The Age is running a story today headed Asteroid Heads for Earth. Which sounds fairly scary I guess. The article then says that the best estimate is that it has a 1 in 909,000 chance of hitting the earth. I guess Asteroid might be heading for earth, like you might win the lottery this week was too long to fit above the story.
But how about “Asteroid 15 times more likely to hit earth than you are to win the lottery this week!” ?
(Which is roughly right, at least for the UK’s national lottery).
Msnbc.com reports that after further review, the probability is 1 in 2.2 million. Funny thing is they headlined their story about this change: “Asteroid strike ruled out for 2014” - I’m guessing that their threshold for declaring something “ruled out” is 1 in a million.
Well, heading towards something isn’t the same as colliding with it. When I say I’m heading to grandma’s house, that doesn’t mean I’m going to necessarily run into and destroy it.
It seems highly unlikely we’re going to be wiped out by this particular asteroid come 2014, but not as unlikely as we might wish it to be. After all, even using MSNBC’s 1:2.2 million forecast, one could say that odds like that pay off 4 times a day in New York City.
Well yes, but we are concerned with 8-9m New Yorkers, and so the comparison would only be relevant if there were 8-9m earths we were worried about. In which case 4 would be zapped. But as we are worrying about only one earth, it is like saying i’m worried about a 1 in 2.2m chance thing happening to a Mr G.A.XYZ from New York, i.e. it’s highly unlikely.
My favourite example of low probability odds was a UK lottery statistic in the LRB, where it noted that to have a higher chance of winning the jackpot than dying before it is drawn you must buy the ticket only 3 1/2 mins beforehand. If you are 16 you get an hour, if you are 75 just 24 seconds.
Matthew has pre-empted me.
Calculating the probability of someone winning a particular lottery is straight forward, and the result — 1 to 13.98 million against in the UK’s national lottery, for example — is meaningful. There are only so many permutations of possible input, and the output will certainly be one of them.
Calculating the chances of an asteroid hitting us is far from trivial.
The “probability” in this case is a function of the uncertainty in our knowledge of the asteroid’s orbital trajectory and speed. And in the case of long-period bodies, which are the most difficult to spot and hence the most dangerous, it is also a function of our estimation of how the presence of other bodies will affect its orbit.
When I see astronomers smiling at television cameras, happily quoting precise odds on our survival, I can’t help thinking that they would be happier running Wall Street hedge funds.
While we’re on the topic of odds and perdition: you don’t need to calculate the likelihood of an asteroid collision to know that the world is going to end soon.
http://www.anthropic-principle.com/faq.html
… no, I don’t really believe in doom soon (near-flying asteroids be damned — http://web.mit.edu/www/writing/doom.html)
Jon: Actually, the phrase “ruled out” does make sense. The distinction is one that the original news story failed to make, because they were misreading the probabilities on the NASA NEO Current Impact Risks page.
These potentially hazardous asteroids often have several possible collisions with Earth. The probability listed on the front page is the cumulative probability for all possible collisions, not the probability for the earliest one, which can be seen by going to the page for the individual asteroid. The initially listed probability for 2003 QQ47’s 2014 collision with Earth was about 1 in 1.75 million; the “1 in 909,000” number cited on the wires was the cumulative probability. As you can see, now the 2014 collision’s probability is nil and it is not listed, but there are still impact possibilities for 2003 QQ47 in the late 21st century.
That NASA page is the go-to place for analysis on giant asteroid scare stories, but you have to know how to read it.
And, actually, I should qualify what I said above: by “all possible collisions” I meant “all possible collisions within the range of interest of the NASA SENTRY program”, namely the next 100 years. Beyond that, asteroid impact predictions usually get too uncertain to bother with, with some oddball exceptions.
I’m actually not sure where MSNBC’s “1 in 2.2 million” number comes from; the cumulative probability for 2003 QQ47 on the NASA site is higher now, more like 1 in 500,000, mostly because of possibilities in 2058, 2067, 2069, 2087, and 2096. Maybe it went up after the story was written.
Let’s keep in mind that, while the odds of a major asteriod strike in any given year are pretty low, when it does happen so many people will be killed, that asteroid impacts are, on average anyway, a major cause of death. Theoretically. In the long run.
Brett-
In the VERY long run. The sum of the impact probabilities (from NASA NEO Current Impact Risks) over the next 100 years is 3.28 × 10^-3, or 0.00328. Which, if the risks remain roughly consistent over the long run, means we have a 6.56% chance of impact over the next 200,000 years (I think I’m doing this math right), or roughly the same amount of time since the speciation of homo sapiens. So chances are excellent that we’ll get an impact within the next 3,051,000 years or so.
But GOSH that’s a long time! Hopefully long enough to figure out how to deal with the risk. (Hey, it’s roughly 61,000 times as long as it took from Watson and Crick’s discovery of DNA to the sequencing of the human genome, which I have to regard as a problem of similar difficulty.)
NEO uses the phrase ‘ruled out’: “The new data have enabled astronomers to refine the orbit of asteroid 2003 QQ47, and so rule out 13 of the 31 potential impacts originally listed on the JPL Current Impact Risk table. Among those ‘virtual orbits’ to be ruled out was the one for 21 March 2014, which gave the asteroid its Torino 1 rating.”
HEY, I just won the lottery! Wow, this is my lucky day, I wish everyone a happy life!
À Gauche
Jeremy Alder
Amaravati
Anggarrgoon
Audhumlan Conspiracy
H.E. Baber
Philip Blosser
Paul Broderick
Matt Brown
Diana Buccafurni
Brandon Butler
Keith Burgess-Jackson
Certain Doubts
David Chalmers
Noam Chomsky
The Conservative Philosopher
Desert Landscapes
Denis Dutton
David Efird
Karl Elliott
David Estlund
Experimental Philosophy
Fake Barn County
Kai von Fintel
Russell Arben Fox
Garden of Forking Paths
Roger Gathman
Michael Green
Scott Hagaman
Helen Habermann
David Hildebrand
John Holbo
Christopher Grau
Jonathan Ichikawa
Tom Irish
Michelle Jenkins
Adam Kotsko
Barry Lam
Language Hat
Language Log
Christian Lee
Brian Leiter
Stephen Lenhart
Clayton Littlejohn
Roderick T. Long
Joshua Macy
Mad Grad
Jonathan Martin
Matthew McGrattan
Marc Moffett
Geoffrey Nunberg
Orange Philosophy
Philosophy Carnival
Philosophy, et cetera
Philosophy of Art
Douglas Portmore
Philosophy from the 617 (moribund)
Jeremy Pierce
Punishment Theory
Geoff Pynn
Timothy Quigley (moribund?)
Conor Roddy
Sappho's Breathing
Anders Schoubye
Wolfgang Schwartz
Scribo
Michael Sevel
Tom Stoneham (moribund)
Adam Swenson
Peter Suber
Eddie Thomas
Joe Ulatowski
Bruce Umbaugh
What is the name ...
Matt Weiner
Will Wilkinson
Jessica Wilson
Young Hegelian
Richard Zach
Psychology
Donyell Coleman
Deborah Frisch
Milt Rosenberg
Tom Stafford
Law
Ann Althouse
Stephen Bainbridge
Jack Balkin
Douglass A. Berman
Francesca Bignami
BlunkettWatch
Jack Bogdanski
Paul L. Caron
Conglomerate
Jeff Cooper
Disability Law
Displacement of Concepts
Wayne Eastman
Eric Fink
Victor Fleischer (on hiatus)
Peter Friedman
Michael Froomkin
Bernard Hibbitts
Walter Hutchens
InstaPundit
Andis Kaulins
Lawmeme
Edward Lee
Karl-Friedrich Lenz
Larry Lessig
Mirror of Justice
Eric Muller
Nathan Oman
Opinio Juris
John Palfrey
Ken Parish
Punishment Theory
Larry Ribstein
The Right Coast
D. Gordon Smith
Lawrence Solum
Peter Tillers
Transatlantic Assembly
Lawrence Velvel
David Wagner
Kim Weatherall
Yale Constitution Society
Tun Yin
History
Blogenspiel
Timothy Burke
Rebunk
Naomi Chana
Chapati Mystery
Cliopatria
Juan Cole
Cranky Professor
Greg Daly
James Davila
Sherman Dorn
Michael Drout
Frog in a Well
Frogs and Ravens
Early Modern Notes
Evan Garcia
George Mason History bloggers
Ghost in the Machine
Rebecca Goetz
Invisible Adjunct (inactive)
Jason Kuznicki
Konrad Mitchell Lawson
Danny Loss
Liberty and Power
Danny Loss
Ether MacAllum Stewart
Pam Mack
Heather Mathews
James Meadway
Medieval Studies
H.D. Miller
Caleb McDaniel
Marc Mulholland
Received Ideas
Renaissance Weblog
Nathaniel Robinson
Jacob Remes (moribund?)
