⋮    ⋮  

The Star Citizen Exclusive Interview: Erin Roberts

Author Photo
Feb 17, 2017
Share Tweet Submit

PC Gaming is an emotive subject, whether it’s fanboy wars between AMD, Intel and NVIDIA, what brand of accessories/components you prefer, what games/genres you like best. Perhaps one of the more emotive subjects in recent years for the gaming community has been that of Star Citizen.

Pitched in late 2012 as a PC exclusive and a return to the long underserved space-sim genre with no publisher oversight, it’s fair to say that the crowdfunding campaign captured the imaginations of a large number of gamers with over 1.7m unique accounts (note, this does not necessarily equal backers of the game since people can have more than one account if they wish) and over $142.7m raised in the last 4 and a bit years.

In that time, scope has grown massively and as the funding dollars have rolled in, the game has spawned a huge following with both supporters and some critics. Star Citizen is also an emotive subject for me as it’s what brought me back to PC gaming in 2012 after my decade and a half in the console hinterlands of occasional and casual gaming at most.

So yes, full disclosure, I am a backer of Star Citizen. I loved playing Chris Roberts’ Wing Commander games as a kid and over the last 20 odd years I’ve repeatedly AltaVista’d, AskJeeves’d, Yahoo’d and of course Googled semi-regularly for “new wing commander” until in late 2012 the search returned the Star Citizen Kickstarter. I watched the video introduction and immediately backed. Since then, following the game has been a wild ride. Seeing a big game studio business be built and the game I’ve been waiting for a long time coming together piece by piece has been great fun. There have also of course been those who think the game is impossible to finish, scope has grown too much and the entire thing is a scam.

Personally, I think there are easier ways to go about running a scam, development hasn’t been the fastest but at the same time when you consider that the team has been putting out regular patch updates and increasing functionality all the time, as well as trying to effectively build a studio from scratch, I don’t begrudge the pace of development.

That much being said, the last year or so feels like the company as a whole has largely been up and running. Development continues, however there is some risk that the community fades away after years of waiting. CIG does a good job of putting out regular update posts and video shows to keep the community engaged and by the visible metrics (money raised annually, new backer accounts annually), there is a steady stream of new backers and money coming into the game, so the risk would appear to be small.

I recently had the opportunity to go on a studio tour to Star Citizen’s largest development studio (a little over 200 staff) in Manchester, UK and also got some time to sit down with Erin Roberts (Chris Roberts’ brother) who runs Foundry 42 (the European development arm of Cloud Imperium Games) and put some questions to him. The tour itself and some of the conversation topics were covered by a non-disclosure agreement so there is a reasonable amount I saw and learnt which I can’t write about (yet). But the interview with Erin was still an interesting one and there is a reasonable amount I can say.

The Interview

Wccftech: Hi Erin, thanks for taking the time to chat to us. I wanted to start with something that is perhaps slightly peripheral to the game itself and discuss Brexit. Obviously we had a couple of emails last year about the topic in advance of the vote and there was mention of a possible risk to the UK videogame tax breaks. Given that we now know the outcome and the UK will be leaving the EU, is there any immediate impact to you guys?

Erin Roberts: As a company, we’d probably rather that the UK didn’t exit the European Union as a lot of our staff are from all over the EU, we’ve got some great people here and there is uncertainty as well as potentially making it harder for us to hire the people we want.

That much said, in terms of the immediate impact to the company, it’s actually been somewhat surprisingly positive because of the change in exchange rates and the fact that we as a company do the majority of our development in the UK, but of course the majority of our revenue is in US dollars, it actually meant that given the Sterling crash, UK development has effectively become cheaper.

Additional to that is the tax credits situation from the UK government, that law is likely to change due to the way it was implemented as a European-wide law, but right now at least the UK government is promising that will stay intact.

So short to medium term looks ok, obviously we have a 2 year period while the UK negotiates its exit from the EU, longer term is of course still unpredictable until we get the detail on how the exit works. Close to perhaps 25% of our staff are from the EU and given that we support a global community, the language skills they come with are also very beneficial to us.

W: The big question is obviously 3.0. There were a lot of appetising bullet points on slides discussed about 3.0 last year at GamesCom and it feels like 3.0 is in many ways the next big thing with the gaming really starting to come together in terms of new gameplay mechanics and a significantly larger number of interest points in the persistent universe. What is the situation there?

