Monday, November 05, 2012

A very different point, but interesting...

In light of my previous post (and obvious rush of stats-and-politics geek adrenaline), it seems fair to note this piece by Michael Gerson.  I think Gerson talks past most of the debate about prognostication, but he does make an interesting-yet-different set of points about whether we should care so much about polls and measurements. 

I think that, ethically speaking, I agree with him a good bit.  As a thought experiment, would the world be better if we had a self-declared blackout on all polling data for 3 months before an election?  I am willing to entertain that we might all benefit.  Deciding whether a set of policies (and what should happen) is right or wrong is simply a very different issue than whether a prediction about what will happen is right or wrong.

More tomorrow.

No comments: