All current existing YouTube videos can be filtered by a single 11-character id. Does that really seem like enough?

Have you ever wondered how the many, many, MANY YouTube videos that apparently exist can really be filtered by such a small id? It doesnt seem nearly enough, especially when we live in a world with systems that use alphanumeric ids of 16, 20, 24, and more. Of course, my grasp of big numbers is limited and deplorable, so I did a little more digging.

According to some brief research it seems like YouTube (as of sometime last year) has about ~15 billion videos, with upload rates each day in the millions. The id that each video uses is an 11-character combination of three character sets: A-Z (26), a-z (26), and [0-9] (10), which makes the number of choices for each character 62. Remembering my high school math lead me to a permutation to figure out:

nPr = n! / (n - r)! -> 62P11 = 62! / (62 - 11)!

This results in about 20 quintillion (i.e. 20 with 18 zeroes) possibilities for a YouTube video ID. The current data set is nowhere near hitting that number. Thus, an 11-character ID really can easily filter all those videos.

This thought experiment has now concluded, but go read this BBC article which gets into it more.