Welcome! This is an in-development one-stop shop for the Pokemon data science you didn't know you wanted, with a concerted focus on specific combat matchups and optimizing parties.
As of March 2024, the current scope of the project is all of Pokemon Generation 1, including:
- Pocket Monsters Red & Green
- Pocket Monsters Blue
- Pocket Monsters Stadium
- Pokemon Red & Blue Versions
- Pocket Monsters Stadium 2
- Pocket Monsters Pikachu
- Pokemon Yellow Version
- Pokemon Stadium
- Data Migration and Processing
- Collect all relevant data from Serebii, Bulbapedia, pkmn.help, The Cutting Room Floor, the JP Pokemon Wiki, and more.
- Transform the data into SQL-friendly CSV tables and battle-sim-friendly JSON objects.
- Data Analysis Suite
- Leverage the transformed CSVs to build charts via Bokeh.
- Expand upon findings in pseudo-academic data analysis articles.
- Host this analysis on site, and organize
- Battle Simulator
- Construct JSON objects of potential Pokemon parties and loadouts.
- Build respectful battle simulator implemented for a specific game or game cluster.
- Conduct extensive simulation to create more data about optimal combat matchups.
- Implement specific trainer data to discover optimal strategies.
- Add findings and more charts to Data Analysis Suite, and link pages to the corresponding games.
- Party Optimizer
- Leveraging matchup data, create tool to generate ideal parties based on opponent.
- Implement modes such as "required mons" (such as a favorite or starter) to personalize the tool to the user.
- Attempt to disrupt the competitive metagame
- Use all available tools and data to simulate a developing Smogon competitive metagame, including Mon relegation and promotion over epochs.
- Potentially discover something cool and/or useful for the competitive scene.
- Repeat steps for the next game