Dev
/dev: A look back at League in 2020

Join Safelocked in reflecting on the past year of League of Legends.

DevAuthorSafelocked
  • Copied to clipboard

Hey all. 2020’s felt like the longest and shortest year all at once, but now that we’re finally at the end, we wanted to reflect back on what’s happened in League this year. Our theme was focusing on making each and every game you play, honestly, just a little bit better. That required us to make good on some longstanding promises as well as face some longstanding problems. This needed to be our focus because we knew we’d need to strengthen League's foundations to support everything we’re planning in years ahead.

We broke down the core game loop, looking at pre-game, in-game and post-game. Let’s start with pre-game first.

Client

client.jpg

Client performance was the biggest focus when it came to improving the out-of-game experience. We started by improving our telemetry to understand more of the why behind problem areas like Champ Select and post-game screens. Unfortunately, it highlighted that client screens could be unreliable due to a myriad of issues, not just one we could knock out with a concerted push. We dug in and made steady improvements to reliability, responsiveness, and load times in the client, gaining momentum as we learned more.

That said, over the years, we’ve learned to stop counting chickens before they hatch when it comes to how noticeable performance improvements are. But now that we can see the combined impact of this year's changes, we can confirm tangible, chunky improvements when you compare the client at the start of 2020 to the client today. It's faster on startup, loading in 16 seconds or less for most players. Architecture improvements led to better responsiveness and reliability of the screens you use most, like Champ Select, End of Game, Runes, and Collections. We reduced out-of-memory crashes by 49%, and addressed over 100 bugs to improve the overall navigation experience.

We're not taking this year's gains as cause to stop working, though. There's still room to improve, and we have a full roadmap of more client improvements we’ll keep pushing out in the new year too.

Rise of the Elements (Last year's preseason gameplay)

ps2020.jpg

It’s been a while since we talked about last Preseason, but wanted to reflect on how things settled.

With evolving SR, our goal was to build on more distinct early, mid, and late phases punctuated with emergent gameplay opportunities you can take advantage of in different ways depending on champion kit and map awareness. On the Dragon Soul side, the intention was to create an alternate win condition to Baron for teams to consider optimizing around, depending on how valuable that Soul is to their comp. Dragon Souls (and Elder Dragon) have created strong moments of tension around late-game Drakes that otherwise wouldn’t be as much of a priority. The game-to-game variance from the Elemental Rifts and different Soul types is less than we anticipated, but overall, we believe these changes have made for richer macro decision-making and objective tug-of-wars that we’re happy seeing.

Mythic Forge (This year's preseason gameplay)

ps2021_v1.jpg

Hahah, this system. There have been a good handful of items that have stayed truly iconic over the years in League, but maintaining unique item identities across the entire shop has been a struggle we haven't always come out on top of. In order to address issues that built up across many seasons—unclear item identities, odd build paths, must-have and never-picked items—we really needed a full do-over. As you've hopefully seen by now, we took a class-by-class approach to ensure the shop serves everyone's needs, including the addition of Mythic items.

We also completely rebuild the item shop, your main interaction point with the item system. Better tech will let us keep improving the shop's usability in the years ahead, and this was also our chance to make the shop responsive to how balance, meta, and player choices shift with time. It doesn’t matter if an item was optimal three months ago if it’s not the right choice now. Surfacing relevant recommendations ensures the shop is a valuable, accessible tool for new players and returning veterans alike.

On the feedback front, we're still assessing initial trends. As a general PSA, games are always bloodier in Preseason because everyone is playing a bit more aggro and experimenting more. Higher damage is a concern we've seen brought up a lot for example. We think the reality is a mix of damage actually going up, leading to changes in recent patches, and games feeling like damage went up because people are dying when they don't expect to. While we're interpreting initial balance feedback cautiously, please keep it coming. Understanding which parts of the system remain frustrating into 2021 will help focus our efforts early next year.

The other line of feedback we wanted to call out is around the new shop. Now that we've done the initial rebuilding work, we've been focusing on issues you've called out. Highlights here include resolution fixes, an Options menu slider to adjust shop size scaling (drag-to-scale coming in 11.1), pin functionality for the quick-buy panels, and improved readability of the Recommended tab. The new data-informed shop recommendations are also live, and we're excited to see how they impact decision making. We're still dedicating time to these features, so keep sharing your thoughts.

Behavioral Systems

behavioralsystems.jpg

This wasn’t an area called out at season start, but I mentioned it in our last Riot Pls. This year, we shifted focus from reforming players onto making a better in-game environment. For early efforts, that meant better int and AFK detection. We’re really happy with the results we saw this year: Detection improvements have led to punishing 50% more players who are intentionally feeding.

Like our work on the client, we’re continuing to refine here, and we’ve begun to tackle detection and punishments for players who AFK or idle in the game too. Mitigation features for players affected by AFKs have also recently shipped—LP mitigation for ranked, and an updated Early Surrender system for all SR matches. More to come in this space next year.

