I agree completely with your logic. But Powerball draws are twice a week so in 62 years one could play 6448 draws and thereby double your odds of one success per lifetime to 1 in about 27,000.
By comparison, Wikipedia’s article on coin-flipping cites an estimate that the odds of a coin landing on it’s edge is about 1 in 6000.
ETA: the study cited gives the odds for a nickel, which is unmilled and comparatively wide; the odds with a quarter would be worse.
In order to cause an apocalypse, the asteroid doesn’t need to be big enough to destroy the ecosystem; it only needs to be big enough to damage civilization long enough to cause huge numbers of people to starve. All big cities consume way more food than they produce, and they import the food by trains and trucks. Disrupt the supply lines and the citizens have about a week to either leave the cities or starve.
Sure, you can estimate that a KT Extinction Event sized asteroid will hit once every 100 million years, but it’s also likely that asteroids 1/10th that size hit ten times more often and asteroid 1/100th that size hit 100x more often. I would argue that 1/100 the size of the KT asteroid would be sufficient to disrupt our food supply long enough that 90% of the humans would starve. But we wouldn’t go extinct.
The difference is that once the asteroid strikes, everyone’s wiped out. So if you look at the thing from a long-term (like, really long term, over a time horizon where it’s quite likely that an asteroid strikes) perspective, it’s possible to have a situation where you have a total of seven billion people (or whatever the world’s population will be at the time) dying from the asteroid strike, but the total number of lottery winners in recorded history will be lower than seven billion. Hence, the probability of being killed by the asteroid strike is greater. The fact that we have, so far, not seen people killed by asteroid strikes is simply a result of the facts that we haven’t made observations long enough for the law of large numbers (over the long run, the percentage of events actually observed will approximate the theoretical probability) to kick in.
We all win the grand prize.
Since we’re clearly not going to accept the GQ answer to the original question, at least this thread has conclusively proven that people who play the lottery have a poor understanding of statistics.
I think this is the correct way to analyze the situation. One day an asteroid will hit Earth, and kill all humans. On that day were there more lottery winners in the history of humanity than the number of people killed? If so, then it’s more likely to win the lottery than be killed by an asteroid, if not then the reverse.
The fly in the ointment is that it’s entirely possible that human beings will be already extinct before that asteroid hits earth. If the sun goes nova tomorrow and wipes out humanity then nobody–aside from maybe a few unknown individuals in prehistory–will have been killed in an asteroid strike, while there have been thousands and thousands of lottery winners.
Also note that “winning the lottery” is not very well defined. Unless we define it as winning a particular lottery, like the Mega Millions lottery mentioned in the article cited by John Mace.
I think you dracoi, had the answer. But I still think that formula is based entirely off population. And I dont think the population makes asteroid apocalypses more likely, nor does it increase my chances of dying in such an event. That formula could not be used as a comparison to winning the lottery.
If the population today was only 100,000,000 we would have an actuarial record of 1 death per year over the proposed 100,000,000 million years between asteroid strikes, making my odds of survival (or at least not being alive when the strike occurred) much better. On the other hand if the population was 100 times what it is now, my chances of dying would be much higher while my chance of winning the lottery exactly the same. That does not compute.
This reminds me of a Simpsons quote:
Which is to say that “recorded history” is perhaps a bad way of looking at events that could wipe out records. (Not that that’s the issue here, but it’s worth keeping in mind.)
Grabbing impact frequency data from wikipedia, throwing it into wolframalpha and extrapolating for the odds of winning the mega millions (given a draw twice a week that you always enter) you have a 1 / 2482515 chance of winning the jackpot (according to the data also listed on wikipedia) each year. According to the graph, an impact of ~964 GT happens on average once in this period of time, this is just less than 1% of the estimated impact proposed for the KT extinction.
LOL, that is funny. I only mentioned the fact that “Of all recorded history no one has died from an asteroid while at the same time there have been probably thousands of powerball winners (or similar lotteries)” because it is accurate to say that lotteries are literally guaranteed to be won on a regular basis and on a relatively short predictable timetable. Asteroid apocalypses on the other hand are not guaranteed or on a short predictable timetable.
Although I do realize that as this thread has advanced, my thinking has changed and i realize we arent talking about just anyone winning the lottery but rather a specific person winning the lottery, which is definitely not a relatively short or predictable process.
Wait, what?
A 964 Gigaton impact happens on average once every year?
Like, were there 65 million of them all at once 65 million years ago to make this average, or is this an error? 
the fraction I gave was the probability of winning on a given year, so I calculated the impact with the same probability.
The quote I’m reminded of is:
what if there were a single immortal human on earth doing nothing but playing mega millions twice a week. He gets 104 games a year with the probability being 1 in 285m of winning. Over the course of a 100 million year cycle he would’ve won approximately 36 times for every 1 apocalyptic strike. That should show that he is 36 times more likely to win the lottery than to die by asteroid strike. With his life shortened to only 80 years his probability drops to 1 in 2482515. At the same time the probability that the asteroid will strike in the same 80 year time frame is 1 in 89,370,540.
Lemur866 writes:
> . . If the sun goes nova tomorrow . . .
We know quite about what point in their lifecycle stars go nova, and it will not happen to the sun tomorrow. In fact, it won’t ever happen to the sun. The sun is not the right sort of sun to go nova.
But presumably there would also be 100 times as many people playing the lottery, and so 100 times as many lottery winners over time. The common sense that is being refuted is that since I have heard of people who have won the lottery but haven’t heard of people being killed by an asteroid, then for an individual winning the lottery must be more likely than being killed by an asteroid, but that is just because the lottery winners are statistically independent of each other so a few can dibble in now and then, but in an asteroid impact, we all go together when we go.
No matter how you tackle this one, you have to make some assumptions with affect your answer. But it certainly is possible to use certain assumptions which make death by asteroid more likely.
NASA estimated that a meteor 1km across or larger could cause the extinction of the human race. [cite] Such impacts occur on average once every 440,000 years. [cite] So the probability of it happening in the next 100 years is approximately 1/4400. Multiply this by 10 billion (expected human population) and we have 2.3 million dead. Do you think there will be more than 2.3 million people who “win the lottery” in the next 100 years?
My back-of-the-envelope estimate is that somebody somewhere wins more than 1 million USD in a lottery about 100 times per week. That’s 5,200 per year or 520,000 in the next 100 years. I’m assuming the number of lotteries won’t increase or decrease. So half a million people win the lottery in the next 100 years, which is quite a bit short of the 2.3 million expected to die from an apocalyptic asteroid strike in the next 100 years.
But that’s looking at the big picture, including all people on Earth, even those who never play the lottery and people who live in places that don’t even have a lottery. OTOH, if you focus specifically on one American who buys one lottery ticket every week with odds of 14 million to 1 against, that person’s probability of winning this year is .00037%, compared to .00023% probability of being killed by an apocalyptic asteroid strike this year.
Poor Charlie Brown. “I got a rock.”
Using your logic it’s more likely that I win the lottery than a woman is ever elected president of the US.