Sunday, 28 August 2016

10:34 – More kit stuff today. The warm weather is back. Highs are only in the low 80’s F, but the heat chill is up in the low 90’s. Lows are down in the 50’s, which means autumn weather isn’t far away.

Exponential figures are interesting. The other day, someone emailed to ask what I thought the probability was of something really bad happening. I told him I thought the probability was on the very close order of 0.9999 that nothing really bad would happen tomorrow, or any given day. One chance in 10,000 in other words, or a 0.0001 probability per day.

The problem is obvious when you extend that. If the probability is 0.9999 that nothing bad will happen in a day, that means the probability that nothing bad will happen in a year is (0.9999)^365, or just over 0.964. In other words, there’s about a 3.5% chance of something really bad happening in the next year. Extend that out to three years and you have (0.9999)^1095, or about .896. In other words, there’s about a 10.4% chance of something really bad happening in the next three years. Of course, all of that depends on the initial estimate. If the probability is instead 0.99999, one chance in a hundred thousand per day, the exponential figures are much lower. But if it’s instead 0.999, one chance in a thousand, they’re much, much higher.

Based on history, I think it’s ridiculously optimistic to assume 0.99999. Assuming 0.999 may be pessimistic, but not by much. For example, our planet is hit by a Carrington-class CME on average once every 100 years. The last time that happened was in 1859, so we’re well overdue. In 2012, we narrowly avoided being struck by a catastrophic CME. NASA estimated at the time that there was a 12% chance we’d be struck by such a CME before 2022. If their estimate is accurate, that’s a 1.2% chance of a catastrophic CME per year. Then we have an EMP attack, or hackers destroying our electric distribution network. It’s difficult to estimate the probability of that happening, but it must surely be at least as likely as the CME, and would, if anything, be more catastrophic than the CME. Then we have a lethal pandemic, which historically occurs about every 100 years. It’s been just about 100 years since the last one, the Spanish Flu of 1918. With modern air travel, the next pandemic is likely to spread planet-wide in a matter of days, something they didn’t have to deal with in 1918. And those are just the most serious threats. Among them, they make my 0.9999 estimate seem reasonable, or even too optimistic.