Third Party and Independent Daily found a great post detailing three rather blatant examples of polling bias against third party candidates. His findings are pretty disturbing to anyone who cares about accuracy in polling.
I noticed something quite disturbing when studying recent polling for elections around the country. Rasmussen Reports will include upwards of 6 or 7 candidates in the Democrat and Republican Primaries, sometimes with candidates that only receive 1% of the vote, yet when it comes to the General Election they will only include the two Democrat-Republican candidates despite there being sometime upwards of 3 or 4 other candidates from Third Party Nominations or Independent Nominations.
Regardless of whether or not these third party candidates in question align with my personal beliefs, or those of this site in general, it should be a general principle that polls should actually aim to represent the will of the people. You can’t do that without asking them about every candidate that can be reasonably expected to have a statistically significant amount of the sample.