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)
Little Demon Tristana will be live any day now, which makes this our last update on the feisty little yordle from hell. Today we’re talking about bug fixin’, PBE feedback, and the behind-the-scenes work of skin design.
But first, here’s a time-lapse video Little Demon Tristana’s splash art illustration. You can find more details on the process in our previous post.
Sending the Little Demon to PBE
One of the last things we do is name a skin and write the skin’s bio. It’s something we think about throughout development, but don’t lock in until the end. In the case of one-off skins like Little Demon Tristana, we usually have a working name (in this case it was “Player Choice Tristana”) and a working backstory. One thing we incorporated due to player feedback was tying her to the Bewitching Tristana skin, which we wrote a little nod to in her skin description.
A witch who is now one of the great queens of Hell, Tristana commands a vast army of lesser demons and rules the twin Aspects of Ruin and Woe. Mortals attempting to invoke her will find a variety of tempting contracts, but signers beware: Tristana’s ultimate goal is bathing her signatories—as well as the world—in hellfire.
For bigger thematics like Star Guardian or PROJECT, the background story is something we keep in mind throughout the entire development of the skin. If you’re interested in learning more about how we think about thematics, check out this Boards post I wrote on the subject or this dev blog about naming skins.
After naming, our next step is to put her on PBE for player feedback.
Getting It Right with Quality Assurance
Alan “Riot KiWiKiD” Nguyen, Senior QA Analyst: Hey friends! I’m Alan “Riot KiWiKiD” Nguyen, and I was responsible for all QA efforts on Little Demon Tristana. There’s a lot to QA, but for this post I’d like to break it down into three parts: ensuring healthy development, testing, and addressing player feedback.
Ensuring Healthy Development
Throughout concept and development, QA want to establish that the way we’re creating the skin will be the least troublesome for development and most valuable for players long term. We always consider how buggy or disruptive certain decisions can be, especially early in production. Any hasty decision can snowball into serious development pain and be a lot harder to deal with if not caught early. A good way to consistently gauge decisions is by asking questions.
For example: If we were to add a pet or familiar to a skin, does it add enough player value considering time needed (to rig, animate etc.), and can the skin support the proposed pet/familiar? Another example: Little Demon Tristana has wings! During development we asked questions like: How big should they be? Do we have enough support for them? Do we want to make them look different during Rocket Jump (W)?
One other part of ensuring healthy development as QA is coordinating playtests! We playtest skins weekly to measure both specific and overall feels. Playtests also make sure we’re not straying too far away from the thematic. Each skin has specific feedback points and questions we’d like the playtester to keep in mind and/or answer. Here’s an example of Tristana’s points during one of her first playtests:
Red flags include anything alarming such as Tristana ult is covering the screen or Tristana’s gun going missing, and we consider a recall “blocked” when the basic animation sequence is in place.
After playtests, feedback is compiled and given to the team, which they’ll consider and potentially act upon. For example…
We did end up acting upon both feedback statements: We cleaned up her explosive charge (E) passive, and also added a fair amount of purples and pinks into her kit! Yay for playtesting.
We conduct actual testing on Little Demon Tristana via test suites designed to make sure no funny business is going on—no bugs, nothing working unintentionally. Put shortly, a test suite is a collection of tests. These test suites are quite extensive and can take up to two weeks to finish.
Sneak peak of small part of a test suite:
This specific test case accounted for Little Demon Tristana being able to be altered as expected when a specific ability or item affect her.
We want to make sure the skin is tested for both common and uncommon interactions. A skin can be quite different in terms of tech and interaction compared to the base champion. Bugs can occur and exist by a myriad of ways: Lulu and Tahm Kench existing (darn you Polymorph and Devour), VFX covering a full screen, a new part of body not yet animated, etc. The main question we’re trying to answer through testing is, “Does the skin work exactly as how a player would expect?” And from that, we ask, “Are all my VFX thematic? Are the SFX appropriate? When I am stunned does the indicator show up? etc.”
Here are some of the bugs we found while testing Little Demon Tristana. Enjoy!
1. Easy to overlook: Little Demon Tristana was using classic VFX for units surrounding the selected target.
2. Clarity Cleanup: Little Demon Tristana’s Rocket Jump was rendering over impassable terrain—which is just fancy wording for appearing over walls—which is something we only do with crucial spells. We don’t want too much clutter among Summoner’s Rift.
3. Most interesting bug award for Little Demon Tristana goes to: Buster Shot (R) overlay applying to epic monsters, whoops! Although a player ulting Baron with Trist hardly ever happens, we still wanted to fix it for cases the player would want to steal or secure baron with ult.
Addressing Player Feedback
Once the skin’s on PBE, we create a PBE thread for players to report bugs and give feedback. We’re always looking over feedback to see how we can improve the skin, as well as monitoring forms like Twitter, Reddit, and League Boards globally for any potential bugs we may have missed. We’ll respond when we can and are working consistently to make sure the skin will resonate with the majority of players.
In the end, the skin is created to make players happy. We do whatever we can to make that possible, and are very grateful for the time, feedback, and suggestions that y’all provide us. Thank you.
Incorporating Player Feedback
Janelle “Riot Stellari” Jimenez, Product Manager: One of the most difficult jobs of a product lead is deciding what changes we should make (if any) based on the player feedback we receive while a champion is on PBE. Most feedback is subjective: I like blue, you like pink. I hate chocolate, you love it. There’s nothing that makes everyone happy, so a product lead has to decide whether or not the team agrees with the feedback.
