Member-only story
Data Science Project — League of Legends Dataset Part 2.
Last time we analyzed the LoL data set for factors that can lead to a teams win from different global objectives. Now we take the data on the team compositions and visualize it.
In our last medium post we took a League of Legends data set from kaggle that contains records of over 50,000 ranked matches. Each record has a number of features including the winning team, the champions used for each team, the amount of towers and other global objectives each team took and more. We analyzed some of that data to understand what global objectives can increase a teams chance of winning a game.
From our analysis it seemed that the teams won 80% of the games where they took the first inhibitor and over 70% of the games when they took the first tower.
Now we look at the features dealing with the composition of both teams to understand if team composition played a role in winning a game. The full code can be found at:
