Two part project for analysing games played in lichess.
Part 1) Data engineering: I created a script to parse and condense the chess games available in the lichess database: https://database.lichess.org/ The games that are included in the analysis are only those that have an evaluation available (around 6% of the games according to the lichess documentation).
Part 2) Data analysis: Naturally I could not help myself and I went and analysed the wonderful dataset that was created in step 1 to see whether I could get insights about the playing habbits of chess masters broken down by theri rating. https://www.kaggle.com/noobiedatascientist/analysis-of-lichess-games?scriptVersionId=55683619
Some visuals from the notebook:
Popularity of openings accross rating categories:
Number of times the balance of a game changes by rating and time control: