The expanding popularity of an Arma 2 mod, Day Z, might have surprised all of us, but imagine the surprise felt by the chap who created it, Dean “Rocket” Hall. There are nearly 48,000 characters now registered in the game’s stats, when he imagined there’d be just a few hundred. His motivation to make the ultra-bleak multiplayer zombie survival mod might not surprise any of you, though, when you read his take on what games should be, and why the kinds of stories experiences like Day Z produce are so important. There were a couple of times in this interview where I hooted in agreement with what Rocket had to say. See if you can spot them.
RPS: Am I correct in understanding that you work for Bohemia – the Arma 2 devs – in the Czech Republic? What do you do there?
Hall: I sure do! I am working there as a game designer in multiplayer. My responsibilities are with multiplayer for Arma 3.
RPS: Are we calling it “day-zee” or “Day Zed”?
Hall: Day-zee, we went with – I actually think the name is terrible, and I came up with it. It was really just a working name originally, but it started as a working name and now it has stuck. It’s grown on me as someone came up with a new logo and so on. So perhaps it’s not too bad.
RPS: I like it! “Daisy” makes it good for puns, too, which definitely works for our audience. So you’re working on this in your spare time? Or are Bohemia allowing this to leak into work life?
Hall: Well, there’s obviously some crossover. It’s good for experimentation, so that has helped. And it’s been something to do on weekends! I don’t speak Czech [Hall is a New Zealander] so it was helpful to have it as a project to sink some time into. And the same was true when I was with the army back in New Zealand – I’ve been at Bohemia since January – where I was holed up in this tiny little town with an internet connection and not much to do. Give a man a computer, an internet connection, and not much to do, well, wonderful things can happen.
RPS: So this idea has been a while in the making? Can you talk a bit about the genesis of the Day Z idea?
Hall: One of the first things I did when I got into modding was to look at making something zombie-related. That was the first thing I did when I started modding for Arma, when it came out. But really I guess it really came from a desire to play more games that give you an emotional connection to your character. I found that with most of the games I was playing, while they were interesting and entertaining, I didn’t find that I had any contextual reference with the character – the character would do things I wouldn’t do, because of the story. It’s like when you watch a horror movie and yell at the characters “why are you doing that!?” I was having the same kind of response from gaming experiences. And I wanted to change that.
RPS: Right, and Day Z is story-free and open-ended, which means players are compelled to tell stories about what happened, because what happened is unique to them – it’s not the same story that everyone will have seen in, say, Half-Life, or whatever. You must have started hearing those stories from Day Z pretty early on…
Hall: Yeah, that has been extraordinary, and it’s an extraordinary experience for me to read around forums and blogs and find people getting into their characters and what happened to them. I read about people talking about shaking from the adrenaline of situations they’ve been in. It all cuts to something quite unique about humans, really, that ability and need to tell stories – that’s how people learn and it’s how people pass on information. The popularity of the mod stems from there, that fundamental place. It only stirs some very simple emotions and responses from people, of course, but it’s given them a burning desire to report what has happened, to tell the stories about the experiences it generates. There was no support structure for the mod, and so it spread through things like social media. There was no announcement, no promotion. The popularity of the game is just down to people wanting to tell stories.
RPS: You must have been surprised that it was so popular. There’s forty-seven thousand people registered on the website stats now!
Hall: Well, I knew it would be popular within the Arma community. The Arma community is a really strong community, especially in its approach to innovation. They like to see new stuff come out, they get on board, and then they make new stuff in response. So I expected that, but what I didn’t expect was the crossover into the mainstream. Arma has never really had any crossover, because people write it off over little difficulties and don’t get to see what is underneath. But it has crossed over and coping with capacity has become my focus for the past two and half weeks. Every time we would upgrade something, it would be filled to capacity. We launched a server and it’d instantly be full up. We moved the website to a bigger server twice, and then to a cloud server. And it was like solving a traffic jam. We’d move the traffic from one intersection and it would move and jam somewhere else up. We haven’t had a chance to really think about what has been happening, we’ve just been dealing with that capacity issue.
RPS: Is there a ceiling you will hit from the persistent architecture of it? Can you keep upping capacity?