ER: We’re working hard on 3.0 right now and you’ve hopefully seen some great stuff we’re putting into it on the tour. We’re still at the start of the year and there is a lot of scheduling work going on so I’m obviously not giving dates today and of course we want to give the juicy information to the community first, but we’re looking at putting out perhaps 2 or 3 big releases this year which significantly push the amount of locations, gameplay mechanics and content that the players will be able to experience and give feedback on.

W: A while ago, Tony Zurovec put out some design posts regarding careers and envisaged gameplay mechanics. They were all quite high level at the time because obviously no systems had been built out and they were more conceptual Obviously as we approach more actual professions being available to backers over the course of the year, presumably there needs to be quite a bit more detail given to the community in terms of how are we going to be doing these things and what are the mechanics involved, is that something that’s coming soon?

ER: Definitely, there are a lot of systems being put into place which will basically support everything the player will do. I can’t remember the name we have for it internally right now but there is a whole system for players etc communicating while in space with regards to jobs. So let’s say you want to send out a distress call. You can send that to just your friends or anybody, some people may come and help you, some may try and kill you, who knows. That entire system is basically the same system in which if you’re out in your ship and you run out of fuel, you’d put out a tender to say “Hey, I’m at this location, please bring me fuel and I’ll pay you X”. This is all the system we’re building out now and it’s part of the whole new mission framework system. Let’s say you’re quantum travelling and you get interdicted by pirates and you communicate for help, it could be players or NPC’s and it’s basically a jobbing system.

The overall system is the same basic one whether it’s for distress, asking people to do something, asking someone to join you and so forth. That system effectively gives a huge capacity for players to get together and do stuff in game. We need that system along with interdiction of course if you want to do cargo runs. There is going to be the “buy 500 tons of this” and we’ll have to balance that and pricing and I’m sure it will be difficult to balance, but if you can just buy stuff, fly somewhere and then get your money, well there’s no risk to that and that’s no fun, there should be a well-balanced risk/reward.

Star Citizen Anniversary Livestream Showcases Star Marine, Capital Ships

When we put in mining, same thing, you’re mining away and then get attacked, distress or need someone to come and haul stuff you’ve left nearby away etc and you’ll pay them X. These are all jobbing type systems and of course there will be mechanics for mining and everything else, but these systems are much simpler to do than all the core tech stuff we’ve been working on like planets, making the game 64 bit etc. The main things we’ve got left now, the planetary stuff is there although we’re continuing to refine it, the AI in terms of both the subsumption walkabout AI as well as the combat AI, although they work it’s not to the level that we’re happy with yet and also then the object container stuff.

Object containers, all they do is say “something is here” and then as you get closer to them, they say “OK, you’re getting closer, now I need to load” and it could be layers like just external stuff you can see initially and then internal stuff. Right now what we’re working on is the intelligence for when things stream in which is what we need for the next big release. Once we have that, core tech is basically done and then it’s about layering on a bunch of very cool gameplay.

So with the next big release a lot of the underlying game is there and then we can look at transferring people between servers so we can have hundreds of thousands of people maybe in one instance, but that doesn’t come online until later.

Planets are great but we need to generate the content there for the players to consume, like the homestead mission you saw, somebody perhaps takes a rescue mission to a planet and maybe it’s a trap and you’re ambushed, or maybe another time it’s just a simple rescue mission, or we have ground vehicle racing on planets etc. We also don’t want to have you coming to planet where you just use your ship and shoot everything easily, the balance should make it so that you have to physically land your ship and come out on the ground whether there are air defences or something else to force you down.

W: One of the big issues that players talk about are PU framerates. There’s been a lot of talk about network code refactoring, regional servers coming online with 2.6.1, number of player limitations etc, what’s happening here?

ER: Regional servers doesn’t tie into this so much and of course is about lag. If you look at Star Marine for example, frame rate is fine and there is no reason why we will not have the PU to that kind of frame rate. The reason is because in Star Marine, we only load in what we need.

In the PU there are several things going on. Some of the mission givers are obviously run by AI and that AI isn’t optimised at all and a lot of what we’re working through now is making stuff a lot faster in the AI.

