I just found this article [1]about polling from last week, by Stuart Rothenberg. It makes some good points about how polling works, what the pitfalls are, and the difference between partisan and non-partisan pollsters. I am not sure I share his bias for partisan polls, and I think he overvalues the cost to partisan pollsters of doing bad polls. But, the article shows how readers need to be discriminating in alanlyzing polling data.