Star Citizen: Calling All Devs - Squadron 42

As Jared said, I'm Brian Chambers. When I first started with CIG, my title was Development Director. So it was me and a handful of guys starting up the office here in Germany. And so I spent a few years really focusing on getting that team solid, and working with Erin on global development and looking at teams and seeing what we can help and push. Mid last year, roughly, we restructured a few things within the company with our growth. It just made sense to look at that and see if the current structure we have still works for everybody. Part of that structure was to pull in studio managers for each office so they could work on the day-to-day stuff that it takes to keep the lights on, and general admin stuff and everything, right? That allowed me to then come and focus more on dev. So at that point, I became VP of Development, and I now work really, really close with Erin on a daily basis to drive the development team globally. Come November, December, me and Chris and Erin were talking and looking across the projects and seeing how things were structured, and we said, okay, with Squadron, the size it is, and what we want it to be as far as expectations, it absolutely needed some more dedicated focus. So I am now primarily focusing on Squadron 42. Erin's now primarily focused on the PU. We both have visibility and say on either side. A lot of stuff crosses the boundaries per se, right? So one call actually works for both. So my primary focus now is Squadron, working with the team every day, seeing what's going on with schedules and with progress, and it's going to be good.

And this kind of reorganization is a normal part of video game development. It happens in projects all over the place. It's not necessarily because of issues or problems. We usually do it at the end of one year, the beginning of another year, to take a step back, to reevaluate how things are going and to find ways to improve, to make things more efficient.

No, it wasn't due to any deficiency. It was just that now we have a studio manager there and they can handle all this other stuff that I needed to take care of, and that now frees up my time to go back to my roots, which is game development. And if I'm free and Erin's free, let's split that focus so we can get some more of that dedicated attention. It's definitely a positive thing. And I'm excited about it. Working with the team, you see progress every day. It's good stuff.

No, the roadmap, honestly, it's one thing that I've looked at more and more recently, and other people on the team, and we were scratching our heads too, going, wait, hold on, what's going on here? The roadmap as it currently exists, starting chapters at the top, going to the bottom, and just spreading out horizontally, it doesn't make sense. There's so many things that we do; a majority of what we do crosses all chapters, right? When you talk about mechanics, when you talk about technical work and so on, when we're working on characters, whatever it might be across most disciplines, it touches all chapters. So in theory, if you were to leave it as it is, all chapters would basically be done all at the same time, right? We've got every little thing, excluding a couple little bits that are going to be very specific and very unique. So due to the nature of the development and what you push on, if we call it AAA or quadruple A or whatever it might be, it really doesn't work. I think when we first started it, it was just logical. It was like, oh, let's put all the chapters down and we're cool.

Follow a linear progression. It's easy to understand. It's not like we had an example to look at. There aren't other AAA games out there with a publicly facing development roadmap during their development. So we had to figure this out as we went. We tried for a linear presentation, but as you're saying, development runs parallel.

Yeah, and I actually scrubbed through stuff, maybe a couple of months ago, looking, okay, what other AAA games are giving the level of detail that we do? And I think some had some roadmaps, but they're just very chunky. Like, hey, we're a quarter of the way done or halfway done and so on, right? A big key thing for us, though, is we don't want to give away spoilers. It's incredibly important to me and it's incredibly important to Chris that we hold that stuff tight, because we want that experience when we get it into people's hands, that they play it and they experience it new. I'd rather not give away, although we may have already most of them, I'd rather not give away any more characters, locations. And it's actually hard when we do monthly updates and so on. A lot of stuff we get from the team that I'd prefer to publish, that is awesome progress, has so many spoilers in it. I'm like, no, can't do that. Can't do that. Nope, nope, nope. So it's a chore. However, the current format doesn't work to show the progress the way that we think is the best, or in a format that makes sense for what we're doing. With that said, we're currently in the process of investigating new ways that we can provide information. Progress is being done every single day. Awesome stuff. We have global reviews with the directors and leads and Chris Roberts every single week where we go over and see people's progress. So progress, really cool stuff, is being done. So it's up to us to find a way that we can present that to everybody publicly. And that's something that we're actively looking into. Absolutely.

