ShuaiGuy wrote:From the article:

Thanks to work done by him and his colleagues, we now know that the odds of the big Cascadia earthquake happening in the next fifty years are roughly one in three. The odds of the very big one are roughly one in ten.

Thanks to that work, we now know that the Pacific Northwest has experienced forty-one subduction-zone earthquakes in the past ten thousand years. If you divide ten thousand by forty-one, you get two hundred and forty-three, which is Cascadia’s recurrence interval: the average amount of time that elapses between earthquakes. That timespan is dangerous both because it is too long—long enough for us to unwittingly build an entire civilization on top of our continent’s worst fault line—and because it is not long enough. Counting from the earthquake of 1700, we are now three hundred and fifteen years into a two-hundred-and-forty-three-year cycle.

The author goes on to say,

It is possible to quibble with that number. Recurrence intervals are averages, and averages are tricky: ten is the average of nine and eleven, but also of eighteen and two.

. But then goes right into the same "OH MY GOSH WE'RE ALL GONNA DIE" in that same paragraph.

I think that the article did a poor job of explaining how the 1/3 and 1/10 figures were arrived at, and could have said more about the problems with average recurrence intervals. I would at least have liked a maximum and minimum of the intervals where we know them. But unless this a really deliberate attempt to mislead the mention of it would sduggest that it's been addressed in the calculations - I expect it was felt to be just too hard to understand for the target audience: statistics are confusing.

Before calling bullshit, let's just look at the figures. We're somewhat over the average of the intervals now, yet the figures say it's twice as likely not to produce an earthquake in the next fifty years as it is not to. There is a 90% chance that there won't be a megaquake in that time, equating to it being around twice as likely that even if there is a quake it won't be the 'really big one'.

Yes, I do find the tone of that section to be somewhat sensationalist (I'd rather the expression '243-year cycle' hadn't been used, for starters), but the actual figures are saying it probably won't happen in the next fifty years and probably won't be the worst case even if it does. The author's inclusion of the caveat implies that the odds have been estimated with all that taken into account. I'd have be an expert on the subject and come up with my own estimate independently from the data to say whether the numbers are good but at a naive glance they look modest enough.

We can't predict when an earthquake will happen - we can't forecast them so as to evacuate the area and begin an emergency response, but we can predict that it will at some point, and to suggest that putting any kind of figure on whether it being within a reasonably coarse-grained period is a remote or a significant possibility is ludicrous is rather a stronger statement.

From

my research just now, the maximum and minimum intervals that we know are 900 years and around 200, with the majority being around 200-400 and 1/3 of the intervals 2-3 times that. There is nothing between 400 years and almost 800. But the number of data points is painfully small: 6 intervals, with two long ones and four short.

The most recent interval (the one before the 1700 quake) was long, and the series is 'short short long' repeated twice. So at a naive glance it looks as though it would imply the next interval would probably be nearer to 300 than 800 years, but it's too few data points to say so - one of the sources for the wikipedia article addresses that in slightly more detail.