Published April 7, 2026 | Version 1

Olympic swimming analysis between 1912 and 2020

Description

This report presents a data-driven analysis of Olympic swimming results spanning over a century of competition, from the 1912 Stockholm Games to the 2020 Tokyo Olympics. The dataset, sourced from Kaggle, contains 4,359 entries covering individual and relay events across all strokes and distances.

The analysis is structured around four main themes:

1. Changes in performance over time
Long-term trends in finishing times are examined across all strokes and distances, for both men and women. A Welch two-sample t-test is applied to formally assess the impact of the FINA super-suit ban (January 2010) on performance levels. A linear regression model is also fitted to quantify the change in the pace of improvement before and after the ban. The evolution of the gender performance gap is analysed decade by decade.

2. Geopolitics of swimming — national dominance
Medal counts are analysed by country and by decade, revealing distinct eras of dominance (USA, GDR, Australia, Japan). A country-level efficiency metric (gold medals / total medals) is computed and visualised. A K-Means clustering algorithm (k=4, scikit-learn) is applied to country-decade combinations to identify typologies of national performance.

3. Individual impact and the Superstar Effect
The most decorated athletes in Olympic swimming history are identified and compared across total medals and gold medals. A dependency ratio is computed for each country, measuring the share of national medals attributable to a single athlete — a phenomenon referred to as the "Superstar Effect."

4. Conclusion and perspectives
The report concludes with a discussion of key findings, methodological limitations (World Wars, boycotts, pandemic, doping), and directions for further research, including GDP-adjusted medal analysis, peak age modelling, and the implications of Cameron McEvoy's 2026 world record (50m freestyle, 20.88s) for training methodology.

Keywords:
Olympic Games, Swimming, Data Analysis, Sports Science, Performance Trends, Gender Gap, K-Means Clustering, Superstar Effect, FINA, Python

Files

report.pdf

Files (3.7 MB)

Name Size
md5:afa2dac174c843e89d1710c2b4132787
3.7 MB Preview Download

Additional details

Dates

Accepted
2026-04-07
First version

Software

Repository URL
https://github.com/perazaf1/olympicSwimmingAnalysis
Programming language
Python
Development Status
Active