Right, so it's not all doom and gloom. We recognize that the roadmap for Squadron 42 isn't as effective a tool as we were hoping in conveying the information without spoilers, mostly due to its linear nature and our parallel development, not just between all the chapters, but our parallel development between Squadron and Star Citizen. It's very hard to present that to backers in graph form and lines without the context that we can't give, because the context is all spoiler-related. But we do have a couple ideas, and we're still looking for more, about how we can be more open about Squadron 42's development, how we can convey more information on its progress.

So we've been thinking a lot about the roadmaps lately, and ultimately we want to improve how we communicate in a variety of ways so we can basically provide a better idea of what's being worked on, why something may have changed, et cetera. So more specifically, we're looking at some creative ways that we think we can better illustrate our progress that ties closer into how we're blocking out our time internally. We're already well underway with exploring some creative options for how we can visualize that and how we can share that with everybody. But of course, it's going to take a little bit of time. But I think we're heading in the right direction. We think we're working on some stuff that you guys are going to like.

As far as blocking out time, some of that stuff we already do on the dev side as well. We block out sprints that guys are working on, and you can see their completion progress. So we even talked about, could that be existing information we have that pulls in, again, as long as it doesn't give away spoilers and so on.

Yeah, absolutely. We also recognize that as we're developing Squadron 42, much of that content ends up making its way into the PU, which is something that we feel we can track and also illustrate better. And so that's another thing that we're working on. But in the short term, obviously you can still look forward to the monthly reports. You may have seen recently we did an AMA on prisons. So we would like to continue doing those AMAs more regularly and actually talk Squadron 42 when we can. Like Brian mentioned, we don't want to spoil the game, but as often as we can, when it makes sense, if we can do some AMAs surrounding different aspects of Squadron 42, we'll take those opportunities. One of the favorite parts of Inside Star Citizen is the Sprint Report. We're going to start leveraging the Sprint Report for Squadron 42 content again as often as we can. You probably see the pattern here: we just want to start injecting more Squadron 42 content into our regular communications, but also getting creative and seeing if we can introduce some new stuff. For example, we might do some fun regular video check-ins with Brian, with you, and the Squadron 42 team. We've been talking about some fun ideas. I think that you guys are going to like the direction we're going to go with that.

No, we're exploring everything. We're looking at it from all angles. We've even had some discussion, as we've said, and everybody knows that a lot of the things that we do from Squadron make it to PU, or PU to Squadron. So could we even have some type of shared map that calls out which ones are shared, which are unique and so on? Because a lot of stuff you see in PU, although we're not loud about it, absolutely makes itself into Squadron, right? So by nature, it already starts showing some of that progress there. It would just be us calling it out.

And then of course on the video side, opening up shows like Calling All Devs to host Squadron questions when appropriate. Maybe even an Inside Star Citizen segment here and there related to Squadron. Again, where we can find stories to tell that don't spoil the story being told. Basically, what it sounds like is we're going to pull back the curtain, we're going to push open the closed door as much as possible. It's going to be slow at first. We've got to be careful not to spoil anything.

Yeah, I'd say the key takeaway is that we recognize that we haven't been able to share progress as effectively as we would have liked to with the way that we've been set up. And so we're in an area now where we're willing to experiment, we're willing to tweak things and see what we can do to ultimately give more information. And I think one of the big parts that I'd like to note is that we're going to be looking at Spectrum for y'all's feedback. We'll start a few feedback threads. And we want to hear what you think of the direction that we're taking it. We want to hear your suggestions, and we want to keep that conversation open so that ultimately we can give you guys as much information as we can.