Hall: Well I know people are having trouble logging in, but that’s more to do with the interaction between the local server and their nodes. The persistence server can handle up to ten thousand concurrent connections, and we have peaked at five thousand. So that should be able to handle a lot more load. And I sort of think surely we couldn’t hit ten thousand concurrent connections, right?
RPS: Haha, well…
Hall: Actually, I did think we had peaked six hundred, so I suppose we will see what happens.
RPS: Yes, let’s see what happens on Sunday night, after another week of intense coverage for the mod.
Hall: There’s two things I dread at the moment. One is the weekend, and the other is when I roll out an update. You’re never quite sure what it’s going to do, and it’s a nightmare to roll it out to 120 servers and then check it’s working on them. You’re never quite sure it’ll work.
RPS: Ah, updates. How regular are they going to be now?
Hall: Well it was supposed to be nightly builds, whenever me or someone else from the team committed assets. We were trying to follow with the whole pre-alpha to alpha model thing. But that’s what you do when you have maybe two servers and fifty people. You can’t really apply it to thousands and thousands of people. That has forced quite a serious rethinking of where to go. We’ve wiped the entire plan and focused on meeting capacity. We could have buried the project or limited it, but instead we’ve decided to try and meet the demand. So it’s just managing the crisis before thinking about anything else. There will be an occasional update thrown in too! [Perhaps like forthcoming dogs, as revealed in PCG’s interview with Hall.]
RPS: What do you want Day Z to do, once the capacity crisis has been resolved?
Hall: The experiment has to continue. Because that’s what big companies can’t do: to take risks and experiment like this. They can’t risk upsetting their userbase, they can’t risk messing with existing formulae. They can’t add radical or brutal features, and risk getting it wrong. But this project can. In that sense the experiment has only just started. If there’s one area I really want to see grow, it’s that the environment is also part of your consideration: the weather, the terrain, and so on. I think you should not just be challenged by your interactions with the zombies and with other players, but you should also be facing the world. That will add a contextual link to a map as wonderful as Chernarus. You will have to start worrying about shelter, thinking about the rain, and so on.
RPS: One of my favourite discoveries was that there were wrecked vehicles that we could then fix up to move about more rapidly. We then discovered that certain places seemed to yield more of the right kind of loot to fix up vehicles and so on – was that intentional? Or am I seeing patterns in random loot spawning?
Hall: Obviously because it was just me placing things, I adopted for a very modular approach. The type of items spawned for you to collect are based on the type of buildings, or types of buildings: whether it’s residential, military, or industrial. Because I wanted it to be intuitive to your personality. You are in a zombie apocalypse, stood on a beach: some people are the kind of people who will say “let’s go to the military base and get tooled up!” Other people might say “I’m going to go to a farm, get a farmer’s rifle and just be careful.” Others still might strap flares to themselves and run into town to try and help survivors. I kind of wanted it to be semi-intuitive, and for people’s responses to make sense, and to work within the world. So I made it class-based, dependent on the buildings. And that means it can be quickly applied to another map, too. It would take about an hour to apply it to another similar-sized map.
RPS: How much feedback have you taken from the way people behave once they’re inside the game. The bandit/survivor distinction has obviously been put in to identify those who who are killers from those who are a little more trustworthy?
Hall: The bandit and humanity system was added for a few reasons. One of them was that I wanted to make systems that do not imply judgement: they should not tell you how to play. However, there also needed to be impact to your decisions. There will be decisions such as “do I pick up the ammo or do I pick up the food?” But you also face decisions like “do I shoot that person, or do I not?” If you shoot the person, there should be some effect from it. There shouldn’t be a direct negative consequence, of course, it shouldn’t tell you how to play, but there needed to be something. Anyway, what happened when we set up the original European server was very different from the original, relatively peaceful New Zealand server. Suddenly everyone was killing each other. I think the language barrier came in there – German, Russian, and English speakers – and it rapidly descended into chaos. The average life expectancy dropped to something ridiculous like thirty minutes. We knew we had to do something!
So what we did was implement that bandit system, which highlights the killers. But I don’t think it works. I think we need to have skins that are based not on your humanity, but on things that you find, craft, and use. That should allow people to craft their characters how they want. To appear as the character you actually play.
