Games and personalities at the Olympics created a major social media buzz this year. This is well-expected from the biggest quadrennial sports events. Fans heartily indulged themselves in some very voluble discussions that could yield a “gold” mine of knowledge for the data scientist – Which were the most popular hashtags? Who were the most popular athletes over the world? Which were the most talked about events? Which was the most talked about nation?
The Flytxt data science team analysed these discussions, from public Facebook and Twitter pages related to the event, to understand the overall sentiment and discover new insights. While some of the insights follow human intuition, many of them are fascinating and totally unexpected.
A lot of big hits; a few little misses
Surprisingly before the event, the fastest man of earth, Usain Bolt was not even among the top three most popular athletes. Katie Ledecky was not even mentioned among the top medal aspirants. Swimmers as expected were the most followed ones. Gabby Douglas – the first African-American woman in history to win an individual all-around gold medal – started out as an athlete with the most positive sentiment. However, her popularity chart dipped because of the US anthem controversy.
Most mentioned male athletes:
- American Swimmer Michael Phelps
- Jamaican Sprinter Usain Bolt
- American Swimmer Ryan Lochte
- British Tennis Player Andy Murray
Most mentioned female athletes:
- Gymnast Simone Biles
- Gymnast Gabby Douglas
- Gymnast Aly Raisman
- Swimmer Katie Ledecky
More female athletes earned positive sentiment than male athletes
Michael Phelps invoked the most positive sentiment in tweets at 43 %, being the most decorated athlete at the games. While Simone Biles and Katie Ledecky had the same medal tally, the overall positive sentiment for Biles was more (26 %) than Katie (6 %). Other top scorers included Usain Bolt (19 %), Gabby Douglas (16 %) and Ryan Lochte (6.5 %) followed by Americans Gymnast Laurie Hernandez (5.5 %), Beach volleyball player Kerri Walsh Jennings (5 %) and Sprinter Allyson Felix (4.5 %). However, on the whole, the top scorers included more females than males.
By looking at the results of the analysis, the athletes were related to specific adjectives.
- The awesome Michael Phelps
- The fastest Usain Bolt
- The talented Simone Biles
- The great Gabby Douglas
Most popular athletes in not so popular events
Gymnastics beat football which happens to be the host country Brazil’s national sport. Track surpassed swimming which happened to have the top 2 most popular athletes in its kitty. Other popular sports included basketball, tennis, beach volleyball, rugby and golf
Most talked about nations
Post event, the popularity chart of the winning nations swung greatly. In the beginning, Britain was nowhere among the most talked about nation at the Rio Olympics 2016. It marched on the 2nd spot, beating the usual favorite China and even surpassed Brazil in terms of popularity, despite the host country advantage.
Methodology: Social media activity data before the Olympics kick off (5th August) and after the games (21st August) was analysed. In this article, top 100 items have been visualized as word clouds.
The data collected was filtered and tokenized. The process includes sentences being converted to words and removal of irrelevant stop words. The number of occurrences was the criteria to find the popular item (such as sportsperson or event) in the games. The sentiment analysis comprised use of a dictionary of over 4000 words that included positive and negative terms.
In a nutshell, the gold winners were not exactly a hit on the social media. There were many other players and nations which were riding the Olympic wave of popularity. Nevertheless, the discussions around the world’s biggest sporting event were fascinating.
Try deep learning using MATLAB