I'm confident that we'll find a way that's informative and gives people a good idea of what we're doing on a regular basis.

Like you said, both Squadron 42 and Star Citizen are completely developed in tandem. So what you're seeing in the flight model right now is what we have in Squadron. What you're seeing in the AI and the FPS side is what we have in Squadron. Again, that's something that we've actually seen as a big benefit for Squadron, because we can refine that experience with the help of the backers. We can take the feedback of what people like and don't like and put it into Squadron, and make sure that we're closer to the bullseye when the game comes out, without spoiling anything, which is cool. We're not exactly where we want to be right now, but we're pretty close. We're getting closer by the day.

I think it's fair to say that many people will share the sentiment that it is not where we want it to be. But as we review, as we re-examine at the top of every year, find new ways to improve stuff, we've done things like create the Vehicle Experience team to help refine the flight model. We're going to talk to Francesco Roccucci about advances in AI that we've made a little bit later. I think we're really poised to make some terrific progress this year.

Yeah, absolutely. We've had some really good hires, some new additions to the company that really made a difference in those areas.

That is obviously one of the pillars of Squadron 42. It's very cinematically driven, and that's how Chris has wanted it to be. We've got the addition of these heavily narrative scenes that drive the story forward. We're shooting for a level of immersion that I guess you weren't seeing in most games previously. You're not just a disembodied ghost spectator on conversations. When you go there, you can break out, you can abandon, you can rejoin conversations, they'll time out. You actually feel like you're part of the scene completely. And that's quite engaging. Obviously, rudeness will carry a negative marker, so if I wander off in the middle of a conversation with you... but we have that even filter into the story as well, so if you upset someone, then later on they might be rude to you.

Sounds very complex.

It is, yeah. But overall, it makes you feel part of the crew. You feel like you're there, which is good.

I guess, if I'm looking for a non-spoiler, which is always quite difficult on Squadron, we've recently been working on the new female player PCAP. As with everything in Star Citizen and Squadron 42, it's more than just a head and voice replacement. We're integrating a full performance into each scene to enhance the experience of a gender choice in the game. So if you choose the female character, you will have a different performance within the game. Chris has been vocal in the past about almost presenting the cinematic and story elements of the game as a director's cut kind of thing. But we've always said we're going to allow the player to wiggle out of that and experience it the way they want to. You will see the nuances of a female performance within the same scenes if you choose that gender now in the game.

And with all the subtle differences in the storytelling and the performance of other characters that that might entail.

Yes. You've got a lot of slightly different interactions as well in there. So it's going to be interesting. We shot quite a lot of scenes. So it's a lot of work, but it makes a big difference.

And I suppose it adds to the replay value, too, because you can go through playing with one gender and then go through and play with the other gender and see the differences. And of course, the recent filming we just did was not the first time. They were pickups. It was additional filming with our female player character.

Yeah, absolutely. And we know in the future, obviously, we are going to be doing pickups here and there for a lot of the story elements as they develop.

So, as you guys probably already know, AI is exactly the same between PU and Squadron, in the sense that we use the same system, we tend to use the same type of behavior structure, we tend to use the same core functionalities. So right now our major focus is split between our three subsystems, sub-teams: we have social, ship, and combat. The social team is mostly focused on improving the stability of our usable system, and how we can improve our pipeline to allow designers to construct and build new behaviors, new content very quickly for Squadron. For example, creating scenes that are required for the story in a much more efficient way compared to what we were doing before, allowing multiple disciplines to work at the same time, like cinematics with level design and so on. Ship has been, in the past year, one major point of improvement for us. We've been working very closely with the physics guys to get to the new 3D pathfinding functionality that we just released in 3.8, and we are going to release new updates in 3.9. 3D pathfinding basically doesn't use any of the classic approach to AI pathfinding. We didn't want AI to have the requirement of building new knowledge about the system, evaluating physics, creating our own graph, and then parsing it to find a point A to point B path. What we wanted to do was to use as much as possible the information we already have, even to optimize performance. So we use the SDF from physics, so the distance field calculation, to understand which objects we have near us, and then we start from the assumption that we can fly anywhere, and then we just need to find the best approach to reach that location. And then we fly around in our close environment, avoiding obstacles, avoiding terrain, avoiding planets, avoiding asteroids. And this is also very important for collision avoidance, to predict where ships are going to move and so on. So this is a major improvement that we released, and it's going to be part of Squadron. Also, for example, to have cases where a ship is leading another ship, or a ship is leading the player, or the player is leading the AI. Then we have FPS combat, which has had its main focus on the refactor of the combat behavior. What we wanted to achieve was a behavior that was very modular. So we structure the behavior so that we have different tactics that can be performed, and those tactics are also able to handle cases where we want to, I don't know, reload because we don't have enough bullets for the next tactic we want to use, or we have different weapons. If we have a shotgun, we want to use specific approach techniques, but if we have a rifle, we want to do something different. Besides that, the team has been focusing a lot on improving the cover generation system, so now we have not only cover locations, but also what we call gray covers, which are covers that are good if nothing better is around, pretty much. So we can use basically any geometry of the environment to progress, letting the AI get close to the player. And then we are also improving the perception system. We want to use more information from the room system to propagate audio around and make sure that the AI reacts correctly to the player.

So basically, what we're doing with the bartender, and why we are working for so long on the bartender, is because the bartender is one of these classic examples that encapsulates so many different problems that are used in different behaviors. So first of all, right now, the bartender is actually just a vendor. So we have vendors in Squadron 42 and we have vendors in the PU, and there will be no difference in behavior except their attitude, their skill and their personality, right? So some bartenders might ask you what you want, some vendors actually might not ask you what you want, and they're just going to wait for you to interact with them. Or some vendors might take back the object that you asked to look at before they show you another one. Some won't. So the bartender has this very complex environment where we have a super realistic bar, right, where you can create cocktails, you can pick up objects from pre-existing locations. Let's say you have a fridge with beer bottles, right? Those beer bottles already exist. You just want to pick them up. But if you are creating a cocktail, well, you have the original sources for the type of drink, right? You have maybe a cola, you have whiskey, and you want to mix them up and you want to put a slice of lemon on top of the glass, right? And this is what we are working on. Our system right now allows all of those. That means that designers will be able to create a lot of content very easily starting from that template, not just for vendors, of course, but vendors will already be a major part of the PU releases, and there will be characters in Squadron providing you weapons or providing you other stuff. So this is why it's taking so long, and it's also taking resources from code and design to work tightly close to each other so that we can get the best out of those systems.

Well, it's very easy to see the benefits of ship AI and FPS combat AI in the combat aspects. The bartender is representative of the social AI aspects. Anybody who's played the old Wing Commander games remembers going into the mess and talking with Shotglass and getting the gossip of the day. And you don't just want them to recite things, you want complex actions and them to feel like a living, breathing person. And the bartender is our gold standard, if you will, for all of our social interactions that happen with every vendor throughout Squadron 42.

Absolutely. Vendors are the central place in every role-playing game, sci-fi or fantasy; the center of the world.

I wanted to do something that pushed the envelope a little bit more as a closeout for the year, so everyone can go home for the Christmas holidays and be motivated about coming back in the new year. It's been a tough one. And so I went through the year of data, the year of visuals, collated all that information and pretty much cut it all together into one edited video to show to the entire company, internally only at that point. I think the original video was just over 10 minutes long. There was an awful lot of content; it had characters in it, it had lots more cinematics and everything. And then of course, it's funny, because I showed it and everyone was quite silent in the meeting. I was like, oh, there we are then. And then I get the email from Chris: "Hey, man, this is really cool. We should totally do something with this. We should show our backers." And then, long story short, we just basically went through all of that footage and removed anything that was too spoilery. Because obviously, with it being a narrative-driven game, it's very hard to show content to the outside world without spoiling certain characters, without spoiling certain story moments.

