A few months ago, Gamasutra brought Conan Exiles creative director Joel Bylos onto our Twitch channel to discuss Funcom's hit online survival game. Recently, Gamasutra's Bryant Francis and Kris Graft had the pleasure of speaking with Bylos again about the game's launch on Xbox One's preview program, as well as the design of the game's new Frozen North expansion and what new features can be expected from it.
Joel Bylos, creative director of Conan Exiles
Bryant Francis, contributing editor at Gamasutra
Kris Graft, editor-in-chief at Gamasutra
Bryant Francis: The climbing mechanic and world traversal, can you talk about how it's implemented right now?
Joel Bylos: We don't have the leeway to give a paraglider, or anything like they have in Zelda, but basically any surface in the world is climbable, unless we deem it not to be, which we've done in some cases like dungeons, just to prevent people from exploiting certain puzzle mechanics. Essentially you can walk up to anything in the game, if you hold the jump button you'll attach to it, you can do it right from the start if you like, and you'll be pushed into a third-person camera view just to make that work, but basically you can jump up and attach to any of the rocks or cliff faces, and you climb then. Your stamina bar determines the amount of climbing time that you get, and so forth.
We had to refigure a large number of the rocks in the world to make their collision more precise, because otherwise the player would be floating in the air while climbing. And there was areas the players could reach now that would be outside of their theoretical reach before. Pretty much in this game you could reach almost anywhere by building there, we kind of knew what when we were making it, so we did take the time to make sure they looked okay when you got up to them, but now we had to make sure that the actual attach surfaces were fairly close to perfect when you actually got on them.
"We had to refigure a large number of the rocks in the world to make their collision more precise, because otherwise the player would be floating in the air while climbing. "
We had to make sure locations worked well with climbing. We had to make sure locations didn't break with climbing, like in dungeons. We knew that some of our puzzles would break so we made some of the dungeon walls unclimbable. We tried to make that realistic by making them slick with water, making them difficult to climb.
A problem we had been trying to solve in the game was that people had been building these bases in places that were basically unreachable and then destroying their staircases each night when they logged off. In PvP, they would build a staircase up to their base [...] on top of a very high rock, then they would delete the staircase so players could never get to their base to raid them. It was clever, but it amounted to an unraidable base game, so we decided to add climbing as a way of at least reaching those bases.
Kris Graft: It's one thing to do this in a game like Zelda, that's single-player, you can kind of take into account what players might do. But in a MMO game like this, you could be thrown some curveballs.
Joel Bylos: Yeah. But it's a relatively robust system, it's one of the most polished systems we've launched. It could use some animation polish but it actually works very well, players have not been finding massive exploits or anything like that. It also solved a problem with world traversal with people, who were up on high cliffs, getting down quickly. It's never that fun if you have to find a way down, it's much more fun to jump off a cliff. So we have this whole mechanic where you do what we call the 'heroic plunge,' where you jump off a cliff and spin and grab the wall on the way down, and you'll just slide with both your hands.
Kris Graft: That's how I jump off of cliffs as well.
Joel Bylos: (laughs)
Kris Graft: The other side of the climbing mechanic in Breath of the Wild is the paraglider. Did you consider putting that in, or could something like that be implemented in the future?
Joel Bylos: We've discussed things like that. We can't obviously do paraglider. I mean we could but it'd be kind of lame, it doesn't really fit the setting that well. I don't know if you've ever seen a movie called Yor: Hunter from the Future? It was like an 80s cheesy barbarian movie. In that movie he captures a giant flying lizard creature and uses it as a paraglider, it's kind of amazing.
The unforgettable film Yor: Hunter From the Future
So we talk about things like that we talked about being able to skin certain creatures and maybe use them to do something like that. But we'll see, we haven't reached that point yet.
Kris Graft: Yeah. You previously said everything has to be 'super violent,' I think like skinning a creature, or capturing a creature and then forcing it to be your paraglider...I think that would fit in the Conan world, that's just my personal opinion though.
