Tuesday, November 06, 2007

Told ya!

My polling efforts proved to be perfect. I knew Luke would get between 63 and 65%. I knew I had 10% of the vote for city controller.

There was no "margin of error." I didn't need a margin. There was no error.

I ran the best and most accurate poll in this election cycle, as a hobby.

I don't like the results nor the predicted outcome, but, I reported them, honestly.

3 comments:

Mark Rauterkus said...

To John K.

I know I didn't win the election. I knew I wouldn't too. I'm over the election -- but will still think again.

You won't hear me blaming the voters. Perhaps I'll dish out some blame to the media -- but not the voters.

No action -- well -- I wish I didn't have to delete your postings here.

MH said...

Mark,

The margin of error is something completely different from how accurate the poll is compared to an actual election. Generally, you need to know how accurate your poll is before the actual election happens.

The margin of error is the spread you need to get a given chance (usually 95%) that the actual population figure is the same as you get with your poll. Usually for a national poll, you will have to get 1200 people or so to get a 95% chance that the actual figure is within +/- 3% of what you get in the poll. The margin of error depends mostly on sample size, but there are other factors.

Mark Rauterkus said...

I understand that the margin of error is a statistical measure.

I was saying that the poll I conducted had NO margin of error as a joke -- and -- there was little or no error.

Truth be told, I did make one mini mistake. But, I'm not going to dive into the data again. It is more of a hunch at this point.