Welcome to JoshHornik.com: Statistically Improbable Phrases

« I saw Eriq LaSalle -- 4 points | Main | Theo Epstein Genius Watch - week 2 »

April 15, 2005

Statistically Improbable Phrases

Amazon has this new feature called "Statistically Improbable Phrases". For every book that you can "search inside", they have found all the 2 or 3 word phrases that appear multiple times in the book but that you would only find very rarely in other books. In other words, it is statistically improbable that the phrase would appear so many times in a book.

The math is unimportant, and it's hard to think of a reason I would use SIP's (as Amazon calls them). But it's fun to look at them, anyway.

For example, you are more likely to see the phrase "sweater pocket", "lame saint", and "seeded womb" in the book The Da Vinci Code than in most other books. (Probably not all in the same sentence, though, I imagine.)

Stephen King, in the book Pet Sematary, must have enjoyed using the phrase "fucking cat". He wrote it 4 times, which is twice as many times as the 2nd place book in "fucking cat" references ("My Soul to Keep" by Tananarive Due).

It got me thinking that a fun game would be trying to guess the book based on the SIP's. So here goes. I'll put the phrase(s) -- you guess the book. For example, if the phrase is "giant peach", the answer is -- obviously, James and the Giant Peach. Click the phrase to see the answer.

1. "Psychedelic movement", "hip world", "acid heads", and "donkey beads"
2. "Hands for the conch"
3. "His tomahawk pipe", "pagan harpooneers", and "white hump"
4. "Objective sensory stimuli", "involuntary ideas", "psychical sources" and "unpleasurable dreams"
5. "Silly noises", "pantomime horses" and "nude organist"
6. "He ejaculated" (Sorry, that is an inside joke that I suspect even the one person it's aimed at may not get.)
7. Guess the subject: "first electric set", "album version" and "harmonica solo"
8. "Poor orphan child" and "bethought myself"
9. This one's a play: "ass head"
10. I wish I could say I would have known this one: "pensive bosom", "absentminded beggar", "seaside girls" and "tooraloom tooraloom tooraloom"