Kris Graft: You hear a lot of people say that one of the major advantages of developing on console is that you just have one or two kinds of platforms to get it running on, and if it runs great one, it's going to run great on all of them, because they're all the same. And you have definitely found that to be true, right? (sarcasm)
"I was naive enough to actually think that. If it runs on one Xbox One S, it'll run on all Xbox One S's, if it'll run on one Xbox One, it'll run on all Xbox Ones."
Joel Bylos: (laughs, sarcastically) Yes, that is definitely our experience. I was naive enough to actually think that. If it runs on one Xbox One S, it'll run on all Xbox One S's, if it'll run on one Xbox One, it'll run on all Xbox Ones. What we found was, for some reason it might be crashing for someone's, and it could be not crashing on someone else's, while they could be running the exact same console. They could have no external hard drives with basically the same setup as you.
So yeah we found that it's been interesting with the console. I honestly believed that it would run the same for everybody. On the day of launch people were telling me it was crashing constantly. And I was running it at home, and I was crashing almost never, my minimum time-to-fail was like five or six hours. People were telling me they were crashing every minute, every ten minutes. From a development point of view that can be slightly frustrating, because tracking down those errors is also quite hard.
Kris Graft: How successful have you been at tracking down those errors? Have you been able to do that efficiently?
"When you're testing for stability you have a MTF, a Minimum Time to Failure, which is basically the average amount of time any of them playing the game before they crashed."
Joel Bylos: Yeah I think so. We've been patching every week, because there can be multiple causes of any crash, right? Patching everything as we go. Our minimum time to failure on every version, which is how our testing unit gives us feedback.
When you're testing for stability you have a MTF, a minimum time to failure, which is basically the average amount of time any of them playing the game before they crashed, and that has increased from two hours to seven-and-a-half. Which is obviously a huge stability improvement. Which doesn't mean on the average that people crash every seven-and-a-half hours, it just means that that's the minimum before they force the game to crash.
Kris Graft: What reasons have you found that has caused a crash or a bug that you didn't expect? Like, somebody had Daytona USA installed, and that was the reason that people were experiencing crashes.
Joel Bylos: (laughs) I don't think we've found particularly other games to be the source. We did find people running external hard drives, was one cause, could have been one of the causes. Here's some juicy stuff. Before launch, in Europe we had the rating to show full nudity. Dicks on consoles, essentially. The way we worked with Microsoft to figure it out was to have DLC that enabled nudity in Europe, so that people who could download that DLC, in Europe only, to get the nudity on their characters.
We had set this all up, we had tested this version. The version we were going to launch with we had tested for two weeks, it was very stable, it wasn't crashing much, the coders had done a lot of things to it. And we had other versions that were waiting to be basically patched out later, because those were the versions that we'd been testing. And those versions [hadn't] undergone thourough testing, at least not nearly as thourough testing as normal.
"Microsoft found that there were people getting around and getting the nudity DLC in places that they shouldn't, and we had to quickly release a version that was newer, and hadn't been tested well, and we had launch problems."
So we had been working on them. And then about two days before the launch, we launched on a Wednesday and Microsoft called us on a Monday night, and basically they found a problem in their store, where people in the US were able to get around the region lock and get the DLC for nudity, because the trial had accidentally launched early. People were downloading the game in trial mode well before it was actually launched. A few weird things going on there. So basically people were figuring out a way around [the region lock], so we had to disable the nudity in Europe or else we'd get fined by the ESRB.
So basically we had to disable that version of the trial, that a bunch of European people had downloaded, that had nudity DLC as well. So we had to disable that, and release a build on launch day that had not been tested in the way that the other builds had been. So on the Tuesday night I had played for six hours with pretty much zero crashing and zero problems. And then on Wednesday we had to quickly change our version.
