Ask Riot

Ask a question about League or Riot, and we’ll try to answer it. Answers go live every other Thursday at 1:30 pm (PT)

What do you want to know?

Something went wrong. Try asking again.

Thank you for submitting a question!

Next Article

/dev: Adventures in Odyssey: Extraction

What if players got to be the Doom Bots? Stories from developing League’s new PvE game mode.

By the time you’re reading this, it’ll be the future. The Odyssey event will be live. (We hope you’re enjoying it and aren’t finding the missions to be too difficult!) When we set out to work on Odyssey, we knew we wanted it to feel like an epic event—a surprise League moment with new skins, a cinematic, and lots of missions and loot—and we wanted to cap the whole thing off with a brand-new PvE game mode. Enter Odyssey: Extraction.


We had some pretty big goals we wanted to hit for this iteration of a PvE game mode:

  • Make a PvE mode that players of all skill levels could enjoy, while still having an awesome end challenge for those who are committed to see it through.
  • Add a progression system that persists over many sessions.
  • Ship it in time for the launch of the Odyssey Event, which meant we had 4-5 months for development.

These goals defined what we could and couldn’t do with the mode pretty clearly, and it helped us make some early decisions, such as having five difficulty levels. We wanted to create four difficulty levels for most players to progress through as they powered up, and then one difficulty for our most hardcore players to test their mechanical and teamwork skills.

Other decisions took more discussion and development. For example, the original Odyssey line-up included five skins (Yasuo/Sona/Jinx/Malphite, with Kayn as the villian), but we wanted the mode to have five players. This meant we needed to add another member to the Odyssey team. Based on the team comp we had, a magic damage dealer felt like the right fit. After much discussion, we landed on Ziggs because he seemed like strongest match for both the gameplay and thematic needs of the mode.

Game system progression (which eventually became the Augment system) went through a lot of iterations. Most of our early ideas were quickly rejected because they became too similar to League’s existing item system—things like gaining extra stats or slight modifications to your abilities. However, we eventually found one simple line that summed up exactly what we were trying to do with the mode:


Doom Bots of Doom was a mode we ran several times in the past, most notably with the Teemoing in 2016. In that mode, the bots were extremely powerful and would cast incredibly overpowered spells. It was a lot of fun, and since we were doing a PvE mode, we thought, “Why can’t we let the player do that?” This led to us creating the Augment system: powerful add-ons you earn over time and equip to make your champion closer to a Doom Bot.

We followed one main guideline while developing Augments, which was, “Augments can’t change how the fundamentals of the champion works.” Sona players should still know how to play Sona, even with lots of Augments equipped. So big mechanical changes, like turning Sona’s Q into a charged shot that did more damage if you held down Q, were out.

Some of our Augments were developed extremely early and lasted through the entirety of production (with slight adjustments of course):

Relentless was instantly a Rioter favorite. It was also one of the Augments most prone to bugs.
Shatter, Delayed, and Leftovers were also added quite early and stuck around.
Jinx’s Runaway was another super early Aug that survived development.

Of course, as we were developing the Augments, things didn’t always work exactly as intended. Some of these bugs were hilarious and fun, so we would do our best to keep their essence in the mode. Other times… they were just bugs.

Cascade often didn’t work quite as intended. Sona never minded though.
Sometimes when you don’t set things up properly, it works to your advantage.
Other times it just makes for some... odd visuals. Wrong traps Jinx.

And of course, some Augments were just too disruptive or powerful, so we had to cut them. We found out early in development that crowd control effects were extremely strong, as they prevented enemies from really doing much of anything. With that in mind, you can imagine how powerful some of these were.

Thumper pulled enemies in every tick. It was basically six seconds of hard CC—so OP.
Afterburner used to work on allies, but it turned into the ultimate “troll your team” Augment.
Cataclysm was an experimental Augment that usually did more harm than good. Pretty cool visually though.


While we created four difficulties for most players, we wanted to make one difficulty just for those who really wanted a challenge. It was time to bring back Onslaught. Two Hazards per room, enemies hitting like trucks, and new giant monsters were just some of the things we added. Getting the balance to feel fair and challenging was… well, challenging.

We decided that doing Onslaught with five Augments should be tough, but not insane. Doing Onslaught with two Augments was the real challenge! (We’re not even sure if it’s possible to do with one or zero Augments…) We designed the score system to really push and reward those who take on that level of difficulty. But sometimes, we got carried away:

The Bubble Hazard used to fire basically nonstop in Onslaught. Way too frustrating.
Often times if you fell behind in damage, you were really overwhelmed. To almost comical effect.


Finally, at the end of the mode players face off against Kayn and his trustworthy-and-not-at-all-psychotically-manipulative scythe, Rhaast. While designing Kayn, we wanted to push the limits of what a PvE boss in League could be. While the Star Guardian boss proved that we could demand finesse (dodge various abilities in various manners) and create compelling DPS-optimization questions (do I focus the boss or the monsters he just summoned), these are primarily individual asks.To that end, we decided to focus on giving Kayn mechanics that pushed players to work together to defeat him. Some of his attacks will force players to group, split apart, run in circles, and even pull him out of difficult combat terrain.

But we won’t spoil anything specific here—so good luck!


With Odyssey: Extraction, the team set out to create a fun and exciting PvE mode for everyone to enjoy. There’s lots to do, and we hope you try out different Augment builds, complete as many missions as possible, and see how high of a score you can get. We’re personally excited to see just how high the scores can go.

Please let us know what you’re enjoying (or not enjoying) about Odyssey: Extraction. We’re hoping to learn a lot from this mode and take these learnings to the future. Good luck taking down Kayn and Onslaught mode, and we’ll see you next time. Thanks for playing!


An experimental Malphite augment basically turned him into a jump machine.
Five Malphites hold hands as a sign of friendship.
Our temporary loading screen, which we used for the first 80% of the project.
Sometimes our testing team would slip into madness as they tried to complete Onslaught. We hope you fare better…
Next Article

/dev: Top Tier Changes to Ranked for 2019