In response to:

# Nate Silver's Numbers Racket

"Now, bear in mind that Jordan's critique centered on what Jordan (a numbers-cruncher himself) argues is Silver's over-reliance on small-state polls." this argument is insanely stupid. 1) previous state poll aggregates, for prior elections, have been shown to be extremely accurate. Those state polls contains the same ones that Jordan complained to be too small in terms of sample. 2) small sample size doesnt matter if you have multiple polls. they have bigger random errors, but those cancel out when you aggregate the polls. 3) small sample -> larger margins of error. That margin of error information is already added into the model with the poll results. that means silver already weighted polls according to their sizes.
Curtis108 Wrote: Nov 07, 2012 5:27 AM
Yes...please explain to me the math behind your assertion that random errors cancel out. There is one huge presupposition in that statement, and I don't think you can make the case for it mathematically.

The random errors, in aggregate, may tend to cancel out, but they don't "cancel out" by some required law of statistics.
yesth Wrote: Nov 07, 2012 7:58 AM
`please explain to me the math behind your assertion that random errors cancel out. There is one huge presupposition in that statement,`

it isn`t. This is basic statistics. Its literally the thing they teach you on the first day of stats 101, i`m not going to bother my time it, google it yourself.
John4278 Wrote: Nov 07, 2012 2:15 PM
Since Silver was weighting polls based on past accuracy and then averaging them together, the "random errors" get averaged out. This raises the signal-to-noise ratio, making them very accurate. Gotta love math!

## Nate Silver's Numbers Racket

In the last week or so, an intense kerfuffle broke out over the poll-prognosticator Nate Silver and his blog at the New York Times, FiveThirtyEight. Silver, a statistician, has been predicting a decisive Obama victory for a very long time, based on his very complicated statistical model, which very, very few of his fans or detractors understand.

On any given day, Silver might have announced that -- given the new polling data -- "the model" was now finding that the president had an 86.3 percent chance of winning. Not 86.4 percent, you fools. Not 86.1 percent, you Philistines. But 86.3 percent,...