LogFAQs > #894442134

Tropic_Sunset posted...

Oh, I would have thought you were comparing the shutdowns to his argument.

his claim was that the rate (before this one, because we're attempting to determine whether or not this one is unusual or not in comparison to something else) was 1 per year.

if I look at the history I see that the rate (before this one) was 1 in the last 20 years.

if the rate, before this one, had been 1 per year, then yes, this one in comparison would not have been unusual, as he implied.

given that the rate, before this one, (looking specifically at the last 20 years, which could actually be argued to be somewhat misleading), is 1 in over 20 years, then no, this one in comparison IS significantly more unusual than he implied.

Tropic_Sunset posted...

But if they occurred like clockwork, exactly once every 4.75 years, what would be a more accurate statement?

A) We've had 2 shutdowns in 9.5 years, or
B) Except for right now, we've had one shutdown in 9.5 years?

that doesn't make any sense, you subtracted it from the average. you don't subtract it from the average, you don't include it in the average in the first place.

think of it as drawing a line on the timeline. you're trying to compare one side of the line to the other, wherever the line might be. it doesn't make any sense to compare two *overlapping* time intervals.

it's like, if something happened a million times this year, and you wanted to know if that was significant or not, you wouldn't include the million in your average. you would take the average *before* the million, and the average of the million and compare the two distinct intervals.
---

The truth basks in scrutiny.
http://i.imgur.com/GMouTGs.jpg http://projecteuler.net/profile/Sahuagin.png