The Billion-dollar Roguelike Game

Roguelikes have been an indie favorite for decades now. Who will be the one to take it mainstream?

Hades. Spelunky. FTL. Slay the Spire. Dead Cells. Vampire Survivors. These are some of the most beloved games of all time, often found in the top 50 charts on Steam. However, for the vast majority of casual players these titles remain relatively unknown. I believe that the founders that can adapt roguelikes for the mainstream will generate the next big hits, likely in one of two ways:

1. A singleplayer mobile game with simple mechanics, midcore meta-progression, and strong monetization.

2. A co-op multiplayer rogue-lite that draws inspiration from roguelikes but is tuned for a native multiplayer experience.

I’m not a designer of course, but I’ve spent a lot of time in the genre and was inspired by recent poker-roguelike hit Balatro to put pen to paper on where this genre might go. Let’s dive in.

Balatro. The sneakily fun poker-roguelike that’s got us all doing multiplication tables again.

What’s a Roguelike Anyways?

For those that are less familiar with the genre, roguelikes are characterized by two core elements:

1. Randomly generated levels, including rooms, loot, enemies, and progression paths that are partially or fully generated by an algorithm.

2. Permadeath, where each run ends with the player losing their progress and starting anew.

These pillars make roguelikes compelling for several reasons. First, player skill is crucial, both in mastering the core combat mechanics and understanding how the interplay of abilities stack to create the best builds. For instance, in Hades - an action-roguelike where you play as the son of Hades trying to break out of underworld - part of the game is mastering the mechanics to dodge attacks, predict enemy movements, weave around rooms, and strike at the right time. And the other part is understanding how to min/max the right build given the randomized boons (powerups) you have in each run. What’s unique about roguelike games is that it forces you to make strategic tradeoffs that are different each run: Should I get an Aphrodite attack debuff? Or a Poseidon knockback? Perma-death means that each decision matters, and good players have many ways to show off their skill expression. However, it also means that the genre is fairly hardcore, which Hades solved well by layering in narrative progression with each run that meant you were always moving the story forward.

Choosing the right boons in Hades can make or break your run.

Second, each session is unique. Because the power-ups, enemies, and levels in each run are different, players are forced to adapt to each run differently. There are different synergies between all the powerups that allow for many types of strategies and win-conditions. It feels really good to finally, through luck and smart choices, get a “cracked” run that feels extremely overpowered, creating a “water cooler” moment that encourages players to share their newly designed metas. Take Slay the Spire as an example, which has a card-based combat system. Even though each character only has a limited set of cards, cards interact with each other uniquely, exponentially increasing the number of decks possible. In the picture below, you can easily see how the Silent might want to stack “Bouncing Flask” and “Crippling Cloud” that both apply poison with “Catalyst” that doubles the enemy’s poison. But if you don’t get those cards in your run, then you might have to run a Shiv-strategy or an energy-strategy, totally different from your initial intention.

Some of the cards you can get as the “Silent” in Slay the Spire.

Finally, roguelikes solve the content treadmill problem with player vs. environment (PvE) games. A lot of games, from casual games like Candy Crush to MMOs like World of Warcraft, have a set amount of content for players to grind through. The speed at which players will churn through content is often faster than the developer’s ability to create it. Roguelikes’ procedural generation and varied combinations allow for much higher replayability given a set amount of content. Of course, that also means they’re much harder to design, since you have to deeply think through how to properly balance all the components of the game.

The Mobile Singleplayer Roguelite

So why haven’t roguelikes gone mainstream yet? Well, actually, some of them already have. According to SensorTower, Survivor.io from Habby peaked at ~25M MAU after launch and still generates $120M+ in revenue annually. Habby took the proven, extremely accessible loop of Vampire Survivors - the only player input is movement through a virtual joystick - reskin it in a more accessible casual art style, and add modern meta-progression and monetization techniques (battlepasses, piggy banks, crafting, lootboxes, etc.). They also ran a smart TikTok strategy for UA with a vertical game format that was perfectly tuned for the platform.

Data on survivor.io from Sensor Tower since launch.

Some of the UA creative for survivor.io

The main challenge that these singleplayer mobile roguelikes face is that they need to tune the roguelike genre for a more casual audience. Permadeath is a pretty hardcore concept because it mandates loss as part of the core loop. Think about Monopoly Go as a comparison point, where even when you lose (i.e., your attack gets blocked), you still win cash in a colorful and satisfying explosion.

That’s why many of the successful mobile games will be rogue-lites; the player still dies but every run feels meaningful because they unlock coins / abilities that permanently power up their character. The challenge that roguelites face is that the metaprogression (more HP, more damage, new weapons, etc.) mean that the game is hardest at the start and gets easier over time, a monotonically decreasing difficulty curve. This is the classic depth vs. accessibility challenge.