Champions

yone_v1.jpg

With the expansive roster we have today, a lot of our conversations around champions have been about more evenly distributing the positions and roles we hit throughout the year, and how to connect champions with events as more powerful and thematic launch windows. As one example, we were happy that Spirit Blossom's additional connective tissue to learn more about Lillia and Yone was well-received. Position/role distribution was better this year too with Sett Top, Lillia Jungle, Yone Mid, Samira Bot, Seraphine Mid, and Rell Support. Several of them can flex between positions, too.

Another recurring topic has been champion launch balance and kit complexity. It's an ongoing effort, but 2020’s launches were more balanced than previous years while also requiring less follow-up to get there. We're going to carry forward 2020's updated approach to new champ balance and continue iterating to get even better.

On the VGU side, we were excited to deliver Fiddle and Voli while giving earlier looks into their development. Unfortunately, we weren’t able to push out a new VGU poll like we had hoped. With Dr. Mundo on track, next year's VGU work is shaping up to go a little more smoothly, so we’re planning to run another poll in 2021. When we do, you’ll see a few familiar faces, but also a few more names that (likely!) won’t be too surprising.

Skins

PP_Taliyah.jpg

Skins have always been tough to navigate when it comes to how we can serve more players. Do you make more skins for popular champions? If you make skins for more niche champions, aren’t we serving fewer players? The difference in how many times a skin gets played is extremely stark depending on champion selection. At the same time, that evaluation can feel like it's missing the spirit and the promise of an evolving roster. If you’re going to take the time to main a champion, you want the game to have your back and trust that we’ll invest time in that champion too.

So, over the course of a few years, we scaled up our content pipeline to gain the ability to make skins for less-played champions without needing to scale back on offerings for higher-playrate champs. We also dug into which skins and aesthetics you like most, what different mains are looking for, why that's the case, and how to use these learnings to make new thematics that feel fresher and more up your alley.

This year, we shipped 144 skins for almost 120 champs. It’s been awesome to see the reception of skins like Cosmic Skarner, Spirit Blossom Kindred, and Pool Party Taliyah. While the difference in playrate due to champ popularity is still there, mains are playing these skins more frequently. So, the creative R&D to match champions to more resonant thematics seems like it's on the right path. We're really happy to see that, and will continue working on this more going into the next year.

Events and Game Modes

nexusblitz_v1.jpg

In Season Start, we committed to running a game mode with every event. This meant bringing back enough modes for you to enjoy a variety of breathers from SR proper. With the reskinned Nexus Blitz as the highest-profile addition, Pick Urf, ARURF, and One For All also joined the stable of ways to enjoy League's in-game events. We’re going to continue looking for opportunities to experiment with, and grow, the modes roster in 2021.

When it comes to events as a whole, we wanted them to feel like more complete experiences. This included game modes, but also things like Spirit Blossom's immersive out-of-game experience which was rooted in Ionian mythos. It also included improvements to how you follow along with the event itself, like this year's Worlds tracker that made it easier to track the tournament from Play-ins to Finals. For Worlds specifically, we also think this year's K/DA was the best event integration yet for a musical experience.

2020 also marked the beginning of collaborations across our games, giving you different ways to experience major moments across the League ecosystem. We know some folks were concerned about what Riot launching other games could mean for League, but one year in, it's only made things better. We’re really happy with how these big events turned out—you really showed up for Spirit Blossom, and Worlds viewership was the highest we’ve seen, ever. We’re all thinking about how to take what went well to future events; fingers crossed that you’ll like the next thing we've got planned.

Clash & Ranked

ranked.jpg

Now that we can finally say Clash has launched, it’s been awesome to see a growing number of Clash teams not only competing together, but in many cases returning for each tournament. Our main focus this year was honestly on securing a smooth launch that could withstand server demand. With more players coming in this year during quarantine, we stretched beyond that, adding more Clash runs for more schedule flexibility. Moving forward, we’re interested in making it easier to participate in Clash and show off your progress, so expect to hear more in the new year.

On to Ranked. This is the environment where we really want to preserve competitive integrity. As the main avenue to test your skills, every Ranked adjustment carries incredibly high stakes—particularly at the highest tiers where even tiny changes can have major impacts on match quality and ladder integrity. Big hitters among the Ranked improvements this year included changes to ensure autofill and premade parity on both sides of the team, keeping queue time impact down to +1-2 minutes in the worst case. And to prep for the new season, we just removed inter-division promos to reduce the frustration of bouncing in and out of series and make your actual game-to-game progress feel more impactful for your climb.

These changes reflect a changed approach to Ranked. Being more incremental lets us make adjustments in concert with your feedback and means we avoid surprise reveal moments where Ranked suddenly works in a totally different way. Overall, we're happy with this method since it enables us to improve Ranked in the most transparent, gradual and natural way possible.

Happy End of 2020

If you’ve read until the end, thanks for going down memory lane with us and hanging in there during 2020. It’s honestly felt like this has all happened in the blink of an eye, but we hope this gave a little glimpse behind the curtain for how this year-long journey went and where we want to improve. You keep us motivated to always improve the game and maintain a high bar going into 2021. Have a great end of the year, happy holidays, and see you on the Rift.



  • Copied to clipboard