Interview: Ex-Blizzard developer Keith Lee on Duelyst

We talk to Keith Lee about why he left Blizzard to bring Duelyst to life...

Interview: Ex-Blizzard developer Keith Lee on Duelyst

Keith Lee can boast a long and accomplished career in game development, having worked as the lead gameplay programmer for the Ratchet & Clank series at Insomniac Games, and later as lead producer for Diablo III at Blizzard Entertainment. But a compelling question drove him to leave the comfortable world of AAA behind. 

“I was very curious if we could make a new version of some of my favourite games of my childhood, including Final Fantasy Tactics, and Fire Emblem, which is much more modern. But instead of focussing on the single player experience to build a game that was competitive. 

“As I’ve gotten older, I’ve had a little bit less time to play sustained one-or-two-hour-type games. I felt it would be great to build a game for folks that wanted to have their competitive fix, but fit within 5 to 15 minutes. So that’s really the inspiration. We still wanted to make a game that was accessible, allowed you to have a lot of opportunity to build any type of decks and squads, and that you can share and to compete online with anyone else in the world. So that was the genesis for Duelyst.”

There are some superficial similarities to Hearthstone. “It’s a collectable card game. You can assemble a deck out of a large number of cards, that gives you a lot of variety. We have six different factions in our game.” Yet there are two key features that make Duelyst unique. 

“The first one is we introduce positioning into the game. So you can actually move your units on a five by nine battlefield, in contrast to some other collectable card games where they’re more sequenced. Secondly, we focus a little bit less on randomness. Our games tend to be a bit more deterministic, and that allows people to have a slightly higher push on the skill ceiling. I think this caters to people who are much more interested in thinking about a game not just from a card collection perspective, but to also think about how you move your units on the board.” 

In this sense, Duelyst has been compared to the classic Japanese strategy game Go. “Once you introduce movement, and positioning into a grid, the combinatorics around that increases exponentially. You have to think about all the different type of moves that you’re going to do, not just a sequence of the cards that you play.” 

Coupled with this system of high-skill-ceiling tactics is a strong visual aesthetic, wrought by such talented pixel smiths as Glauber Kotaki, the artist for Rogue Legacy. Keith pointed out that this look wasn’t chosen for mere hipster street cred, but for more practical reasons. “We really loved the expressiveness, and the type of animations that you could do for pixel art. Because you could do art that’s very elaborate and very interesting, and we wanted to stylise that with very saturated colours, and to attract people that way. 

“In contrast, if you were doing something much more 3D-oriented, typically in game development you would need to have a very specific rig, or a skeleton, and what happens is that if you have 600 different minions, they will all look very similar. They might be humanoid, they might be bipedal, simply because it’s exorbitantly expensive to do 600 different 3D models and characters. In contrast, we can build any number of pixel art cards. Some could could be flying. They could have four legs, six legs. Stingers. They could look like a scorpion. All of those things can be very expressive, and feel a lot more like a cartoon than a regular 3D model.”

Keith has been surprised by the degree to which these animations have resonated with users. “In our social media for Duelyst, you would notice that the posts that are actually the most popular are the ones based on the animations of the pixel art.”  

The original Kickstarter pitch described Duelyst as a paid-for stand-alone game. Yet as development progressed, Keith and his team realised that the only way they could do the game justice was by adopting a freemium business model. It was a difficult decision to make, but Keith assured us it was the right one. 

“Our vision for Duelyst was to make sure that we can continue to maintain the game. And we’ve had 55 updates and patches in 365 days. That means we have a new update every week. The ability for our team to be able to push out the frequency of the updates required us to continue to have a more consistent income stream rather than have a big burst up front, for a typical paid product. 

“It was really our desire to create a long-term game rather than a bursty game that did well in its first month and then started to decline. We wanted to grow it month by month, and to be able to build continual content, that would allow us to sustain that. It was really about frequency. Sure, people can say: ‘Well, why don’t you do an LCG or DLC system?’ Well, again, it takes three to six months to launch a DLC. And for us, we have much more frequent updates. We release new content every month. Not only that, but we also have expansions. Using a free-to-play model allows us to budget, and allows us to calculate how much content we can produce with the team that we have, and to offer a higher frequency of updates, so that our community would continue to stay interested in our game. And our game has continued to be out there for over a year already, by doing that.” 

From day one Keith was aiming for 99.99% uptime, and to that end the game is built completely from Javascript. “The alpha would actually completely work without plugins on the web. What that did was allow us to learn how people were playing it, how long they were playing the matches, and allow us to tune the game for the first-time user experience so they could jump into the game as quickly as possible, and to allow us to balance the cards and everything else. 

“That would not be possible if we built a game specifically as a binary, as a downloadable client immediately. Javascript allows us to mod our game very quickly if we need to.” The goal has always been accessibility. “As mentioned before, our games average eight minutes. And we have about 250,000 matches that are played per day. And so we had to make sure that people can jump right in, and if they have to go to dinner in 15 minutes they can still play a game, and that it’s predictable for them. We actually put 80% of our time on the early user experience, to make sure that it’s really smooth and quick for people.” 

As for Counterplay’s strategy for updates and expansions, this has largely revolved around monitoring which of the games 12 Generals are used the least, and creating new cards to bolster them. “Also we noticed that there are some deck archetypes that would start to get popular, but then they would start to fizzle out over time, because people would discover new ways to counter those. 

“Our goal is always to have the maximum type of variety in the game. If you play four matches, or ten matches a day, you don’t see the same General, you don’t see the same cards. That every single game that you play feels unique, and very different.” 

In addition to creating a freemium economy that is very generous (compared to, say, Hearthstone) Counterplay has been organising cross-promotions to draw in new players from complementary fandoms. “For example, we might work with streamers and influencers that play a MOBA. A MOBA would be completely different from our game, but the nice thing is that we’re not competing with their audience for the same game. But at the same time, people might actually discover our game and say: ‘Hey! This Duelyst could be a good game because I heard that it’s 5-10 minutes long, and I could play this game when I’m in queue for my MOBA game.’ So it’s a benefit to both parties.” 

For more details visit

Copyright © PC PowerPlay, nextmedia Pty Ltd