Re:ZestCash is Not Good.
by jarek on Friday October 15, 2010 @12:04AM
(#1793662)
It rather destroys the supposed correlation between crappy jobs and awesome people.
[ Parent
| on HN ]
No, it doesn't. Correlation is not causation. If awesome people have worked crappy jobs more often than not, there can still be a positive correlation between the two variables even if there are dramatically fewer awesome people than people who have worked crappy jobs.
Yes, it does. Crappy jobs won't be correlated with awesome people -- which was andrewljohnson's original claim -- to any meaningful degree given the billions of unawesome people. Since correlation is symmetric, awesome people will not be meaningfully correlated with crappy jobs, either.
Plot every person in the world on a two dimension grid. Dimension x is awesomeness, dimension y is crappiness of crappiest job. Vast majority of people will be clustered around top left of graph (unawesome people with crappy jobs). But if the mean y value fluctuates upwards as you move to the right along x, then there's a positive correlation, regardless of whether or not there are significantly fewer data points further to the right. Read the "Correlation and linearity" section of the Wikipedia article: