Riot Games | Valorant
WELCOME TO
RIOT GAMES | VALORANT
Duration | December 2021 - January 2022
Project Type | Personal Project
0.0 | MEET THE TEAM!
Tim Joo ( Das Me! :D )
Role: Project Manager | SQL Database Engineer, Designer, Analyst, & BI Engineer
Brian Kyoungbari Kim
Role: Data Scientist, Data Analyst
0.1 | why me and Brian?
Brian, a recent graduate at General Assembly as a Data Scientist/Analyst Student, needed a capstone project to finalize his time at his data science bootcamp.
With my past background in teaching Data Modeling, Data Engineering, and Data Analytics, my personal goal was to give him exposure to this crazy data world through genuine, thoughtful, and impactful experiences.
0.2 | topics we covered
SQL
Data Modeling
Database Creation / Administration
Data Engineering / Analytic Methods
PowerBI Dashboard Creation
Python Data Science Techniques
End to End Data Collection
1.0 | PROJECT SCOPE
1.1 | PROBLEM STATEMENT
As the world of Esports rises and more viewers are tuning into video game streamers, there is a major need for a fantasy league for professional gamers.
If a reliable database were to be created, what would the data model and design look like and how can we solve deep questions to predict upcoming Valorant competitions?
1.2 | PROJECT GOALS
Create a fantasy football style environment for the Valorant community
Create more involvement and engagement within the Valorant community
Initiate more awareness and increase resources as only a few of these exist
2.0 | VALORANT TOURNAMENT
Hosted a Valorant Tournmanet
20 people | 4 teams | 1 winner
2.1 | Initial Issues
Lacked the access to personal Riot Games API / Datasets.
Had access to basic Schema of their database, but realized it was JSON based in a NoSQL Format.
Due to this, determining relationships between entities became nearly impossible.
2.2 | Issues Resolved
Decided to host our own Tournament to collect Data for a Cloud Based Azure SQL Relational Database
Hosted Custom Matches to have access to all match data and history immediately after each game.
Hand inputed each data point in Google Sheets as we still had no access to an API.
3.0 | DATABASE DESIGN / CREATION
Final Entity Relationship Diagram
why a Relational Database?
- adheres ACID principles
- quick & easy access for data visuals
- 3rd Normal Form for no unneeded duplication of data
4.0 | DATA VISUALIZATION
working on a updated visual! should be arriving within a week or two! :) - Tim Joo 1/30/24
KEY INSIGHTS
ravenjuzo played the best in terms of most first blood, kills, and the highest average score
Team 4 on average played the best out of all the teams in terms of kills and assists however Team 1 has the most plants and defuses
In terms of first blood, gold players played better than platinum players.
Astra had the highest average score of all the characters played
5.0 | FINAL THOUGHTS
Working alongside my friend Brian has been INCREDIBLE!!
We both were able to combine our Data Science and Data Engineering skills to build an end product solution that is actually both incredible and powerful.
On a more personal level, it felt incredible to be able to use my past knowledge and teach Brian the ins and outs of Data Modeling, Data Engineering, and SQL as a whole. It is a skill set that will empower him further to do truly game-changing analytics throughout his career and I can’t wait to see how it all unfolds for him.
This is the true definition of passing on what you learned so others can succeed as well!