to play a builder game, you don't invite Montezuma and Genghis [Khan] to the party. You go into "advanced setup," you make sure that you're setting the civs that are really going to all be builders. You choose all the builders in the game, and then just have a build-a-thon, and go from there. But if you want the party, you go random and see where they land.
So the idea of distinct AI personalities returns? Like, different Civilizations have pre-programmed behavior sets that will correlate with historical precedents?
Even more so now. Like in Civilization 5, Montezuma always played a very specific way. He'd probably be just rampaging and invading. But we have very specific historic agendas now, which allow the way you interact with the civs to be much more interesting. Like Theodore Roosevelt's is, when he's on his own continent, as long as other civs that are on his continent are not causing trouble, he's probably going to be friendly towards them. But if they're causing trouble -- they're going to wars or starting wars -- he's probably going to be their enemy.
So you can actually have some fun with this, because you might have Emperor Chin nearby, and he's causing you problems. You're trying to play that builder culture game. He's not necessarily going to let you do that because his historic agenda is that he's a wonder builder. He wants to have the most glorious civilization and as many wonders as possible. And if other civilizations are beating him in that wonder race, that's a problem for him. He might get jealous enough to go ahead and start stealing wonders from you, in terms of taking cities.
So maybe you do a little bit of baiting, and you start playing with this information, and you draw him into a war with you, because then Teddy Roosevelt will most likely come to your aid. And now you've basically got this formal war that nobody is going to be angry at you about -- because you didn't start it -- and in the end, you've come out on top. So there's a lot of things that you can play around with these agendas now that make the game so much more interesting.
Are civilizations' historical agendas going to be apparent to players, or is it something they're just going to have to figure out over time?
No, the historic agendas are the one thing that they can see when they come in. It's discovering all the rest of how [leaders are] playing they have to figure out.
How do you actually pick which leaders represent each civilization?
This time around was based around what we thought the personality would do for the game because certain leaders are going to fill certain gaps. We want X number of leaders that act this way, some that act this way, some that act this way, so you can have that balance going into the game. So it's not just about if they were a great leader, it's what they do. Whenever we have a leader that we think would bring a really interesting historic agenda to the table, like Teddy Roosevelt with his Big Stick policy, that's usually how we choose them. Of course there's going to be a balance. Some are Civ stalwarts. Some places really expect certain things to be in the game. We also always want a percentage that have never appeared in the game before, so we try to keep a percentage of those as well.
So obviously players can set their own political agendas as well, but it seems like that system has changed slightly. Can you describe the new system a little more deeply, and explain how it will impact the actual gameplay?
Well, we split the trees now. Before you only had the tech tree, so culture players or builders were at the mercy of people who were driving hard science because they'd always have more advanced units and eventually could steamroll you if they wanted to. So I really wanted a way for the cultural player to compete in the world. So we have a culture tree and we have a tech tree. But the thing is, the culture tree is where all of your policies live: the ability to unlock new governments, the ability to unlock more cards, these policy cards.
For example, you have militaristic cards, economic cards, you've got wild card slots, things like that. Now those are only unlockable in the civics tree, so if you're playing a hard culture game and you go deep into that tree, you could unlock some really advanced governments and a plethora of cards to choose from because you're the shining beacon of cultural awesomeness in the world. Whereas if you're playing a hard science game, you don't necessarily get to go as deep, you've got a more primitive government. You're not as enlightened, you're just strictly on the tech, you want more advanced units.
So as a result, if somebody is threatening you -- one of those science players -- they may have infantry and you've only got riflemen because you're not as far technologically. But because you've got all these great military policies plugged in, your riflemen are about half as much to produce and they've got all these benefits and they're 50 percent stronger in certain areas because you've got that flexibility of government, and now you can stand toe-to-toe with these more advanced civilizations.
Will this affect diplomacy at all? Can you talk about anything that's been added, or any layers of depth or new mechanics that have been added to the diplomacy system?
We're not talking too much about diplomacy, but what I can say is, [leaders] also have random agendas, and they're hidden. An example would be, maybe [a leader] loves industry. That could be the second agenda that he has. You can't see it because you don't have exposure to the information. So if you're this weak-kneed culture civilization, his opinion will start going down; you don't know why.
So maybe to expose some of that information, you establish a trade route. And now that unlocks a little bit of visibility. You start getting rumors and gossip coming back. Still don't know what that hidden agenda is. The next thing you do is you send a delegation. He likes that, that's great, you get a modifier from sending the delegation. Plus the delegation starts sending you information. Now you realize, oh, he's an industry lover, that's why he didn't necessarily like me, that's why that negative modifier is there. So now you can kind of adjust to that as well.
And then of course you can keep going up the information chain to spies eventually, things like that. But [with] a lot of that information trading, you can become a merchant of information if you want, depending on the certain civ that you're playing -- because you've got these different levels of visibility based on your engagement, you've got these different things that you're trying to discover.
Do you judge your coworkers based on which Civilizations they tend to prefer?
Silently, yeah. We don't say it verbally, but we always do.