Predicting Bad Bunny's Setlist

Overview

We're embarking on an ambitious project to predict Bad Bunny's Super Bowl setlist by leveraging machine learning and data from popular music APIs. Our approach will involve collecting comprehensive data on song characteristics, historical setlist patterns, and popularity metrics to build a predictive model.

Our Plan

1. Data Collection

We'll utilize two key APIs to construct our dataset:

• Spotify API: Access to track metadata, popularity scores, and audio features (tempo, energy, danceability, etc.)
• getSongBPM API: Precise BPM (beats per minute) data for songs in Bad Bunny's discography

You can explore the getSongBPM API documentation here: https://getsongbpm.com/api

2. Feature Engineering

From the collected data, we'll extract and engineer relevant features:

• Song popularity and streaming metrics
• Audio characteristics (BPM, energy, danceability, acousticness)
• Historical setlist frequency
• Album and release date information
• Song duration and key signatures

3. Model Development

Our prediction model will:

• Analyze patterns in historical setlists across multiple concerts
• Weight songs based on recency, popularity, and tour themes
• Consider venue size and concert type
• Predict the probability of each song appearing in a setlist

4. Validation & Iteration

We'll validate our model against known setlists and continuously refine it as we gather more data.

Next Steps

1. Set up connections to the Spotify and getSongBPM APIs
2. Build a web scraper for historical setlist data
3. Preprocess and clean the collected data
4. Develop and train the prediction model
5. Create a user interface to visualize predictions

This project combines music data science with machine learning to answer an interesting question: Can we predict what songs Bad Bunny will perform at his next concert?