Yesterday the Labour Party released details of its own investigation in to why they did not do as well as they expected in the general election and the pollsters also released the results of their inquiry why the polls they carried out did not show the Conservatives in the lead.
Even the final exit poll was not completely accurate as it only suggested the Conservatives would be the largest party rather than have an outright majority. Few will be surprised by the finding that they polled too few Conservatives and too many Labour supporters. Getting the sample right is something which is much easier said than done. Modern society is so complex that identifying a representative sample with around a thousand people is incredibly difficult.
I tend to look just at the general trend of polls. If there are several polls all using different methods which point the same way then that suggests to me that finding could be right. In the end in elections there is only one poll that really matters and that is the one on polling day itself!
David Nuttall 
It shows what a waste of time and money is spent on these commercial polls. It all depends on the phraseology used by them and what results they appear to want. I’m sure grass-roots canvassers are better predictors of people’s interests and concerns. Remember Gordon Brown and his opinion of the concerned voter in Rochdale?