“Isles is a storm spotter for the National Weather Service and was on his ham radio at the time.” (From the KCTV story.)

I think the probability of getting struck by lightning conditional on being someone who’s regularly in their yard on the radio during a storm is probably a bit higher than the probability for the population at large.

But you’re ignoring that the probability depends on the condition of there being a lottery (or a thunderstorm). If there are thunderstorms everywhere that lottery tickets are purchased, that would tell you an informative odds ratio.

That’s why I said “of what to what.” It actually is a great deal more complicated than the mere existence of lotteries and thuderstorms at the same time…. that’s what makes conditionality fun.

Sadly, no one ever gets probabilities right, not even this story.

Part of the problem is that most people don’t even coherently define what’s being compared. This article has a common pitfall of failing to make sure that it compares quantities of the same dimensionality: the probability of being struck by lightning (1/835,000 yr^-1) and the probability of winning a jackpot (1/175,000,000, dimensionless).

If you buy exactly one Mega Millions lottery ticket (worth $1) for every drawing, and you have nationwide-average odds of being struck by lightning, your chances of being struck by lightning are 1/775,000 per year, and your odds of winning the jack pot are 1/1,690,000 per year.

Alternatively, you can compare the odds of winning the jack pot on a single ticket vs. the odds of being struck by lightning during a time period between two lottery drawings, in which case both probabilities are 104 times smaller but their ratio remains the same.

Nameless: Great catch. So the (properly equivalent) odds of winning the jackpot are only about a half of the odds of being struck by lightening. But this also excludes the fact that there are very some strong selection effects for being struck (e.g. holding a radio antenna at the time, like this fellow). Can there be any doubt that “you are more likely to be struck by lightning than win a jackpot” is simply false for 99.99% of the lottery players who might hear it?

There are fairly large geographical variations. Your odds of being struck by lightning are more than doubled if you live in one of the coastal Southeastern states (Louisiana to South Carolina) and greatly reduced if you’re in the Northeast or on the West Coast. There hasn’t been a single lightning fatality in Washington or Oregon in the last ten years. And yes, there are selection effects, too. Your odds are increased if you’re in construction or agriculture, or if you’re a hispanic male.

That is fantastic!

“Isles is a storm spotter for the National Weather Service and was on his ham radio at the time.” (From the KCTV story.)

I think the probability of getting struck by lightning conditional on being someone who’s regularly in their yard on the radio during a storm is probably a bit higher than the probability for the population at large.

Yes, but only one guy got hit by lightning. Three winning tickets were sold. A 3:1 ratio. (Of what to what?)

But you’re ignoring that the probability depends on the condition of there being a lottery (or a thunderstorm). If there are thunderstorms everywhere that lottery tickets are purchased, that would tell you an informative odds ratio.

That’s why I said “of what to what.” It actually is a great deal more complicated than the mere existence of lotteries and thuderstorms at the same time…. that’s what makes conditionality fun.

Sadly, no one ever gets probabilities right, not even this story.

Part of the problem is that most people don’t even coherently define what’s being compared. This article has a common pitfall of failing to make sure that it compares quantities of the same dimensionality: the probability of being struck by lightning (1/835,000 yr^-1) and the probability of winning a jackpot (1/175,000,000, dimensionless).

If you buy exactly one Mega Millions lottery ticket (worth $1) for every drawing, and you have nationwide-average odds of being struck by lightning, your chances of being struck by lightning are 1/775,000 per year, and your odds of winning the jack pot are 1/1,690,000 per year.

Alternatively, you can compare the odds of winning the jack pot on a single ticket vs. the odds of being struck by lightning during a time period between two lottery drawings, in which case both probabilities are 104 times smaller but their ratio remains the same.

Nameless: Great catch. So the (properly equivalent) odds of winning the jackpot are only about a half of the odds of being struck by lightening. But this also excludes the fact that there are very some strong selection effects for being struck (e.g. holding a radio antenna at the time, like this fellow). Can there be any doubt that “you are more likely to be struck by lightning than win a jackpot” is simply false for 99.99% of the lottery players who might hear it?

There are fairly large geographical variations. Your odds of being struck by lightning are more than doubled if you live in one of the coastal Southeastern states (Louisiana to South Carolina) and greatly reduced if you’re in the Northeast or on the West Coast. There hasn’t been a single lightning fatality in Washington or Oregon in the last ten years. And yes, there are selection effects, too. Your odds are increased if you’re in construction or agriculture, or if you’re a hispanic male.