12-06, 19:00–19:30 (UTC), General Track
This talk concerns data analysis techniques applied to the Pokémon Trading Card game. From statistical odds of key card draws to insights from 100 matches and a dashboard, I'll show how data, code, and ChatGPT improve my card game strategies.
I love the Pokémon Trading Card game, and unfortunately, I'm bad at it. In an attempt to get better, I've been collecting data on my games with the goal of analyzing them using a mix of tools like Python, R, JavaScript, ChatGPT, and Google Sheets. This presentation details three key projects I've worked on with this data.
First, I will delve into the statistical analysis performed on the probability of drawing a crucial card that players can only use at the beginning of the match.
Then, I'll discuss what I discovered from analyzing 100 matches featuring one deck archetype. The insights I'll share here reveal fun facts, such as my best matchup and how bad my bad games are.
Finally, I'll present how I created a dashboard from a continuously updated spreadsheet file. This dashboard, rich with statistics, visualizations, and scripts, details my evolving performance in the game, offering me up-to-date insights and trends.
With these projects, I aim to demonstrate how data analysis can be creatively applied to light-hearted use cases like card games, potentially inspiring others to embark on similar analytical journeys.
No previous knowledge expected
Juan is a Trust & Safety Software Engineer working at Bitly. His role at the company is to develop solutions to ensure the quality of the links created or extended by the users. Besides fighting spam, fraud, and scam, Juan's an avid writer. In 2021, he published Practical TensorFlow.js and has written countless articles covering topics such as machine learning, quantified self, and quirky ones such as data analysis done using Pokémon data. Juan holds a BSc in Computer Science from the University of Puerto Rico - Rio Piedras Campus and an MSc in Computer Science from Uppsala University in Sweden.