Recently the blog Slate Star Codex debunked a recent Buzzfeed article about the chances of a man being falsely accused of rape during his lifetime. Our Skeptic Ink Network sister blog Hume’s Apprentice provided the relevant link.
The post is a “listicle” ticking off things much more likely to happen in one’s lifetime than being falsely accused of rape. One of them is “getting killed by an asteroid or comet,” which it says is as low as one in 3,000 or as high as one in 250,000 over a lifetime.
That seemed really high to me, since, in the entire recorded history of humanity, no one has been killed by a meteorite or asteroid. Not one person. Last year’s meteor shower in Chelyabinsk, Russia, which caused citywide chaos, didn’t kill anyone. The 1908 explosion over Tunguska, in Siberia, also didn’t kill any humans, although millions of trees and probably thousands of deer and other wildlife met their maker.
The closest we’ve ever come to a direct fatality was Ann Hodges, a Sylacauga, Alabama, resident, who in November 1954 was struck by a meteor on her leg and hip after it (the meteor) crashed through her ceiling, bounced off her radio, and hit her where she lay, napping, on her couch. She got a nasty bruise, but recovered.
The Buzzfeed source is this paper from a Tulane professor named Nelson. He bases his figures on the known periodicity of impacts, as seen in the geologic record; the estimated number of current Earth-crossing objects in the solar system; and the rate of near misses since we’ve had the technology to keep track of such things.
There’s a chart on that page which ranks the odds of dying of various causes, including motor vehicle accidents, murder, drowning, airplane crashes, earthquakes, and food poisoning. Strangely, the same chart also includes the odds of winning the Powerball lottery, which I wouldn’t say is a risk.
Winning the lottery is the least likely thing on the list, with a “risk” of winning at 1 in 195,249,054.
Phil Plait, the Bad Astronomer, took a look at this back in 2008. He comes up with odds of one in 700,000. I’m not sure that’s equal to Nelson’s one in 3,000-250,000 over a lifetime or not; it doesn’t look right.
I don’t know whose calculations are better (Phil Plait borrowed his from astronomer Alan Harris, but neither Plait nor Nelson showed his work). They’re expressed as an annual risk of death. In the case of the other dangers, such as lightning strikes, I assume the risk is based on how often it happens divided by the size of the population. Though it’s rare to drown, to be killed by a shark, or to die in an earthquake, it does happen several times each year.
It doesn’t make sense to me that we’d assign a personal risk to death by asteroid, since by doing so we’re basically saying that it happens to several hundred people every year, and again, it has never happened to anyone.
Of course it can happen; we’re overdue for another impact the size of the infamous Chicxulub object, which, if it hit the right spot, could kill every living human. But there’s no way to manage our personal risk of this happening. I can increase my chances of winning the lottery by buying more tickets. I can lower my chances to zero by not buying any tickets. Similarly, I could reduce my chance of being struck by lightning by avoiding tall trees during rainstorms. I could stay away from fault lines to make it less likely I’ll be caught in an earthquake. Et cetera.
There’s nothing I could do to reduce my chance of an object from space killing me. There are no “meteor-safe” cities anywhere on the planet. I can’t go stand under a doorframe if I know one’s coming.
As defined, I could reduce my risk of being killed by an object from space simply by encouraging everyone to have more children, thus making the denominator bigger. How silly is that?