That's been a theme of this episode so far, the spoilers.

Yeah, I mean, everything's quite sensitive. There's not really much that goes on in the story that wouldn't give stuff away, if that makes sense. So, for example, if we just showed you one character, you'd just go, okay, I get it, I know who he is in the script. Want to avoid that. So that's how we ended up with that, basically like a show and tell. It's certainly not finished, it's not polished. I'd say visual quality wise, probably about 40% in that video. There's still a lot to do visually. But I think it's coming together. And it was really nice actually seeing the video where it shows how diverse the locations are; the game's not going to suffer from visual fatigue in the slightest. It's one of the key things I wanted to avoid.

So visually, we're pretty much on a curve up. Game dev, if I just draw a curve, it's kind of like this for a long time, and at the end of the process, it goes straight up to the top. So it's not a linear curve, but I think we're just approaching that crest now. But we've come in in the new year. It's always nice to get that break and then just really resurvey how the team's operating, how the teams are split up and what they're focusing on. We have our internal reviews as well. So it's a good time for us all to speak about the project, get a lot of stuff out in the open and just really work on that feedback as a team. But as for the game as a whole, it's genuinely really beginning to come together now. We had Chris here the week before last, and it was really refreshing to be able to sit down for what was a six-hour review, and at the end of it, the game, okay, it crashed once. That's sort of unheard of at this point in dev. And pretty much everything was beginning to work. One of the most technical aspects of the game was running, and people were doing their things, NPCs were doing their things. You could speak to people, have conversations with people. You're in the Idris floating around the system and looking outside at the epic view. And when you see these little nuggets coming in, it's really helpful and really motivational. So visually, it's just continuing to grow on that momentum now. Obviously, we've got planetside locations in development, we've got stations in development, ships, other ships from other races we're trying to get into, all sorts, a bit of everything. And part of the review process is also always really trying to make the most of people's passion. So I'd never want someone on the team to be working on something that they don't want to work on, if that makes sense. You're not going to get much out of them. So that's generally what's been happening. And of course, dare I say it, there's a lot of cool stuff being put together for a certain show at the end of the year.

What I've normally got on my screen is I'm hunting for faults. I'm jumping in and finding things that we can improve on or anything like that. I like to lead by example, so I try and stay completely in touch. I've got one of our stations in front of me that I'm going through. It's not 100% finished yet, there's like two areas which I can actually show you, so you can see what this sort of looks like mid-dev.

We're not usually a visual show, we're usually a talking head show, but let's end on a high note here. What can you show us? And I haven't seen this. I don't know what we're about to see here.

No, I'm just in here balancing the lighting at the moment. So this is Eido Station. You've seen this briefly in the... I'll just pull you out and go through some of the setup of this stuff. So it's one of our comms arrays, set up by Arsiedo. You'll have to excuse some of my shadow issues. We've got static shadow region issues at the moment. And obviously you'll notice this is in completely black space, because this is an object container setup, and these object containers get stuck into the larger system later. We don't alter them. But without further ado, just jump in, I can show you the setup of how this thing lives and breathes. It is very much similar to what we would do on a ship. It has its own power systems, it has its own comms, it has everything that it needs to function correctly. On entering this space, you'll probably notice it's not a sterile station. There are obviously story elements that are going on here. And the main story element for this station in particular is that it's in the process of being decommissioned and ultimately recycled. So this is the comms center room. You'll notice the dark side is generally the bit that's being decommissioned. Everything's starting to get covered in cellophane and cloth and whatnot. But if you come through the other side, you'll see that this side is still operational and is functioning as you'd expect it. So yeah, I'm sorry if I've made people dizzy, because I'm one of these people that just looks around at everything as I'm going around.

This is, of course, just one room of a fully explorable station.