When Jon Chey announced the formation of Blue Manchu in 2011, the Irrational Games co-founder said he wanted to spend time on more "genre, nichey games."

The studio's Card Hunter, released last month, is the manifestation of his efforts. It's a free-to-play, browser-based board and collectible card game, complete with dice rolls and a nerdy game master. That's a bit different from narrative-driven first-person shooter games like System Shock 2 and BioShock Chey worked on at Irrational.

Now that a couple of weeks have gone by, Chey tells us some lessons he and his small team has learned from Card Hunter's launch, and what he might've done differently.

So how is the launch going? I saw that there were some server issues -- which might be a good sign, as far as popularity goes.

Online games are weird. You just want to have enough of a player base, and not too much. This is the first time I've done this, and I was aware of the need to get ready, and be ready. And we tried to get ready. But I suspect the first time you do an online game, you always will bump into problems. We did underestimate demand a little bit. It's been kind of a crazy week.

We brought the game live fairly early in my day [Chey is in Australia], because I had to be around and watch what happened, to make sure nothing fell over. I sat there for the first day, and I watched. We had a decent amount of people coming in, and everything went fine.

So I went for dinner and had a celebratory drink and went to bed. I got up the next morning [laughs] and discovered the rest of my team had been up all night, fixing and rebooting the server and reconfiguring the number of threads on the server and reprogramming and re-provisioning a new database.

I felt pretty bad, because they didn't wake me up [laughs]. I'd just been sleeping through the whole thing.

Well at least you got a nice dinner and some sleep.

I felt fine! I haven't had so much sleep since then.

As a developer with a small team, what have you learned about launching an online web game?

I've been thinking about that in the back of my mind this week – how do we do this better next time? I don't actually know what the answer is yet.

We did do quite a bit of testing on this. We spent several months writing a test system. Our game is relatively simple to play, being a turn-based game – there are very few actions a user can take. We actually wrote bots to play the game. I don't think they can actually finish it, but they can certainly sit there and churn away at countless battles. We load-tested with thousands of bots playing the game.



For some reason of another, actual human behavior was somewhat divergent from the bots, and people were putting a lot more load on it. Obviously, there's no real substitute for having real people play your game. So you run a beta. And we did run a beta as well! We ran a six-month beta, but that was a closed beta.

So I guess maybe the lessons is we shouldn't have launched [when we did]. We should've gone into an open beta where we didn't really publicize it quite so much. We did two things at the same time. We opened up registration and we sent out a press release saying we launched, and ran a special offer on in-game items.

We did the traditional game launch, which is like a full-press, trying to get as much attention as possible in a short window of time. We picked a narrow window too, between PAX ending and Grand Theft Auto launching, when we thought press would have time to cover the game.

That's a very traditional game launch. And I guess that's not a good idea if you're a small developer. You can't pre-scale on the expectation that you might have 100,000 people who want to play your game.

Is that the kind of numbers you're getting?

Yeah. The numbers are small if you're comparing to a major commercial launch.

...Well, for you guys, that's a whole lot of players.

Yeah, 100,000 people is a lot for a game that's still running on one server.

So you would suggest staggering the build-up to the release a little bit more.

I guess so, but there's a reason you do this kind of full-press thing of course, which is you only get to launch once, and [game press] is reluctant to cover a game that's in beta, which is fair enough. And they're also reluctant to cover a game that's been out for several months, because that sort of seems like old news. It's a really tricky problem. It's something I'll have to think about.

It's an interesting point though. The way that a lot of consumer press covers games still revolves around embargos, marketing, hype. Everything is meant to lead up to the launch. And what does "launch" even mean anymore?

Yeah, from our point of view, it was just like flicking a switch – turning off the requirement to enter a beta key to get into the game. There's a rather long period where you get enough bugs out, and you could stay in beta, or you could launch. What I'm saying is that maybe we should've stayed in beta, and tested the aspect of the game that hadn't been tested in beta, which is how does it handle with a sudden flood of people.

Free-to-play, that's new to you as a developer too, right?

Yeah. There's a little bit a relationship to the last point, in that the press is still quite reluctant to cover free-to-play browser games. If you're a free-to-play browser game, it's actually a lot harder to get anybody to write about your game – for good reason – there really are a lot that aren't even worth logging into. Once you put yourself in that category, you have to work extra hard to get any kind of coverage. That was again a reason for us to try to build a lot of hype and get an event around the launch. Otherwise, I think a lot of sites might've just passed on the opportunity to look at the game at all. We've had positive coverage, but it's still somewhat sparse.

Nobody says, "Oh, a free-to-play browser game…that sounds like it'll be great!" [laughs]

This seems well-suited for tablets.

We do have a tablet prototype. You just build the Adobe Air version at the same time as the Flash plugin version. We have that up and running on an iPad. It's great, we just need to do all the touch controls on it. I can't see any reason why we won't [release that]. The only reason we haven't yet is because of manpower.

How many people worked on the game?

It's a little hard to count, because it's mostly contractors. There are only three full-time employees. The programming team is basically three people, but they're not full time.

I assume that you guys prototyped with a physical board game?

Yeah, we did. We did a lot of prototyping on this project. We did initial, entirely physical prototyping. Then we used an engine called Vassal, which is developed to enable people to play board games with each other through the internet. It doesn't implement the rules of the game, it just allows you to create pieces and boards and cards whatever, and the players agree upon the rules. So that's a really great tool for protopying. Then we did the actual programming of a prototype.

I really enjoyed doing the physical prototyping. You can't really do a physical prototype of a first-person shooter. [laughs]

What've you been learning as far as free-to-play goes?

I don't think we would've had those numbers if we hadn't been so generous with the free-to-play. And that's a great strategy [giving a lot to players], but you don't want to go too far [laughs]. Because if you make no money, obviously you aren't going to grow your franchise or support your players.

Goodwill goes so far in the business model though.

Yeah, but there also has to be something that people pay for. Sometimes we have people saying, "I love your game, and I really want to support you…but I don't really feel that buying anything gets me that much." [laughs] Whenever you hear that, it seems that you lost a bit of an opportunity. But we're pretty happy, we're making a decent amount of money.