And so that's shenanigans, but that's what happens sometimes. It's nobody's fault, it's just that Microsoft found that there were people getting around and getting the nudity DLC in places that they shouldn't, and we had to quickly release a version that was newer, and hadn't been tested well, and we had launch problems.
"Some people have twenty thousand hours in the game, and some people have a hundred, but they still have an opinion about this stuff, right?"
Joel Bylos: Today I did a temperature test with people. On Fridays we have a meeting with the entire team, where we just talk about what everyone's been working on, so people all in the loop. Often what I do in those meetings is we thumbs-up/thumbs-down/neutral on a feature, just to see how people feel about things.
And so I did a temperature test today on the dev team, for several of the features that are left to complete for the game. So it was quite interesting. I was like 'how do you feel about this feature,' is it thumbs-up, thumbs-down, neutral, and there's also, if you'could only pick one of these features on the list, which one would you do? And so, very diverse and interesting results based on the type of player that our devs are. And many of them are very different, there are people who play a game solely in PvE, they like to be on a server with other players, but they don't like to do PvP. And they're not interested in interaction on that level of continually struggling against other people.
There are people who literally are just giant jerks, who go around killing as many people as they can find on a server, and destroying people's bases, and teabagging them. So it's interesting to see what people feel about things and what they think is important for the game. And it's super interesting because some [of our devs] have 20,000 hours in the game, and some people have a hundred, because it's a product assistant, who gets coffee for the project manager. He might have only played the game for a little bit, he might have only been working a phone job for a few days. But he still has an opinion about this stuff, right?
Kris Graft: That was going to be my next question, about cutting features, about how you get to that decision, about which of your darlings you're going to kill. Is it just super democratic, like that?
"I try not to do democracy in game development. Design by committee can work, but it's difficult."
Joel Bylos: No no no, I try not to do democracy in game development. Design by committee, it can work, but it's difficult. I think it's more like design by merit, so people come with really cool ideas and implement them, and then you're like 'oh yeah, that was an amazing idea. Great work.'
We don't democratize it, what we do is, I don't know if you saw that presentation where they were talking about their feature scorecards? We have a similar system, where basically we say, does this feature market the game? Will it sell copies in the market? Will people buy a game if they see a video of this? That sounds super cynical, but it's one of many factors, these are all points on our scorecard. So as a marketing feature, is this feature going to cause the current community playing the game to be mad, or happy? Does it add value to that community or does it piss them off?
"We have a system where basically we say, does this feature market the game? Will people buy a game if they see a video of this? As a marketing feature, is this feature going to cause the current community playing the game to be mad, or happy?"
And does the dev team want to work on this feature? Are they excited about this feature? Does this feature meet other development goals by developing it? An example of that is, we have the slavery system with the thralls that you knock out and drag back to your base. That gives us 90 percent of the work for a pet system in the game, having pets that you bring back to your base. That would be an example of a feature that helps development of another feature.
So you have these questions, you give them a point scale, we rate them. I fill out those, I talk to the community manager to get the temperature of the community about each feature, I talk to the marketing team to see what they think, I talk to the developers and get their temperature on things. I make judgment calls on a couple of things like: is this a feature we can do well?
For example, climbing was a feature that I believed we could do well, and we did fairly well with it and I think it meets the standards for our game. And then there's stuff like actual estimates, how much time will it take to implement this feature? And then all of those things we put together in a point score, and we look and say, 'Look, this feature is a minus-six, it's probably not a good feature to add to the game. This feature is a plus-seven, it's a no-brainer, we should have it.'
Kris Graft: So it's the exact same questions that you ask at different points in development, to make these decisions about features?
Joel Bylos: The discussions about which actual features to add to the game, that's more of a discussion that I have with the executive producer. Once we have a list of features that we think define the vision of the game, that's when we start asking these questions about them.
For more developer insights, editor roundtables and gameplay commentary, be sure to follow the Gamasutra Twitch channel.