My yearly effort to collate the album of the year lists from music publications into a single spreadsheet. It started as a good way to catch albums I missed over the past year, and to have all the lists in one place for easy browsing. One could also use it to make some arguments semi-supported by numbers about music criticism and trends and so forth, if one wished.
It's mostly data entry. When a publication posts a list, I put the top 50 in the spreadsheet. I only choose lists that meet the following criteria:
- A list voted on by a staff, not a single person.
- A list that covers multiple genres, not just metal, rap, etc.
- At least 50 albums are listed (though I only take the top 50).
- It's a website or magazine that I've heard of (or it looks legitimate).
I just list albums 1-50 according to their rank on the list, no 50 points for 1st, 49 for 2nd funny business. So low numbers are good, like golf. If a list is cowardly and unranked (looking at you, NPR), I just give every album listed 25.
The columns to the far right are some simple stats, created with valuable input from Ben Greenfield and Alessandro Panella, for measuring consensus across lists. For the first three, as mentioned above, lower is better. First is just the average ranking the album received. But that's distorted by the fact that most albums don't appear on every list. So the second column is a "weighted" average, with the average ranking divided by the number of lists on which it appears. But some albums place very highly on only one list, so that's biased too. Thus, the "Consensus Score" which penalizes an album for every list it doesn't appear on (by assigning a pseudo-rank of 75) and then taking the average ranking, divided by total number of lists. Typically, I sort the list by this final measure.
The last two columns are self-explanatory: The number of lists an album appears on, and the number of lists where it placed in the top ten.
Sure! At the most basic level, if you notice a mistake, shout it out...maintaining the list is a lot of typing. If you know some fancy statistics we can run on the data, I'd love to hear it. If you want to play with the numbers yourself and can't get them off of Google Sheets for some reason, let me know. Most especially, I'm always looking for interesting ways to visualize the data, so if you can do that, please contact me. @robmitchum or rdmitchum at gmail. Thanks!