Top 1000 Twitch Users — Exploratory Data Analysis

Topic: Twitch is the world’s leading live streaming platform for gamers and other creative content.

Dataset: I used a dataset with the top 1000 Twitch streamers in the past year based on total watch time in minutes. The dataset included whether the streamer was partnered with Twitch, whether the streamer’s content was for mature audiences only (18+), and the streamer’s language. In addition, the dataset included other useful information such as follower count, followers gained, stream time, and views gained.

Guiding Questions:

• How does language affect a Twitch streamer’s performance? Are English speakers more popular than non-English speakers?
• Does maturity, or content that is for adult audiences only, reduce the success of Twitch streamers?
• How does a partnership with Twitch affect the performance of twitch streamers?

First, I explored the effects of language on Twitch streamer’s performance.

Language Count of the Top 1000 Twitch Streamers:

In this visualization, I took a count of each language that the top Twitch streamers speak. As the graph shows, there are overwhelmingly more English speakers in the top 1000 list of Twitch Streamers.

Amount of Followers Per Language:

In this visualization, I saw that the streamers with most followers tend to be English speaking. However, Portuguese and Spanish speaking streamers have some higher follower counts.

Through analysis of the visuals, it is clear the the top Twitch users mostly speak English. Another question that could be explored later would be what countries do the top Twitch streamers come from.

After analyzing the effects of language, I looked at how maturity affected Twitch streamer performance. Mature content in this context means that streams are only for individuals 18 years old and older.

Maturity Count:

This visualization shows that the top 1000 Twitch streamers are mostly non-mature audience accounts.

Followers Gained By Maturity:

In terms of followers gained in the past year, streamers without mature audiences tend to have gained more followers than streamers with mature audiences only. This could be due to more adolescents being able to watch streamers without maturity restrictions.

After analyzing the effects of maturity, I analyzed the effects of whether a streamer is partnered with Twitch or not. Before creating visualizations, my hypothesis was that streamers with Twitch partnerships were more successful than ones without a partnership.

Count of Partnered and Non-Partnered Streamers:

Out of the top 1000 Twitch streamers, nearly all of them are partnered with Twitch. This shows that a partnership with Twitch can boost a steamer’s watch time and overall success.

Views Gained Per Followers Gained (Partnered vs. Not Partnered):

Although partnered streamers have gained more followers in the past year, views gained has mostly been the same for both partnered and un-partnered streamers. So while partnerships help streamers become more successful in terms of followers, it does not affect views gained.

Lastly, I wanted to study the correlation between multiple variables relating to the top 1000 Twitch streamers.

Correlation Heat Map:

This heat map brings attention to many factors that correlate with each other. Some interesting insights that I found in this map is that watch time is positively correlated with followers and views gained. The longer that people stream a streamer’s content, the streamer will most likely get more followers and total views. Additionally, stream time seems to have little to no effect on other factors.