What did Rock Radio play? An explanation.

This post was supposed to give a bit more analysis of the old Rock Radio data but as I wrote it, I discovered a good bit of explanation was necessary first...

In my original post, I noted the number of tracks played, the number of artists that featured and how often the different artists were featured across the Breakfast and Drivetime shows over a period of seven days. (Although I missed one of the Breakfast shows, so I only have the data from 13 shows.) The next thing I wanted to do, but never got around to, was to look at the genres of music that featured. But, before I can look at genres, I need to make a couple of disclaimers.

First, I realised after collecting the data that not everything that is played on the radio features on the website. It seems that if the DJ played something that was not in the station's database of songs, it didn't feature on the "what was played" page. Sometimes the webpage claimed that the same song was played twice in a row and my guess is, that when a track was played that was not in the database, the system just reported the previous track had just been played again. It looks like Real XS does something similar - it has solved the repeated track reporting problem but there are still gaps when a DJ goes "off piste". For example, on the last Friday before Christmas, I submitted a trio of festive songs for Tom's Godfather of Rock feature (or Godfather-Christmas in this case). I asked for The Black Crowes - Back Door Santa; The Darkness - Christmas Time; and Black Stone Cherry - Santa Is Back; but as you can see from the screen dump, Back Door Santa is conspicuous by his absence!

So any analysis of what the station played will be incomplete because I am relying on what their website says they played rather than listening to the output and recording what they actually played.

From Wikipedia entry

The second disclaimer relates the the way I have decide on the genres of music played. Picking a genre is, inevitably going to be subjective, so I tried to assign a musical style in a way that was as fair and transparent as possible. But the main problem was that I had data on 361 different songs. I quickly decided that I had neither the time nor the inclination to go through 361 songs and assign a musical genre to each one! Since there were only 185 artists ("only 185" - ha!) I decided to assign a genre to each artist and then assume that each track of theirs that was played was an example of that genre. Clearly this is daft, for example I have classified Kiss as "Hard rock" (for reasons I'll explain in a moment) but had the misfortune to hear the abomination that is I Was Made For Loving You on Tom's show the other day, which couldn't be considered Hard rock by any stretch of the imagination! I hope though that reporting on genre by artist will give a reasonable view of what the station played since it is as good as I can manage in the time available.

The last piece of the puzzle therefore is how did I allocate genres to artists. The problem is, of course, that if you got five Rock fans in a room, you are likely to have seven different opinion on how to categorise any given artist. Also, some artists defy easy categorisation and over the course of their career, or even the course of one album, they can range across umpteen genres. I decided therefore to rely on the wisdom of crowds and turned to Wikipedia for help. Pretty much every artist listed in Wikipedia has a number of genres associated with that artist listed in a sidebox along with basic information such as when and where they were formed; years active; and past and present members. I decided to allocate the genre listed first in this sidebox as the genre for the artist. So, to take Rush as an example, I would probably have categorised Rush as "Progressive rock" but accept that they are a band that are difficult to pin down to one style. As you can see from the screen shot here though, the sidebar in Wikipedia, while listing "progressive" as one genre, puts "Hard rock" first in the list. In this instance, the list happens to be in alphabetical order but this is not always the case. I cannot find any information to explain the convention in the order of listing genres but have assumed that the first listed is the one that is considered by the Wikipedia community, to best categorise the artist and in general, on the face of it this seems to be a reasonable assumption to make. There are one or two exceptions I made to this method. If an artist does not have a Wikipedia entry (and the overwhelming majority do) I chose a genre based on information either from the artist's own website or (preferably) a site such as AllMusic. The other main exception was where the first stated musical genre in the artist's Wikipedia entry would have left them in a category of one. In those instances (again this applied to fewer than ten artists) I chose the next listed genre instead.

I hope this makes sense and that you don't disagree too violently with the approach I have adopted but let me know what you think. And now, after this long attempt to justify myself, I hope to actually post some new analysis of the old Rock Radio data in the very near future.