Last week, we discussed how the iPod Shuffle was the new radio. Its a straight forward thesis, hard to argue against.

Have a gander at a this meme making its rounds on the net -- evincing even more evidence that the radio business model of playing music is fading:

This week, we are discovering proof of this all over the
web, via what's been called The ShufflePlay Game.  (I put up my own version of the ShufflePlay Game over here and here).

Here's a short list of blogs posting variations of Shuffle Play:

The Republic of T.:   Music for the Masses    Scott's Place:  Shuffle play   

Matthew Yglesias: Shuffle!

essays & effluvia: Shuffle Play

The Fly Bottle:  Shuffle Game

Stumbling and Mumbling    Shuffle and see    

Crescat Sententia:   Potpourri

divine angst: my ipod's shuffler

Grammar.police: iPost

Class Maledictorian:  Terrible Taste

Knowledge Problem:   WHILE WE'RE IN A MUSIC VEIN ...

Mansfield Fox:   Know Me by My Music

Death in the Afternoon: Six Hour Exams 

aworks :: "new" american classical music: Wedge (1961). Roger Reynolds

essays & effluvia: More Shuffle Play

Cobb:   Self Indulgent Twenty

The Republic of T. gets props for being the earliest of these I've found -- dating all the back to April 2, 2004. Scott's Place gets special mention for having assembled at least 7 random shuffle collections (Kudos!).


If you know of any other shuffle play posts, please let me know, and I'll add them to the list . . .

(listed bloggers feel free to trackback to this post).

Lastly, check out Professor Booty's analysis of the Math of Random/Shuffle Play: 

Okay, let's say you have 2500 songs. To simplify, let's say that's 10 songs per album for a total of 250 albums. And we'll define "repeat" as "hearing a song from the same album you've already heard a song from". How often will you hear a repeat (by that definition)?

By the time you get to the 20th song, the odds of having heard at least one repeat by the same artists is an extremely surprising 76%!

I'd love to see that math verified . . .


UPDATE:  January 24, 2005 6:30 am

Tobias Brandt weighs in on the subject of the Math of shuffle play:

I disagree with Prof. Booty's analysis of the probability of getting a repeat album within 20 songs and my own analysis puts the odds at 54%.  (See spreadsheet used to get the result).  The spreadsheet also has the benefit of being able to account for excluding repeats of previous songs, a feature many  media players have.

Download song_repeat_probability.xls

Prof Booty:  Not sure what accounts for your mistake but generally speaking one can't just sum probabilities and it's best to  look at things like this in terms of a probability tree diagram.

I wish I could get all my questions answered so quickly and succinctly!