Christopher Sheil
Red Ted
Time Travelling Is Easy
Brian Ulrich
Shana Worthen
Computers/media/communication
Lauren Andreacchi (moribund)
Eric Behrens
Joseph Bosco
Danah Boyd
David Brake
Collin Brooke
Maximilian Dornseif (moribund)
Jeff Erickson
Ed Felten
Lance Fortnow
Louise Ferguson
Anne Galloway
Jason Gallo
Josh Greenberg
Alex Halavais
Sariel Har-Peled
Tracy Kennedy
Tim Lambert
Liz Lawley
Michael O'Foghlu
Jose Luis Orihuela (moribund)
Alex Pang
Sebastian Paquet
Fernando Pereira
Pink Bunny of Battle
Ranting Professors
Jay Rosen
Ken Rufo
Douglas Rushkoff
Vika Safrin
Rob Schaap (Blogorrhoea)
Frank Schaap
Robert A. Stewart
Suresh Venkatasubramanian
Ray Trygstad
Jill Walker
Phil Windley
Siva Vaidahyanathan
Anthropology
Kerim Friedman
Alex Golub
Martijn de Koning
Nicholas Packwood
Geography
Stentor Danielson
Benjamin Heumann
Scott Whitlock
Education
Edward Bilodeau
Jenny D.
Richard Kahn
Progressive Teachers
Kelvin Thompson (defunct?)
Mark Byron
Business administration
Michael Watkins (moribund)
Literature, language, culture
Mike Arnzen
Brandon Barr
Michael Berube
The Blogora
Colin Brayton
John Bruce
Miriam Burstein
Chris Cagle
Jean Chu
Hans Coppens
Tyler Curtain
Cultural Revolution
Terry Dean
Joseph Duemer
Flaschenpost
Kathleen Fitzpatrick
Jonathan Goodwin
Rachael Groner
Alison Hale
Household Opera
Dennis Jerz
Jason Jones
Miriam Jones
Matthew Kirschenbaum
Steven Krause
Lilliputian Lilith
Catherine Liu
John Lovas
Gerald Lucas
Making Contact
Barry Mauer
Erin O'Connor
Print Culture
Clancy Ratcliff
Matthias Rip
A.G. Rud
Amardeep Singh
Steve Shaviro
Thanks ... Zombie
Vera Tobin
Chuck Tryon
University Diaries
Classics
Michael Hendry
David Meadows
Religion
AKM Adam
Ryan Overbey
Telford Work (moribund)
Library Science
Norma Bruce
Music
Kyle Gann
ionarts
Tim Rutherford-Johnson
Greg Sandow
Scott Spiegelberg
Biology/Medicine
Pradeep Atluri
Bloviator
Anthony Cox
Susan Ferrari (moribund)
Amy Greenwood
La Di Da
John M. Lynch
Charles Murtaugh (moribund)
Paul Z. Myers
Respectful of Otters
Josh Rosenau
Universal Acid
Amity Wilczek (moribund)
Theodore Wong (moribund)
Physics/Applied Physics
Trish Amuntrud
Sean Carroll
Jacques Distler
Stephen Hsu
Irascible Professor
Andrew Jaffe
Michael Nielsen
Chad Orzel
String Coffee Table
Math/Statistics
Dead Parrots
Andrew Gelman
Christopher Genovese
Moment, Linger on
Jason Rosenhouse
Vlorbik
Peter Woit
Complex Systems
Petter Holme
Luis Rocha
Cosma Shalizi
Bill Tozier
Chemistry
"Keneth Miles"
Engineering
Zack Amjal
Chris Hall
University Administration
Frank Admissions (moribund?)
Architecture/Urban development
City Comforts (urban planning)
Unfolio
Panchromatica
Earth Sciences
Our Take
Who Knows?
Bitch Ph.D.
Just Tenured
Playing School
Professor Goose
This Academic Life
Other sources of information
Arts and Letters Daily
Boston Review
Imprints
Political Theory Daily Review
Science and Technology Daily Review