Hi everyone! I got sick of everyone complaining about Spotify's shuffle function not being random and not being sure myself so I did a simple test to see how it worked. I basically just kept reshuffling the "Up Next" of an approximately 1000 track playlist, adding that to a new playlist many times, and seeing how many tracks were new/repeated between them.
Note: Original playlist was 1157 tracks long, and "Up Next" is 80 tracks long. The tracks that remained didn't have anything in common, but I did notice they tended to huddle in clusters separated by on average 10 tracks.
To be honest, I don't know enough about math or computer science to know whether this is a good result or a bad result, but personally I'd prefer it if each pass had completely new tracks until the playlist was exhausted.
It would help if others could perform similar tests on Spotify and other services, so here's the procedure:
1. Find a playlist with a lot of tracks.
2. Shuffle play it
3. Go to Queue
4. Shift click "Up Next" and add it to new playlist.
5. Turn shuffle off and on again to reshuffle.
6. Repeat step 4 and click "Skip Duplicates"
7. Go to new playlist and subtract the new track total from the old track total to get the amount of new tracks (80 on the second pass)
8. Record this, the new track total, and any other data you feel is relevant.
9. Repeat steps 6-8 until the new playlist track total matches the source playlist track total or until bored.
If you're an OCD weirdo like me and your source playlist was sorted by Artist, Album, etc. you can follow these to search for a non-existant pattern in the excluded tracks.
A. Do steps 1-11
B. Shift-click the entire source playlist and add it to new playlist (click "Skip Duplicates")
C. In new playlist, sort by recently added, then shift-click the new arrivals to highlight them.
D. Sort new playlist in the same way as the source playlist. The excluded tracks will be highlighted.