Secondly, we’re basically loading a huge amount into memory so there’s no real streaming in. Lumberyard will stream textures but the reason we have these object containers, right now you can turn the object containers on or off, but when you’re in the PU, you can be over here and your friend can be 100,000 miles away and some stuff which happens at one location is affecting the frame rate at the other. There is some extremely basic culling right now but we can do more in this space. We’re not massively render bound right now, the PU is much more CPU bound and we have all these draw calls we’re still drawing even though we don’t need to.

The next big release improves things significantly because of the new object containers which turn stuff on and off, not just turn it off but everything else which is going on in there so people should see a dramatic improvement in performance, not necessarily the finished article, there are still other things to come but this should make things a lot better.

W: Quite a while ago, we were shown very early stage art for the Pegasus Escort Carrier. Since then we’ve not heard much about it, is this still being worked on?

ER: It will still be a ship we use but it will appear in the 2nd instalment of Squadron 42. In the first part of the trilogy you’ll still see some really good ones, the Idris obviously, as well as the Bengal and Javelin, plus one or two others we haven’t really talked about yet.

I personally can’t wait to see this one in game…

W: Economy/AI then, a concern for the community is that players want to feel like they’re progressing and that the game is a challenge, but don’t necessarily want to have to play for a year to load torpedoes on their Polaris or something. How do you combat inflation, giving players a sense of progression, gold farmers, mission giving etc so that one month after the game launches, not everyone has every single ship under the sun and there is a sense of if I pull the trigger on this there is a real cost to me in doing so?

ER: It’s really about the way the game is intended. There will be a lot of balance iterations on the economy side but if you really want one of the bigger ships, you’re kind of going to need a team of people. Like on the Enterprise, you’ve got Kirk and Spock etc. Running some of the larger ships will cost significant resources and you’ll need people to contribute to that. Missiles will of course also cost. Right now, if your ship blows up, so what, in the future though it will carry real consequences. People are really going to need to work together to figure this stuff out and fund it.

W: So this touches on something that gets asked a lot in terms of multi-crew mechanics. Pilots are obviously taken care of, same with gunners, both are quite interesting jobs. Everyone else, the concern is that they’re basically going to be playing a bunch of whack-a-mole minigames and then the rest of the time just sitting around doing not a lot. How do you keep things interesting?

ER: So let’s say you’re on an Idris, being the engineer I think is probably the most interesting job and it’s their job to balance the entire ship and keep everything running. It’s like Scotty in Star Trek. When the ship is taking damage and we have the whole power node system (W: I saw this earlier on the tour, very cool), power or systems go down and people have to run around the ship and go to locations to reroute power etc. Sure there may be some guys who are marines and waiting to be boarded but the number of different things there are to control. I suspect there are going to be some people that when we first release the Idris are going to be spending weeks running training scenarios about what happens if this happens and should we put more power to the main railgun but then maybe that takes the shields down for 30 seconds, ok so what can we do about that etc.

Star Citizen Switches Engines: Move Along Folks, Nothing to See Here

Some people will of course say “screw this, just give me a gun and I’ll be a marine” but there are going to be a lot of interesting things to do to keep players involved.

W: There have been a lot of questions about APIs and what kind of tools the community will be able to create that hook into the game. Looking at the moment we have the ARK StarMap but if you want to create anything out of that, it’s obviously a manual combing through it process. Similarly, if my Org has say one member in system 1 and another member in system 2, maybe I’d like to have some kind of commodity ticker which tells me the prices of goods there etc.

ER: There are obviously APIs and these kinds of scenarios are definitely the kind of thing we want the community to be able to do. In terms of priorities, it’s obviously not the highest at the moment so I don’t think we’ll look at exposing any of this in the next year or so.

This is all great stuff but at the same time it’s the biggest game I’ve ever tried to build in my life. It’s more like 4 or 5 games in one we’re building here in terms of the technology, the company, lots of moving parts etc.

For me personally it feels like the last year maybe is when we’ve actually bedded down and the maturity of the team is there that we can really get better at predicting what we’re producing etc as we get rid of the big technical hurdles.

Our ship pipeline is awesome. I can tell you right now with a good degree of accuracy, how long a ship will take us to do. 2 years ago we were still trying to define the style guides and what they were and the brands etc. Now we can say “it’s a small ship aimed at doing xyz, this will take 12 weeks, 16 weeks etc” and be reasonably confident in it.

