The purpose of this web application is to effectively and heuristically optimize fair teams for games based on skill level and factors such as team synergy (who works well with who, and who doesn't).
Inspiration -The inspiration of this project was the constant annoyance of having to make teams in sports such as basketball, or video games such as Valorant.
What it does - The purpose of this project is to find the optimal fair teams for a game, with no limits on anything such as amount of teams or if the teams will leave out any subs (the program does work even if there will be people left out!). The program will output the optimal teams for your game in a matter of seconds to a huge precision!
How we built it - We used a genetic algorithm for optimizing the best way (heuristically within reasonable time limits) of creating teams. Using various technologies, primarily being Python, but also Bootstrap and SweetAlert for the front-end, the product works well within its runtime limit of approximately under 15 seconds (around 3000 generations with a population size of 30 games)
Challenges we ran into - As this is the first time we have done a hackathon, delegating tasks were a new thing for us that we were able to crack using our specialized skills. This was the first time the front-end developers (Joshua and Patrick) connected to a hefty backend for a full product. This was also the first time Kevin was delegated work to parse between the two using a consistent format. Finally, this was the first time the back-end developer (Rohan) created a genetic algorithm, or any Machine Learning algorithm, a fun challenge, but challenging nonetheless for the whole team.
Accomplishments that we're proud of - Being able to put together a multi-person project for the first time, additionally using some techniques we are unfamiliar with, was an amazing experience, and as a team we are proud of what we have developed.
What we learned - Technical algorithms and such aside, it was amazing to learn about what it is like working in a team, and learning how to string files, modules, and people together to create something amazing. It's a new dimension none of us have experienced in the technical world, but are grateful to have.
Log in or sign up for Devpost to join the conversation.