Development of a crawler to collect online game playing traces

Data is generated by technology every day, one of which that has been generating a lot of data is eSport. In the last 10 years, eSports have become more popular than ever before. And with the rise in Machine Learning and Data Science, eSports have given the data to analyse certain human behaviour with numbers. In the game League of Legends, players are categorized by rankings and object met by players in games are stored as numbers. Games constantly receives update and tries to balance the game to ensure fair play. Through the different patches of the games, what stays constant are the objectives met by players determining their skill. As such in this project, the aim is to analyse the players in games and determine how higher ranking players have different attitude towards a match in game like how professional sports player treat their game. Prediction of the result of a game will also be done using Neural Networks. The project will visualize data into proper information showing trends and patterns found through the analysis. Part of the scope of this project includes of the process of data cleaning and reducing the size of the data by removing redundant information that does not show much of player skills as well as transforming data from objects to integer or bool to save space. Results of this project has shown that teamwork is more important than it seems in a game as well as player skill generally does affect the result of a game. Higher skilled players also tend to continue a game even if they are losing whereas lower rank players give up on the matches.Bachelor of Engineering (Computer Science