Active Users:343 Time:12/07/2025 07:01:26 PM
You're completely missing the point. - Edit 1

Before modification by heartbreak at 07/10/2012 03:35:07 PM

"the most extreme of double digit trials ALMOST matched the poll results; therefore the polls suggest no advantage."


This is not what I'm saying. What I have demonstrated is that with only 100 data points our results per trial can vary enough to lead to different conclusions. Thus, making a conclusion on any given trial of 100 subject to a great amount of error.

Within a minimum of 10 trials, the upper bound of inequality was 69%, less than the polls (which have 36% more samples than each of our trials.) Therefore the collective polls most likely show(ed) a genuine Obama advantage.


136 is not much larger than 100. You have consistently stated that 136 is an ample or huge sample size, can you demonstrate how it is sufficient?

Return to message