“Ugh, I want a rematch now,” my character mutters as he emerges from a mystical pool of blood. He’s just arrived back at the lowest chamber of the underworld. Again. I couldn’t agree with him more; I had Theseus and the Minotaur on the ropes, until I dodged out of the way of the Minotaur’s axe — straight into Theseus’ spear. But once I died and returned to hell, I could indulge in one of the best parts of Hades: talking to each character, getting their reactions to my most recent defeat, and learning about them in turn.

It didn’t feel like my attempt to escape the underworld had been wasted effort. On the contrary, it felt like an opportunity to discover more about this enticing world and the well-written characters within it.

You May Also Like:

This is the loop in Hades, Supergiant Games’ rogue-lite dungeon crawler and the company’s first Early Access project. It’s just the latest example of how the studio is changing the concept of defeat with its games — making it an integral and valuable part of the player experience instead of a time-consuming punishment. As a longtime fan, I was excited to get in touch with Greg Kasavin, lead writer and designer at Supergiant, to discuss this approach to development. I wanted to know how Supergiant removes the “game over” from its games without compromising challenge.

“We’ve always been fascinated by the seemingly inevitable moment of player failure,” Kasavin said. “I think part of our interest in this comes from the simple observation that we often learn from failure, both while playing games and across many other aspects of our lives.”

In video games, Kasavin explained, that moment of failure is vital. A player that dies is often excited to dive right back in and give it another shot. Other times, they might want to quit the game and never play it again. Knowing the difference between those moments is significant. Supergiant has always felt a need to give it careful consideration as part of each game the studio has worked on — “to do everything we can to weave it into the experience, and make it feel connected to our games’ broader themes,” according to Kasavin.

Game Over… Or Not

But what informed the idea of removing “game over” from Hades altogether? The studio’s first two titles, Bastion and Transistor, were both isometric action-RPGs with more traditional fail states. I asked Kasavin how and why this has changed over the years, especially with regards to Supergiant Games’ more modern projects: Pyre and Hades.

“Our games take inspiration from a lot of time-honored traditions in game design, so we know we have these conventions such as extra lives, continues, and game-over screens to draw upon,” Kasavin responded. “But we never want to leverage these conventions thoughtlessly, sight-unseen, so instead we look closely at how they can be adapted to fit the particular experience we’re trying to create. We also consider what opportunities we have to subvert the convention and player expectations in an interesting way.”

Pyre, for example, has no real game over state. As a story-based RPG, but with gameplay similar to NBA Jam, each match the player loses simply factors into an overall record. Characters react to victory and defeat in unique and persistent ways. Sometimes you can only learn an opposing team’s backstory by losing to them. Still, Kasavin explained, most players didn’t even experience that aspect of the game.

Punishment Vs. Reward

“The decision to make sure Pyre had no game-over state created a ton of extra work on that project that we knew full well would not affect most players, who have been conditioned to retry in games until they succeed,” he said. “It was still the right decision for that game, and I’m really glad we did it… We wanted a game where you had to confront and move on from your mistakes, and become stronger from them.”

Indeed, it seems to me that defeat is actively rewarded, rather than punished, in both Pyre and Hades. Hades even requires the player to die a number of times before unlocking specific features and meeting new characters. I asked what it was like to design a game around the idea that the player should lose more often than win, but Kasavin said this isn’t necessarily Supergiant Games’ intention.

“In general, we don’t really approach player failure as something to be rewarded, but rather we do our best to make sure it isn’t disproportionately punishing, and that it fits with the narrative,” he stated. “Above all, we want to acknowledge that failure is part of the experience, and therefore it should be something that’s interesting.”

The Depths of the Underworld

Kasavin pointed out that in the case of Hades, the rogue-like structure means dying and starting over is an integral part of the experience. Supergiant wanted to add a sense of narrative continuity to what the studio already knew (from other games in that style) could be an inherently compelling structure. The narrative helps downplay the consequence of dying. That aligns with the typical player experience, since you die all the time in rogue-likes.

“Hades is really different from our other games because it’s designed first and foremost around replayability, whereas our other games are designed more to be played through as a discrete narrative with a more traditional beginning, middle, and end,” Kasavin continued. “Our previous games each offered reasons to come back and play through the story successive times, but Hades is fundamentally designed around playing through repeatedly.”

This got me thinking about which other games (or types of games) could be most improved by removing game-overs from the picture. So I gave Kasavin a more hypothetical question: How would he apply this design philosophy to other games in the industry if he could?

Cost-Benefit Analysis

“I don’t think removing game-over states is something more games should necessarily aspire to, or would make games better,” he answered. “But I do think player failure is an endlessly fascinating topic in game design, and that there’s limitless potential for games to do interesting stuff around player failure, up to and including removing the ability for players to fail at all.”

Why not, I wondered? Surely there are plenty of games that would benefit from having more narrative justification for failure. Even taking “game over” out of the equation altogether could surely improve some games. Right?

“Our focus is on making the best games we can, which can spark your imagination like the games you played as a kid,” Kasavin said. “I don’t know about you, but I for sure hit a high number of game-over screens as a kid. I have fond memories of many of them. I also played games like the original Wing Commander, which let me lose and let the story keep going, and that’s always stuck with me.”

What Can Be Learned

Ultimately, Kasavin explained that he still loves and respects the traditions of classic games. He grew up playing a number of classic games that have been significantly influential on Supergiant Games’ work. Rather than rejecting their conventions and ideas, Supergiant wants to examine what made those ideas work so well — to see if the company can build on them in interesting ways.

“The concept of ‘game over’ is core to almost any game with a sense of traditional gameplay challenge to it,” Kasavin concluded. “For many games, a traditional failure state is vital to how they work. I would never want to see that style of game go away, and I think it never will. Of course there is also room for games that do not have traditional challenge or game-over states. Games can be almost anything.”