Every year at the Eurovision Song Contest numerous speculations arise about the hidden factors of a country's success.
Running order has become a persistent debate - does it really matter if a country performs first, last, or somewhere in between the line-up of 26 final entries?
This project explores whether performance running order in the Eurovision Song Contest final can influence the country's final place and score. It includes a combination of data exploration, correlation analysis, and linear regression modeling.
- Significant positive correlation between later running order and final points (r = 0.410, p ≈ 0.038).
- Inverse significant correlation between running order and final place (r = -0.184, p ≈ 0.014).
To load the Eurovision data, a local Flask API I built is called which processes a CSV dataset and returns a list of Eurovision entry DTOs. /entries API is implemented in a separate repo: eurovision-data-api.
- Evaluate score and place prediction accuracy of the fitted models
- Explore non-linear models
- Compare jury and televote scores and their correlation with running order
- Introduce new features to the dataset for multivariate analysis