Facts aren't decided by opinion poll

I've been trying to work out all day what I make of the results on the Wings/Panelbase poll question about whether people believe there was a conspiracy against Alex Salmond.  (A plurality of SNP voters think there was, but a plurality of all voters think there wasn't.)  The conclusion I've reached is that the numbers don't actually tell us anything much at all, and I'm not even entirely sure there was much point in asking the question at this stage when the facts have yet to reach the public domain.  There are a great many topics on which public opinion matters tremendously, but a dispute over a point of fact is not one of them.  It's either true or untrue that there was a conspiracy - what matters is truth, not opinions or hunches or guesswork.  Any evidence of a conspiracy should be thoroughly investigated, and if wrongdoing is discovered, the people involved should be disciplined accordingly.

I was at an event in Edinburgh a few months ago, in which a well-known comedian made a highly provocative comment that implied Alex Salmond was guilty on a "no smoke without fire" or "he just comes across that way" basis.  The audience roared with laughter, and I was left feeling deeply uncomfortable.  Doubtless if there had been a Panelbase poll of that audience, it wouldn't have been very favourable for Mr Salmond - and that wouldn't have mattered a damn, because when a jury actually listened to the evidence, they acquitted him on every single charge.  Exactly the same principle applies to the allegations of a conspiracy - it really isn't important what members of the public think if they don't have the facts in front of them.

I would just make a couple of technical points about the poll, though.  One of the problems with polling that relies on volunteer online panels is that there will always be a disproportionate number of politically engaged people in the sample.  In practice, that doesn't seem to distort voting intention results too much, but I do wonder if this is the sort of question where there may be an issue.  The SNP voters who were polled may, for instance, be significantly more likely to have read Craig Murray's account of the trial than SNP voters in general.  Of course that probably means they're much better informed than most voters, but the flipside of the coin is that we should be cautious about assuming their views are fully representative.  So I'm a bit dubious about Stuart's crude extrapolations that "one million" people must think this or that.  He's also guilty of a sleight of hand by lumping together Don't Knows with people who think there was a conspiracy, and saying "more than 70% of SNP voters either definitely believe that the party tried to fit up Alex Salmond or consider it a real possibility".  Don't Know means Don't Know.  It's not a synonym for "yes, I consider that a real possibility".  Many people who said Don't Know may have literally known nothing about the subject, and were in no position to offer any sort of opinion at all.

For what it's worth, I'm deeply concerned by what I've heard about the sequence of events that led to Mr Salmond being put on trial, and I hope the truth does come out.  But truth isn't determined by public vote or by opinion poll.