Introducing: The Book of Odds

What’s more dangerous: a playground jungle gym or your office chair? As it happens, one in every 3,759 fatal accidental falls starts from a piece of playground equipment. You’re 85 times more likely, meanwhile, to fall to your death from a chair. That’s one of the many odd pairings waiting to be discovered in The Book of Odds, an online statistical encyclopedia launching tomorrow.

Some other probabilistic tidbits I found digging through the site:

  • The average American is more likely to live without ever visiting a dentist than to live without a TV in her home.
  • A married man is about as likely to cheat on his wife as he is to experience a flight delay.
  • You’re eight times more likely to have your ashes abandoned at a crematorium than to see a new book on personal finance be published in a given year.

The Book of Odds is a searchable online database of “odds statements,” the probabilities of everyday life. You can search it by keyword or by the odds themselves — for instance, how many things stand a 1 in 142 chance of happening to to you. As a special treat for Freakonomics readers, you can try the beta version of the site by clicking here and entering the username “brownian” and password “motion.”

The site’s founder, Amram Shapiro, says he wants to create a reference tool for better understanding the endless stream of odds that confront us:

For people to understand a probability, they have to find something they can relate to. Our ambition is that anyone should be able to look up any probability and find something in the list of equal probabilities which they’ve personally experienced and therefore can relate to their own empirical sense.

In the midst of this kind of gigantic outpouring of data that surrounds us, there are a lot of interesting things to be found if you only look for them. If you ask an interesting enough question, you have a reasonably good chance of finding data to help you make some sense of the answer. To me that produces a kind of optimism. It means we’re not just drowning in data.

Sifting through the Book of Odds to find intriguing pairs of odds statements (say, the likelihood of drowning in a pool versus being killed by a handgun) can be fun exercise. Shapiro likens it to punning or metaphorical poetry: “You find a subterranean relationship between two unlike things you had never thought of before.”

We hope you’ll share your favorite finds with us in the comments here. It might even spark some interesting research. Shapiro says that’s his secret desire: “I would love someday to discover there were a whole lot of PhD theses triggered just by somebody wandering around Book of Odds and saying ‘wait a minute, that gives me an idea … .'”


Damon

"A married man is about as likely to cheat on his wife as he is to experience a flight delay."

Is this the joint probability that a married man will go on a flight AND experience a delay.?

Last I checked the conditional probality of experiencing a delay while flying was 100%, or maybe it just seems that way in Atlanta,

Ian Stanczyk

Ryan,

Many thanks for the write-up. I hope you all enjoy the early look.

Best,
Ian Stanczyk
Book of Odds

Alan

Saw this guy speak and demo at a Web Innovator's Conference in Cambridge, MA. Really interesting site platform!

Spooner

Damon (comment #1) points out a common problem with most "odds" statements: ambiguity. And what does "85 times more likely" mean? If the odds of some event are 2%, does an event 85 times more likely have a probability of 172%? Clearly not, so just what does it mean? I hope the book states things more clearly.

Statler

You should really specify if you mean experience flight delay on a given flight or ever experience it.

Bill

Commenter Damon, I'm guessing that there is a significant number of married men who have never/will never fly, bringing the numbers of married men experiencing flight delays below 100%.

Snide

@Damon, Perhaps it is because there is a reasonable percentage of married men who never have flown, and thus have never experienced a flight delay.

I'm curious about the personal finance book tidbit. Am I to read that as "not read a new book on personal finance?" If so, I'm more inclined to believe that one then.

Christian

What's more dangerous: a playground jungle gym, your office chair, or your interpretation of these factoids?

Good grief, you can't judge which of two activities is more dangerous with only the numerator. You have to a have a denominator. That is, how many times or how many hours are spent doing each activity.

By your logic jumping out an airplane without a parachute is much safer than siting in an office chair because it accounts for a very tiny number of fatal accidental falls!

KarenS

"What's more dangerous:"

"You're 85 times more likely, meanwhile, to fall to your death from a chair. "

For some reason, the link doesn't work for me, but I'll assume they didn't control for time spent in a chair vs time spent on playground equipment. I'll bet my next paycheck that most people spend more than 85X more time in a chair than they do on playground equipment. Thus, it should be "more dangerous" to spend an hour on playground equipment than it is to spend an hour sitting on a chair.

I think a similar disconnect happens when we talk about the safety of traveling by plane vs car. Statistics show it's safer to fly, which (no pun intended) flies in the face of our intuition, that flying is more difficult than driving, and thus more dangerous. The reason flying is statistically safer than driving is because of the conditions in which we fly--the planes are piloted by highly trained professionals, strictly monitored by professionals on the ground, and are (in most cases) rigorously inspected and maintained. Most cars are driven by bleary-eyed, minimally trained average Joes and Marys, who are operating their cars while talking on their phones and drinking coffee.

If these average Joes and Marys flew in unregulated air space with minimal training, as well as several distractions, people would be falling out of the sky like rain in a monsoon.

Read more...

Eric

Aw come on. I'm sure the average person spends a lot more than 85x the amount of time in a chair as on a playground. So, even if you're 85x more likely to die from a chair fall than a playground one, it could still be that playgrounds are (a lot) more dangerous than chairs.

(Plus, what about non-fatal injuries?)

In other words--one perhaps unintended consequence of using this site might be people abusing or misinterpreting statistics.

Dan

They must mean the probability of cheating PER MARRIAGE versus the probability of a delay PER FLIGHT. Gotta be, right? Otherwise, that's a lot of airport-lounge rendezvous.

frankenduf

i don't see how this applies- drowning odds vary according to swimming ability, not averages

Damon

@ Bill & Snide,

For the general population who do not understand the difference between Joint and Conditional probablities, stats like these can be dangerous. I seriously doubt the adultery rate of married men is anywhere near 100%.

Perhaps my tounge in cheek jab at Delta's lock on Atlanta and their poor on time performance didn't come off very well. Afterall, DELTA is an acronym for Don't expect to leave the airport.

Ken

@Damon,

I took that to mean that a married man, throughout his entire marriage, is as likely to cheat once as he is to experience a flight delay *in any given flight*.

That seems a lot more reasonable to me than the explanation that some men never/rarely fly.

Mark

Here's my favorite tidbit so far, just for fun:

1 in 115,300
The odds a person will visit an emergency department due to an accident involving a pogo stick in a year are 1 in 115,300 (US, 2007).

nobody.really

“A married man is about as likely to cheat on his wife as he is to experience a flight delay.”

Last I checked the conditional probability of experiencing a delay while flying was 100%, or maybe it just seems that way in Atlanta.

What does that tell us about the state of marital fidelity in Atlanta? Hm.....

Jonathon K.

I'm amazed at the amount of odds that are just about 1:1.

Kevin

Aren't many many books on personal finance published every year? That would make the chance of having ashes abandoned 800%...

Snide

@Damon,

I got your joke and enjoyed it. I just didn't think you were exaggerating THAT much, so I took some seriousness to it. :)

d

"The odds a fatal firearms accident will occur to a white man 20 - 54 are 1 in 2.47 (US, 1999 - 2005)"

This also seems misleading. As it reads, I take home that 40% of all white males 20-54 were killed by guns in the span 1999-2005. It should more likely read that 40% of fatalities in that demographic were firearms-related. The overall probability of death is left out of the odds ratio.