Dead Cells has become a hit since it was released on Aug. 6, but it was relatively successful even before that. It was in Early Access beginning in May 2017 and around one year later, it had sold over 730,000 copies, according to developer Motion Twin's head of marketing Steve Filby.

When most games hit Early Access -- if they decide to, anyway -- there’s usually a lot of the game already there. It’s rough and incomplete, of course, but it’s usually a version of the final product. With Dead Cells, however, Early Access and early testing essentially created 1.0. Before that, it was a completely different game.

Before Dead Cells, Motion Twin was creating free-to-play games for browser and was planning on making a sequel to Die2Nite, a multiplayer zombie survival game. Those who’ve played Dead Cells may notice this isn’t quite like the final product. The game they know is a roguelike, Metroidvania platformer about a prisoner trying to escape from a mysterious prison.

That’s because at some point in development, something had to change.

“We tried it and tested it and we found it wasn't fun,” designer Sébastien Bénard said on a recent Gamasutra stream. “At this point, it was quite a complicated cause because we were not sure what to make out of this game.”

It was thanks to player feedback, both from Early Access and from friends, that Dead Cells as we know it came to be. It’s a single-player game because one of Bénard’s friends enjoyed a single-player version and over time, it became the accepted version. At one point, the game had a points system, but after player feedback, they threw it out.

The previous game was “really like shit,” Bénard explained. “It took one year -- a little bit less than one year -- and we switched it from a totally different [game] to this action platformer.”

Why go Early Access?

There are a lot of reasons why a studio would go through Early Access. With Motion Twins, there was definitely an advantage since it took a lot of player feedback and made major changes. However, according to Bénard, there was one important reason.

“We really needed to get this game out because the company was in a bad situation,” he said.

He noted the “indiepocalypse” -- the moment of time in 2015 when indie developers discussed the over-saturation of games markets like Steam and mobile -- as also being a key factor in the push to Early Access, even if they were discouraged from it.

“Everybody was talking about the indie apocalypse,” he noted. “We also knew that we wanted to have lots of feedback on the game because it's a roguelike, so we know ... we have to test a lot of approach on how the players play the game.”

Okay, so getting the money from Early Access was a priority, but going into the early release gave the developers what they really wanted: feedback. To truly figure out what was working with their audience, the developers wanted to know what was working.

“It was not just about getting money because if we did that, it wouldn’t have gone right,” he added. “We knew we actually had to take care of players and their feedback. We spent a huge amount of time listening to people and actually getting their feedback on the game.”

For example, the aforementioned points system was removed after negative feedback. But what’s impressive was how quickly they got rid of it. Bénard said that the team was happy with the decision until one weekend, when he saw a small group discussing it on the Steam forums.

“It actually worked fine and no one was saying bad things about this system,” he explained. “There was a very small group of people on the Steam forum that were discussing the system and how it was not strategic or not interesting. It was a single thread on the forum, [but] they were on point.”

Bénard said he saw the discussion on the weekend. By Monday, he met with his team and said he wanted to throw it all out. By the next build, the system was gone.

But whose opinion matters?

Player feedback was important to Motion Twin -- especially since that influence exists at the foundation of Dead Cells, but not all feedback is created equal. Being in Early Access for around a year meant that the team got a lot of comments. So how did they know that the small group discussing a points system, for instance, should be prioritized?

What if it’s something only one player wants? If it’s a large group, should their opinion matter more?

“I don't have any kind of mood recipe for this, it's up to you to decide: is it something that will make the game better? Is the change for the good, or is it something one person wants? It's a difficult balance to find,” he explained.

Besides developer instinct, it also comes down to the player experience. Finding the right balance so that you’re making the game you want while also pleasing your target demographic is the ultimate goal, so everything should go towards it.

Balancing difficulty, for example, is a struggle for developers in many genres, but it’s especially important for roguelikes, which tend to pride themselves on a difficulty spike in the beginning that becomes easier to manage over time. That thought resulted in a strict “two-second combat” rule. The player encounters dozens of enemies per level, so combat has to be quick. High-risk weapons can kill enemies faster, but the standard was set from the beginning.

Bénard also noted the care taken for even basic moves. Jumping, for instance, should work. If the player jumps from one platform to another they should be able to land. Unless the jump is meant to be complicated, there shouldn’t be any issues.

"Is it meant to be a challenge? If it's not a challenge, you should succeed,” Bénard expressed. Noticing when those don’t work and listening to players that say they run into those issues is key.

The goal of every interaction needs to be clear. If a player is going to fail at something, even in a roguelike, they need to know why. Motion Twin didn’t want to punish players for certain styles of gameplay, especially when the game wasn’t going to encourage fast or slow players. When that balance was off, and was noticed in Early Access, then it could be fixed. That’s why Dead Cells rewards both fast and slow players.

“If you punish someone for something that he wasn't aware of -- going fast in a level -- it's not something that feels fair because you couldn't know what you had to do before actually doing it,” Bénard explained.

Going into Early Access isn’t for every developer. However, for Motion Twin and Dead Cells, it was essential, not only as a tool to raise money for more development, but to actually make the game. Without it and other player feedback, we’d have another zombie, survival, multiplayer game.

“When we went on Early Access, we were sure that our current version was quite fine. We just planned to just add more content,” he said. “After the first two days, I think, we did have very interesting feedback from people who actually played the game more than we ever had… At some point, we really decided to act [on] the feedback much more than we expected."