And I was so relishing the prospect of some entertainingly inventive sophistry in response to my challenge to AM2/Scottish Unionist - he's delivered so reliably on other occasions, but perhaps the poor chap is tired. (Understandable, he doesn't half get around.) Having finally let my comment through moderation, his response was indeed to update his table, but with the figures miraculously rearranged in a completely different order so that the most recent poll features in the middle rather than at the end! His justification? The latest poll looks like a 'blip' to him. Fair enough, let's dispense with all the hard statistical analysis, AM2's intuition is far more reliable.
I think AM2/Scottish Unionist urgently needs to acquaint himself with Mike Smithson's 'second golden rule' - "a rogue poll is one where you don't agree with the numbers".
To illustrate this point, let me ask another question of AM2 - do you think you would you have written a post full of shameless innuendo attempting to smear TNS System 3 and undermine its credibility if by any chance the latest poll had shown 20% in favour of independence and 60% against? Do you not think in those circumstances your post would have been rather more to do with the contents of what you'd have taken as read as being a rigorous, scientifically-conducted survey? A simple yes or no will suffice.