RPS: I was surprised, actually, by the range of niches that people have taken on. I anticipated the sort of PvP aspect of it, but actually a lot of people just take the woods and try to survive more peacefully. Are you surprised by that diversity of roles having sprung up?
Hall: Well, I guess I thought that kind of person was the sort of person who would play it. And certainly I anticipated that something like this would happen, but I didn’t anticipate the scale of it. I wanted to create the kind of game where not only you had to live with the decisions you made, but you also had to live in a world where other people had made those decisions. The system has to support the range of responses. People exaggerate how much player-killing goes on, too. They say “oh there’s slaughter of newbies”, but only about ten percent of the population is a bandit at any one time. I also don’t think it really fits traditional descriptions of PvP, either, because there isn’t a systemic approach to it. It just says: play it however you want. Sometimes killing other players is a part of that.
RPS: The persistence aspect of the mod is really fascinating to me. You could have made a mod where the character simply existed on individual servers. Why create a system which meant you could log in with the same character on any server?
Hall: Well I guess it’s something I’ve been really passionate about for a long time. I’ve wanted more games to do it. It would make my life an order of magnitude easier if it wasn’t persistent across all servers, and a lot of people say it shouldn’t be, but I really think it would take a way a lot of what makes it.
RPS: Oh it definitely should be. That’s the most interesting thing you’ve done with it. Really, the persistence is absolutely key to the experience.
Hall: You can’t just do things because they’re easy. So that’s what I’ve stuck with it even though it’s made things complicated. I think it provides context for you as an individual player, which is something I think was missing from a lot of the games I’ve played in the past. In those games things didn’t seem to matter, but if it’s persistent, then things sort of matter. It switches something over in your head. The decisions you make are different if you know that this character is going to be there tomorrow. Your decisions become more complicated. This mod is less about what’s in it – there’s not much to it – it’s about what is going on in your head.
RPS: Yes, and elements like persistence are a sea-change for that kind of mental disposition towards a game. Anyway, the upshot of all that is that Arma 2 is selling on Steam again. That’s quite an accomplishment. What’s the reaction been like at the studio?
Hall: Well to be honest we’re all just a bit shocked! And it’s only just started to sink in that it’s real today. Before that it was just numbers on the website. And I was overwhelmed with problems! So I was dealing with those. But it started to become real as I got more and more messages, until I got a message from a friend saying “Hey, have you checked out this Day Z mod?” And I had to reply “Yeah, I made it actually.” And he sort of wet himself! So it’s been crazy.
RPS: Yeah, the way it spread like wildfire has been quite the thing to see. I knew it was going to be interesting just from waking up one morning to find so many links to it in my inbox.
Hall: The reaction of the community to it has actually been the most interesting aspect of this whole experience. I think there’s a real lesson out there for game developers, a lesson about how gamers take things on when they can generate their own stories. People want to get on board when you say to them “we’re going to build the world, you’re going to populate it and decide how it actually functions.” People really do want that.
RPS: Yeah, and we shouldn’t have to prove to publishers and developers that we want that, anymore. A whole bunch of games now prove it. And one of the main responses to Day Z from fans has been “why hasn’t someone made this game before now?” It seems sort of obvious. Obviously we all want to play a zombie survival game like this, so why hasn’t anyone made it? Do you feel vindicated?
Hall: I am trying to think of a way of answering that question which is G-Rated… Because I had a response, but it’s probably not appropriate. I’d been pitching this kind of idea for a long time. And I pitched it when I worked as a producer. I was told things like “players say they want stuff, but they don’t really want that stuff, and when they get it they will hate it.” So I crated Day Z a bit as a “take that!” in response to that. One of the guys I pitched it to has since emailed me saying “Well, I guess I was wrong.”
But yeah, there are a lot of gamers out there who want this sort of game. I guess like just like me they want their experience to have context. But also they want games to take advantage of the technology that they already have. They want games to create something unique to games. For people to generate stories that are unique to the player – books and movies can’t do that, only games can – so why are we still trying to make games which act like books or movies? It doesn’t make any sense.
RPS: A perfect point to end on. Thanks for your time.