League of Legends 2022 World Championship Visualization

Information Visualization - School Project

TL;DR

Skills: Data Cleaning, Data Analysis, Data Visualization, Usability Testing, Prototyping, Tableau, R

Top Lessons Learned: 

  • This was the largest data set I'd worked with so that was a great challenge and sparked my interest in doing more quantitative work in the future!
  • Conducting and interpreting usability testing on visualizations was a new and educational experience - particularly when a visualization like this needed to target both expert and casual players.

Project Background

League of Legends is one of the most popular games in the world with over 150 million registered players, and has a thriving E-Sports scene. It contains highly complex and regularly changing “meta”; game meta is the best strategy to win, and analyzing key variables of the game drives better chances to win at both team and individual level. We found some visualizations online, however they were overly complicated and contained a lot of irrelevant data that we believed could be simplified.

This project came from an Information Visualization course at the University of Washington. Our team decided to create a dashboard from a large data set collected during the 2022 League of Legends E-Sports season. This dashboard includes visualizations that cover champion stats (such as pick and ban rates, champions by position, etc), player stats (KDA, gold per minute, average damage, etc), and team stats that detail neutral objectives.

User Research

Research Goals

There were two main goals in our user research:

  • Understand League of Legends E-Sports data to understand what kinds of visualizations we can and/or should create.
  • Understand what kinds of data visualizations League of Legends players currently use and are interested in using in the future.
Secondary Research

I started off by conducting secondary research and learning more about League of Legends, the game meta, the E-Sports leagues, and more. This gave us insight into what players might want to see in our final visualizations, and helped guide our initial data exploration.

Expert Interviews

I conducted several expert interviews with long-term League of Legends players and E-Sports followers to help guide our analysis. I asked clarifying questions on variables present within the data, what kinds of information they would be interested in seeing from the type of data we had, and what websites or sources they used to keep up with the current season and meta.

Data Cleaning

Before starting to create our visualizations we had to do some data cleaning:

  • I profiled the data in excel and then used RStudio to clean the data.
  • First I cleaned the columns by eliminating spaces and filtering out columns that were unnecessary for analysis (such as url, gameid, etc) or that we identified as irrelevant to our analysis (such as wards placed, specific dragon types, etc).
  • We filtered the data so that only players and team data from the top 25 performing teams were in our data set.
  • We checked for misspellings and outliers, and scaled the data.
  • The final cleaned data was written into two CSV files - one for team data (objectives gained by the teams, etc) and one for player data (KDA, first blood, etc).

Early Visualizations

After our initial research, we created some early concept visualizations, which we took as a chance to explore different ways of presenting the same data or concepts. Throughout the quarter we continued to iterate and test these visualizations.

User Flows & Testing

We worked on some mockups, flows, and initial dashboard pages with visualizations which you can see here:

After seeing what visualizations we had created separately, we came back together to see which ones would work best together, and decided on our final direction, information architecture, and what other visualizations we would need. We initially decided to divide up our visualizations between team data and champion data, however we later decided to have a player page that would focus on individual players to allow comparison across teams, a team strategy page that focused on neutral objectives, and a champion stats page.

Once we put all of our initial visualizations into tableau and created a low-fidelity navigation, we began user testing. We conducted five usability tests and did rapid iteration after each test.

After our third usability test I designed the navigation in Figma, and began designing the rest of the dashboard by adding pro players and champions images, customizing the colors according to the League of Legends brand guidelines, and more.

Project Demo

Reflections

Limitations: We had quite a short amount of time to clean our data, visualize, test, and iterate. We were also limited to data from 2022 which is not very current in the world of E Sports. If we had more time we would have loved to look at this data longitudinally, to see at how the variables we visualized vary by region, and to look at more individual game stats rather than overall averages.

What I Would Do Differently: If I were to redo this project I would take on all the data cleaning myself, or have my teammate Connie do all of it. Because of the class structure, our team did data cleaning and initial visualizations in groups of two, which lead to several data file versions and visualizations that were difficult to piece back together in Tableau.

I would also have switched to Tableau Public much more quickly. We didn't realize until halfway through that we could all access a shared Tableau Public file through a browser, and had previously been working with our own individual Tableau files, which caused a lot of frustration.

LinkedInInstagram