The other important facet is making the game easy to intuit and learn. That’s why games like Vampire Survivors or Survivor.io do so well, because there’s literally one mechanic, movement. However, there’s more depth to the combat than you would think; learning the various skill trees, positioning attacks properly, and slipping through groups of enemies take time to learn, and better players will impose their own limits like no-hit highscore runs. .

That’s also why Balatro presents an interesting take on where roguelikes might go in the future. Most of the roguelikes we talked about up to this point are focused on combat. But Balatro takes the 52-card deck that we are all familiar with and gives it a unique roguelike twist. Add to it the simple satisfying loop of “number go up” and you get a smash hit. And with all roguelikes, there’s also depth: trying to quickly assess what your best hand can be, what cards are left in the deck, and making some risk-adjusted guesses on the next play.

I’m curious to see what other non-combat roguelikes might spin up because of Balatro. Maybe there will be interesting platformers, runners, or board games that people experiment with. Technically, even Flappy Birds is somewhat of a roguelike with its algorithmically generated pipes!

The one other important callout with mobile games is monetization and meta-progression. Currently many of the indie roguelike games have traditional premium-box monetization for $20-30. While this is certainly one of the best player-value deals out there, it won’t result in the billion dollar outcomes that VCs and publishers are looking for. For that, we’ll need to look to traditional F2P monetization techniques while balancing the game to not be too pay-to-win (P2W).

Yang at End Game is doing some interesting exploration here. One of the early game prototypes (see below) they’re developing combines the roguelike simplicity of survivor.io and the hero collection / monetization of Genshin Impact. Having a roster of heroes solves some of the pay-for-power issues by allowing for different meta team comps rather than one overpowered weapon that you can buy. Banners and live events can mitigate some of the power creep issues that roguelites encounter. It also introduces more depth of spend; Genshin Impact, one of the top grossing hero collectors, has some of the highest revenue per download (RPD) at $95 in Japan and $35 in China. Meanwhile, Azra Games is another company that’s using roguelike mechanics to enhance the replayability of its next-gen hero collectible RPG.

The Multiplayer Genre Mashup

The other direction that I believe will bear a lot of fruit is experimenting with what a multiplayer-first roguelike might be. Singleplayer roguelikes encounter a cursed design problem when you add another player. If one of the players die, do you continue the game onwards while the other person watches? Not very enjoyable for the weaker player. If you have a revive mechanic, how often can you use it? Use a revive too frequently and you remove permadeath and the tension of the game. Some of the games like Spelunky or Rampage Knights solve this by allowing the dead player to turn into a ghost that still has some core actions until they’re revived (similar to Among Us), but it’s still not a perfect solution.

Instead, it’s likely that a roguelike that’s designed to be co-op first will be more successful. However, there’s many different directions that this could lead. There’s No Man’s Sky, which generated near-infinite levels at the start for players to explore and colonize together. There’s Diablo, one of the original ARPGs, which took a lot of inspiration from roguelikes for its procedurally generated dungeons and random loot drops. There’s Don’t Starve Together, which has a co-op twist on the original survival crafting roguelike game with its large, procedurally generated map. And of course there’s Risk of Rain, which is the most traditional co-op roguelike that scales the difficulty, monsters, and loot up and down depending on how many people are playing.

The key you might be noticing with these games is that they’re more rogue-lite than rogue-like. Because of the hardcore nature of the most punishing, skill-based roguelikes, most of the multiplayer roguelites will likely borrow features from its parent genre to enhance its gameplay: procedurally generated levels, randomized loot / enemies, a variety of synergistic abilities to collect, and death / revivals as part of the loop. These multiplayer roguelites will likely have some sort of genre blending (FPS, ARPG, survival-crafting, etc.) and have a premium + liveops model in order to smooth the revenue curve over time. Helldivers 2 is a good example of what the future could look like here, with a galactic war map that changes and evolves based off the collective of player actions. There’s quite a few startups that are going after this thesis. A few are in stealth and so I won’t reveal them here, but Fuzzybot is one that’s public and combining roguelite ARPG combat with the crafting/collection loop of a life sim.

The galactic map in Helldivers will shift based off of where players choose to deploy and how well they perform.

Conclusion

Roguelikes are one of my favorite genres of games, and I’m excited to see where they evolve from here. I must also mention the obvious potential that AI has here to change the kinds of games that can be created here. Levels could be generated with more detail and with more immersion, NPCs could react in more intelligent ways, and live ops could be personalized more effectively for the player. It could also change the kinds of games that we play. Suck Up! for instance, is a roleplaying game where you masquerade as a vampire and have to talk innocent townspeople (powered by LLMs) into letting you into their houses. There’s no perma-death with this one, but talking to LLMs is inherently random with many branching paths, similar to the level and enemy randomness of a traditional roguelike.

If you’re building in the genre and innovating on what’s new, feel free to reach out!