W: So are you guys working in an agile/scrum based workflow? Earlier you mentioned about huge planning exercises for the year etc?

ER: We do work in 2 week sprints and have a reasonable forecast for stuff that is 1, 2 or 3 months out. The planning is more around what are the main goals and prioritisations, of course stuff which blocks may mean reprioritisation at given points in time.

W: So at some point, when/who says (and we all really love Chris obviously) to Chris “please stop giving dates to the community!”?

ER: To be fair, Chris isn’t really giving hard deadlines, he says we’d like to do XYZ and we would, we’d all love to give everything as soon as we can, but this is something that gets brought up repeatedly and this was why at the end of last year we really started opening up with our production schedule to the community.

I’ve done this all my life though and it’s often, you know developers saying “oh yeah, we can do that!”, it’s not Chris making up dates, people want to give their best but then a roadblock appears, some contingent work isn’t finished in time or sometimes priorities change. It’s like last year when we delayed Star Marine and everyone reported that we’d killed it (W: We didn’t! In fact we reported it was still coming!) but we needed to reprioritise for other things.

So now sure, now there are often frank conversations saying “OK, I know you think you can get X out in 2 days, but what happens if all of a sudden you need to fix bugs for a day or some blocker happens or how sure are you that the other team delivers you what you need when you need it?” etc.

As we mature as a company, things continue to improve. When I was doing the Lego games for 10 years, it was so easy. There was no tech change. It was literally, ok this time we’re going to do Lego Batman, so reskin it all and put in 2 or 3 new gameplay features, bang, we could knock those out every 4 to 6 months. This is an order of magnitude more complex.

W: Thanks very much for your time Erin, it’s been great to see the studio and meet you.

Wrapping Up: Where Next for Star Citizen?

The studio tour itself was interesting but unfortunately I can’t really report on any of the detail of it yet since it was covered by NDA. What I would say is that there is a lot of very cool stuff coming in the pipeline, so whether you’re looking forward to capital ships, alien ships, or human ships, new FPS modes, PU improvements or new gameplay mechanics, there really is something for everyone.

What I would comment on is the sense of pride that is evident at Foundry 42 and which its people take in showing you what they’re working on. The level of detail in what is being developed is amazing, certainly beyond any game I’ve ever seen and the people there are truly passionate and solving problems every day.

Personally, I do feel that CIG have been slightly naïve in some of the ways they have done things over the last year. It sounded originally like there would be less events than they have traditionally done and I think it made sense, particularly when GamesCom went from originally being a stand with pop-up social events to a full blown presentation event, that left little meat to be released at the end of the year and some subsequent criticism to the end of year stream.

When you look at the project in its early stages, there were lots of disparate strands to demo and release (hangar module, dogfighting module, PU, FPS module, PU persistence etc) but as the last major “module” to be released (Star Marine, the FPS mode) was released last year, there really aren’t the kinds of levels of work being produced which allow for such huge event reveals anymore. There is iterative work taking place, primarily in the PU and of course there is also Squadron 42 (the standalone single player campaign) to come, but Squadron 42 has limited reveal potential without giving spoilers to what is primarily a storyline driven campaign.

Am I still a believer in Star Citizen? Absolutely. The game still has problems to overcome but given what has been achieved thus far, there is little doubt that they can be resolved assuming funding doesn’t dry up. The company feels like it’s in the right place to be able to do what it needs to do and deliver the final push towards launch (I suspect sometime in the next 18-24 months). It does feel like perhaps there could be better understanding of some of the principles of agile/scrum based development and forecasting at some levels but I remind myself that in my 18 years working in financial services, I’ve seen a lot of game developers/project managers etc make the jump to finance but never seen the reverse be done. When those people come, on a fairly regular basis, significant training is needed on “professional” development principles. Spending a fortune on agile training is likely not a high priority in the game industry (even where it occurs in the financial services sector, a lot of people still don’t “get” it) and as long as the developers continue to put out good content, overcome their blockers and deliver as a backer, I’m happy.

It’s been an interesting glimpse into the development of one of the big games currently in development. As ever, I’ll be following the coming stages with great interest.

Share Tweet Submit