Hacker News Reader

Comment by hbn | Hacker News Reader

I remember hearing Apple had to make a "random-seeming to humans" algorithm with the iPod's shuffle feature as well for the same reason. Grabbing a truly random song every play doesn't feel random to humans. What people really want with their song shuffle is something new they haven't played in a while.

View on HN

I don't have many memories of experiencing this with my own iPod, but I recently came across this paper - "Does Your iPod Really Play Favorites?" (https://www.researchgate.net/profile/Jessica-Culhane/publica...).

In the paper, the authors examined evidence of nonrandom behavior and created several probability models for these events under the assumption of a random shuffle. Their conclusion was:

> Much of the evidence of nonrandom behavior reported by Steven Levy and others does not hold up when the probability models of the events are determined. Our results show the probability models for a random shuffle in many cases do not match the intuition of users. In addition, our statistical tests show the long-term occurrences of these events are within expectations under the assumption of a random shuffle.

I'm not sure if this was something Apple implemented after 2009.

I think it's gone now, but older versions of iTunes had a "randomness" slider that ranged from truly random to a biased randomness that "felt" more random by avoiding recently chosen songs/albums

I remember this, and I also remember reading anecdotes online about people listening on shuffle, hearing two songs from the same artist or album in a row, and getting superstitious about it like it was trying to "warn them" something bad was about to happen.

So Apple had to take their generally true random algorithm and make it intentionally less random so it worked like people implicitly expected it to. Makes me wonder how many things we really are "holding wrong" because dev expectations didn't line up with how people actually use something.

Random is random, shuffle is shuffle. They are different things.

"Random" doesn't necessarily mean "uniformly distributed random." Shuffling a deck of cards randomly is random, but you're not going to get the same card twice in a row, or even the same card twice until you go through the whole deck.

Do you re-shuffle after the deck is exhausted, or after every hand? It's not excessively unusual to be dealt the same card twice in two consecutive hands.

Who said anything about hands? I'm trying to make an analogy between shuffling a deck and shuffling xkcd comics. I thought having a visual, physical analogue would perhaps help in seeing that randomness still exists in a shuffle.

Sorry, the point I was trying to make was "shuffle" could possibly mean "re-shuffle after every track" - so you could still get the same track twice in a row.

(It was a bad way of stating it, that's on me)

Shuffling the whole playlist to choose a new song every time is no different than choosing uniformly from all songs every time. What would be the point? I can't tell what you are getting at.

My point is that, for a not-unreasonable interpretation of "shuffle", it is the same as "random".

nvm

And a uniform random shuffle chooses any of the N! orderings with equal probability. One is looking at a single element as the output of the process, and another is looking at the entire list as the output of the process.

Also, there are plenty of non-uniform random distributions.

Grabbing a new song with probabilities weights based on past outcomes is just as "truly random" as picking a new song with a uniform distribution on all songs every time. https://en.wikipedia.org/wiki/Stochastic_process

> What people really want with their song shuffle is something new they haven't played in a while. Speak for yourself

I mean I'd agree with OP, shuffle should take all songs, put them in random order and play them start to finish without any of them repeating. If I wanted repeat then I'd turn on repeat.

People are pattern seeking, even in "random" data.

Reply on news.ycombinator.com