Metrics for rating albums
goverat's thread on the top albums of 2007 had me think about the sort of metrics that can be used to rate how much you enjoy an album.
I tend to think that an album (as opposed to the songs on the album) is good if you listen to it over and over. You have only listened to an album as often as you've listened to the song with the lowest play-count. With many play-counts comes the implication that the album was worth listening to that many times.
I consider an album poor if I am only willing to listen to a few songs on that album (no matter how much I like those few songs), so this is my reason for choosing the minimum play-count of a song on an album as my metric of choice.
This is not a perfect metric, however. A long album may be listened to less than a short album for obvious reasons. And if you find many good albums in a short period, this may also decrease the apparent worth of an album based only on play-count. It's difficult to deal with the second problem, but the first might be addressed by multiplying the minimum play-count with the album-length.
Here are the results of my analysis of my own library (only albums purchased in the past two years since I began using iTunes as my only music-player):
1. The Most Serene Republic - Phages (2006) (m.p.c. = 100)
2. Wolf Parade - Apologies to the Queen Mary (2005) (m.p.c = 78)
3. The Most Serene Republic - Population (2007) (m.p.c. = 68)
4. Broken Social Scene - Broken Social Scene (2005) (m.p.c. = 67)
5. Bloc Party - Silent Alarm (2005) (m.p.c. = 54)
Or, applying time-scaling to adjust for the lengths of albums (i.e. minimum play-count * album-length)
1. Broken Social Scene - Broken Social Scene (2005) (70.48 hours)
2. Wolf Parade - Apologies to the Queen Mary (2005) (62.18 hours)
3. The Most Serene Republic - Population (2007) (61.01 hours)
4. The Most Serene Republic - Phages (2006) (55.08 hours)
5. Bloc Party - Silent Alarm (2005) (48.59 hours)
I'm not particularly surprised by my top 5. If I had began using iTunes six months earlier, Metric's Live it Out would knock Bloc Party out of my top 5, but these are certainly my favourite albums.
Are there better metrics for gauging the quality of an album? What are your top 5 by either of my metrics?
I tend to think that an album (as opposed to the songs on the album) is good if you listen to it over and over. You have only listened to an album as often as you've listened to the song with the lowest play-count. With many play-counts comes the implication that the album was worth listening to that many times.
I consider an album poor if I am only willing to listen to a few songs on that album (no matter how much I like those few songs), so this is my reason for choosing the minimum play-count of a song on an album as my metric of choice.
This is not a perfect metric, however. A long album may be listened to less than a short album for obvious reasons. And if you find many good albums in a short period, this may also decrease the apparent worth of an album based only on play-count. It's difficult to deal with the second problem, but the first might be addressed by multiplying the minimum play-count with the album-length.
Here are the results of my analysis of my own library (only albums purchased in the past two years since I began using iTunes as my only music-player):
1. The Most Serene Republic - Phages (2006) (m.p.c. = 100)
2. Wolf Parade - Apologies to the Queen Mary (2005) (m.p.c = 78)
3. The Most Serene Republic - Population (2007) (m.p.c. = 68)
4. Broken Social Scene - Broken Social Scene (2005) (m.p.c. = 67)
5. Bloc Party - Silent Alarm (2005) (m.p.c. = 54)
Or, applying time-scaling to adjust for the lengths of albums (i.e. minimum play-count * album-length)
1. Broken Social Scene - Broken Social Scene (2005) (70.48 hours)
2. Wolf Parade - Apologies to the Queen Mary (2005) (62.18 hours)
3. The Most Serene Republic - Population (2007) (61.01 hours)
4. The Most Serene Republic - Phages (2006) (55.08 hours)
5. Bloc Party - Silent Alarm (2005) (48.59 hours)
I'm not particularly surprised by my top 5. If I had began using iTunes six months earlier, Metric's Live it Out would knock Bloc Party out of my top 5, but these are certainly my favourite albums.
Are there better metrics for gauging the quality of an album? What are your top 5 by either of my metrics?
Comments
I like chocolate more than turkey, but guess which one I eat more of.
If that is the standard then Chromeo & Interpol put out my two favorite albums this year.
I like chocolate more than turkey, but guess which one I eat more of.
Sounds more like denial to me. And I'm not sure the Turkey/Chocolate thing is a good comparison. Perhaps if Arcade Fire cost more, made you fat, and rotted your teeth, I might understand why you were listening to Interpol more often.