For example, we saw very early feedback expressing that fiery VFX would be too much like Dragon Trainer Tristana, so we made it part of the product vision to make it meaningfully different. When we pivoted to blue and violet VFX, feedback was largely positive—however, there were people who felt that it strayed too much from the concept. Another pivot we made was removing the Teemo head bomb on her E. We got a lot of player feedback that the focus of the skin was too much on Little Devil Teemo and not enough on Tristana—and further, the bomb turned this into a joke skin. As much as we loved the idea, this was a Player Choice skin, so we decided to implement that feedback and changed her bomb to a spirit orb.
We also made some minor tweaks while she was on PBE, including making her skin tone more of a vibrant pink, her fangs smaller, and her hair more white. While some players praised those changes, other people liked the original better. As a team, we found these changes made the skin look more polished, so we implemented them. We’re constantly trying to balance the feedback we get with the decisions we think will make the skin stronger.
After everything has been addressed to our satisfaction, the skin goes live.
Research and Meta-Choices: What Does “Success” Look Like?
Chelsea “riot aeneia” Hughes, Insights Researcher: Hello again, everyone! Riot aeneia returning with another update from Insights. ‘Cause that’s how we work. We’re there, start to finish, through the process of iterative feedback!
As an Insighter and Strategist, there are two main parts to my job: First, working with the team to identify what “success” looks like; and second, helping the team to figure out if we achieved that success. While the Player Choice Tristana campaign seems pretty straightforward, there were a lot of decisions that went into it: How many choices should we offer to players? What type of voting style feels best? What kinds of rewards should we offer players who participated? And finally, How do we ensure that this experience is enjoyable for players? At Riot, we pride ourselves on being data-informed. We actively seek out data to help guide us towards the right decision.
That, my friends, is where I need your help.
As the Player Choice Tristana Experience comes to a close, I’ve been working on a survey to address some of these questions. So keep an eye out in your inbox! The information we get from that survey will inform those decisions for the next Player Choice Skin.
Bringing It to the Finish Line
Harald “Agile Harry,” Senior Development Manager: Hi everyone, Harald “Agile Harry” here, and I’m a development manager on skins. I hope you’ve been interested in the process of making a skin; that’s what I geek out on. One of the things I do is to make sure all the amazing artists are working together towards a shared understanding of where we all want to end up: The best version of the Tristana skin you chose from the vote a while back. Here are some basic structures we put into place to help guide the team.
Timeboxing: We decide how much time we want to spend working on the skin. Let’s say we were to spend one year on making Little Demon Tristana. We probably could make the best skin we have ever made, but many players would probably be super unhappy for the long wait. On the other hand, if we were to spend only a week, we would get the skin to players quickly, but you would probably be disappointed in the quality. The sweet spot is somewhere in between.
We guessed we need about four weeks on Little Demon Tristana in development time. By having a timebox, we are training ourselves to make hard choices and be smart about what creates the most value for players in the given time.
Art and Product Pillars: We consider a skin “done” when we hit our product and art pillars for the skin, among a few other essential things. For Tristana, our pillars were:
- Adorable and demonic, not “edgy”
- Right-hand yordle of Little Devil Teemo
- Fiery and mischievous
Once we hit those pillars, it’s time get the skin into players hands as quickly as possible.
Cross-Functional Teams: Here’s something really cool: All the artists working on the in-game skin actually sit together, with the exception of sound effects designers because they need a specialized sound room to create awesome SFX. We organize in cross-functional teams and the artists sit directly next to each other to easily talk about the skin from all angles. This makes communication quick and easy and allows for collaboration and agreement when changes need to be made. Here is what that can look like:
Team Happiness: While it’s important to let players immerse themselves in new skins as fast as possible, we also want to pace the work for our developers at a sustainable way. It’s important that people are happy at Riot. We do what is possible to take care of our team members.
Many things can contribute to fluctuating happiness levels—they may drop when other team members don’t come through with their commitments or it may get boosted when people acknowledge each other for their great work. We ask team members to rate their team happiness at every Retrospective, a recurring team gathering to inspect where we can improve as a team and decide on action to take. Together, we come up with ideas to help each other maintain high levels of happiness throughout the year so we can do our best work when creating new skins.
That’s a Wrap
And with that… we’ve reached the end of the journey. From the whole team, thank you for helping shape Little Demon Tristana into the best demon yordle ever (don’t tell Teemo). But really, creating a skin with players all over the world has been an incredible experience for all of us. To everyone who voted—be on the lookout for a Little Demon icon in your inbox when the skin goes live in patch 9.11 (it might take a few days to distribute them all!) If you missed the vote, you’ll be able to pick up the icon in the BE shop in the latter half of this year.
If you want to revisit any of the steps in the process, check out the links below.
And the Winner Is… : Where we announced the poll results and kicked off Little Demon Tristana’s production.
Splash Art & VFX: The beginning of VFX and splash art exploration, plus why we chose Tristana.
Model & SFX: The first look at Little Demon Trist’s character model, animation, and spooky SFX.
The Latest on Little Demon Trist: The update where it all starts to come together.
Splash Art and Chromas: More on creating Little Demon Tristana’s splash art and chromas.
The Final Update: You’re already here. Thanks for reading :)