Architecture
The general architecture of the project.
This is still in progress and we're working on it.
🗓️ Tracker
The tracking feature is a classic form input. Once the form is filled, it is sent to the backend and stored in the tracker_results table. All the stats and charts are client-side generated and updated in real-time.
📊 Analysis
Analyse from log files generated by the scrim workshop code : DKEEH from Scrimtime.
Once imported, the files are parsed and analyzed to provide a better understanding of the game. Here's a sequence diagram of the process :
The .txt files are sent to the backend, then to the Python processing server. Once the analysis is done, the backend sends the alert to the frontend via websockets so the user doesn't have to refresh the page.
🔎 Scouting
Our scouting part is dedicated to fetch players, teams, matches information. For that, we have 2 main sources : FACEIT and Blizzard.
Our frontend calls are made with Tanstack Query so we have a nice cache system and we don't have to worry about the data fetching.
🥊 FACEIT
We use FACEIT API to fetch data from the platform. We are able to get players, teams, and matches information.
Some of the data provided by FACEIT seems not to be accurate (results in player history for example), so we have to be careful with the data we get from them.
❄️ Blizzard
To fetch data from Blizzard, we use the OverFast API provided by TeKrop.
Blizzard let us get some players information that have a public career. When the data is available we can get stats about the players and its heroes, but also about the winrate and the number of games played on the different roles. I recommend to check out the OverFast documentation which is really well done.
👥 Teams
Teams are a way to group players together. Nohting complicated behind it : a user can create a team or join a team.
If a user create a team, he's automatically considered as ADMIN. An ADMIN can manage its team. Therefore, he can promote other players to ADMIN or KICK them from the team.