Panelbase datasets confirm it's 50-50 on independence

In the unlikely event that anyone from Britain Elects is reading this, could I just point out that you still have a tweet up from Saturday night wrongly suggesting that the new Panelbase independence poll showed figures of Yes 49% (-4), No 51% (+4). The correct figures, which have now been confirmed with the publication of the Panelbase datasets, are in fact Yes 50% (-2), No 50% (+2). I know the mistake will have been made in good faith, but the frustration is that large sections of the London media treat whatever Britain Elects post as gospel, so until there's a correction journalists will continue to wrongly believe and claim that Panelbase was a No-majority poll. 

I don't pay the Murdoch Levy, but I gather that the article from the Sunday Times (who commissioned the poll) didn't even mention the headline numbers with Don't Knows excluded, which is highly unusual, so it's difficult to escape the impression that all this 'confusion' has come about partly by design. The datasets show that, after weighting, 415 respondents said they would vote No, and 412 said they would vote Yes. You can hardly get much closer than that - it works out as Yes 49.8%, No 50.2%. With Don't Knows left in, it's Yes 46.4%, No 46.7%, so it was just bad luck that those numbers were rounded to Yes 46%, No 47% - but the figures with Don't Knows excluded are generally regarded as the headline numbers anyway. So, just like the recent Survation poll, it's a genuine 50-50 outcome.