As some background, I have worked in market research for more than 10 years so I do know about surveys, polls and statistics. The people surveyed, how they are surveyed, and how they are weighted to produce the results are just as important as their answers. For example, in this poll, it appears that no weighting by political party is being used. With a sample of 41% Democrats vs. 34% Republicans, it is not surprising that Obama is up so much. Additionally, I admit that Democrats are more “excited” about this election than Republicans. This will definitely affect polling numbers as Dems would be more likely to respond to a survey than Republicans. If anything, I think these polls are hurting the Dems by potentially increasing the opportunity for complacency come election day.

I have reviewed many of the polls’ methodolgoies and would say that Rasmussen does the best job of weighting and reporting (3-day rolling average). However, their poll is not without issues. Their party weighting is based on their own historic tracking from previous polls. I don’t blame them – there are not a lot of ways to obtain this data – but it certainly lends some reason for concern.

Do I think Obama is up nationally? Absolutely. Do I think the race is closer than many of the polls show? Absolutely.

I do think Obama will win – maybe even by a decent electoral count. But it will be a lot closer state-by-state than projected. Obama may still win a lot of states, just not by as big of a margin as people think.

Just my two cents – and maybe a little wishful thinking!