The main question that motivated me to look at the ages musicians die was whether 27 is a particularly dangerous age for musicians. Several well-known ones like Brian Jones, Jimi Hendrix, Janis Joplin, Jim Morrison, Kurt Cobain and more recently Amy Winehouse joined the so-called Forever 27 Club ending their lives.
Looking at the line-chart for all musicians above you can see a small spike at 27, but there are other ages with more prominent spikes and 27 doesn't look so special.
But there has to be something about that number, right? Maybe rock musicians, those who play the guitar or those who commit suicide are more likely to die at 27.
Looking at these individual types of musicians, it turns out that 27 is more probable than in the overall distribution, but no matter which of the above types you look at, there is always another age at which more musicians died.
The data is a subset of the Freebase Musical Artist collection, where dates of birth and obviously death are known. In addition I queried gender, musical genres, instruments played, and causes of death as optional columns resulting in a total of 12,114 data records being analyzed.
I limited the types in the musical genres, instruments played, and causes of death categories to the 10 most common ones each, otherwise there would have been too many choices with low death tolls.
Too download and process the data I wrote a Python script making use of the google-api-python-client and pandas libraries, the latter being used to calculate statistical measures like mean, median, standard deviation and variance for the individual data series.
The line charts are rendered with Google Chart Tools. You can select up tp 5 types to display at the same time. The more you choose the harder it gets to recognize and compare the distributions.
To learn who died when, you can click on a data point within the chart and the corresponding musicians' names will be displayed in a box below the chart.
The data above doesn't really justify the special perception of 27, does it? I guess it still feels special, because some members of the 27 club were very influential and became icons for whole generations.
Moreover, a death at an early age is less expected and thus attracts more attention than when someone dies at, say, 74, which is the most frequent age of death in this data set.
What do you think makes the 27 club special?
Ramiro is a developer who likes open source, data mining, visualization, and writing. To learn more about him and this site, see the about page.data analysis visualization music death