> You tried to because you disingenuously tried to say that the drop would "disprove" my argument.
"It went down 19%" disproves your argument if you don't treat it as an outlier.
It doesn't matter how that compares to 2019 in particular.
I don't understand why you think there's anything disingenuous about that.
> Like, we can see WWI and WWII in those figures, I'm not trying some weasel word sleight of hand to fool people into believing I meant monotonically increasing, as though the actual increase doesn't prove my point in substance.
I assumed we were treating those as outliers. Hence my confusion at not treating covid as an outlier.
> Every one of the full 4 decades after the 1970s was substantially higher than the 1970s, taken as an average (which is slightly weird but it'll do). And not just slightly elevated, the next lowest decade is more than 6.25% above the 70s, and the rest are 10-11% above.
But a lot of the interesting points don't align well with decades.
If we offset by exactly half a decade, to get a different view but in the least cherry-picky way possible, 1975-1984 is 34.6 average, 1985-1994 is 37.3 average, and 1995-2004 is 34.6 again.
Then 2005-2014 goes up again, but only to 36.8, less than the peak.
Seems to me, the person is claiming that a linear trendline increasing causes "steadily" to be an appropriate proxy for "values went up and down some, but overall went up".
If we can agree that the trendline is linear and that it goes up, then yes, we can agree on "steadily".
However, if we observe that the actual data goes both up and down over a specified period, then no, the data cannot be characterised as "steady".
For those interested, perhaps I could suggest drawing a process capability chart? The sigma variation lines would seem to clearly demarc outliers.
"It went down 19%" disproves your argument if you don't treat it as an outlier.
It doesn't matter how that compares to 2019 in particular.
I don't understand why you think there's anything disingenuous about that.
> Like, we can see WWI and WWII in those figures, I'm not trying some weasel word sleight of hand to fool people into believing I meant monotonically increasing, as though the actual increase doesn't prove my point in substance.
I assumed we were treating those as outliers. Hence my confusion at not treating covid as an outlier.
> Every one of the full 4 decades after the 1970s was substantially higher than the 1970s, taken as an average (which is slightly weird but it'll do). And not just slightly elevated, the next lowest decade is more than 6.25% above the 70s, and the rest are 10-11% above.
But a lot of the interesting points don't align well with decades.
If we offset by exactly half a decade, to get a different view but in the least cherry-picky way possible, 1975-1984 is 34.6 average, 1985-1994 is 37.3 average, and 1995-2004 is 34.6 again.
Then 2005-2014 goes up again, but only to 36.8, less than the peak.
Is that facts enough for you?