Is Ian Smart a malfunctioning android?

Now don't bite my head off, I'm merely throwing the question out there.  I know we've long suspected that Smart isn't quite of our species, but I now feel that the sheer regularity with which he follows up tweets telling us that "Victory Is Inevitable! #itsover" with the not-exactly-logically-consistent message that "There Is No Room For Complacency" probably demands an even more drastic explanation.  If by any chance he isn't a malfunctioning android, my only other thought is that we could be witnessing an incredibly advanced type of performance art.

Here are some of Smart's "insights" from earlier this evening, which as ever take a fair bit of imagination and a shot of vodka to make much sense of -

"Slowly sinking in to the cybernats that their public meetings speaking to each other doesn't actually reflect the real world. 28%."

I don't know about you, but I'm struggling to think of a single bona fide "cybernat" who doesn't believe both publicly and privately that Yes are going to win this. My own view is that the referendum could easily go either way, and accordingly I sometimes find all the misplaced euphoria a bit wearing. So heaven only knows which "cybernat" Smart thinks he's been talking to - maybe it was Tom "It Is I, LeClerc" Harris in his latest brilliant disguise?

"Not close to 49%. Not even close to 40%. Phone banks don't lie. That's why they are consoling themselves in their own company."

Well, Ipsos-Mori use phones to conduct their surveys, and even they (as the extreme No-friendly outliers) produced a Yes vote of 36% in their last poll two months ago. Surely even a malfunctioning SmartDroid Z-3000 can see that 36% is considerably closer to 40% than it is to 28%?

"The polling is nonsense. They keep adjusting the sampling to try and pretend it's a contest. Look at the real election results."

Yes, the entire polling industry (which for the purposes of this referendum is 100% English-based) is quite clearly in cahoots with the pro-independence campaign and is trying to make the race look closer than it is. And if Ian could just supply a single reason why they might be doing that, his theory would sound even more plausible.

"At a funeral c 2 months past was worried when I was told by 4 others they were thinking of Yes. Turned out they were taking the piss."

What an astonishingly classy anecdote.

"So, Kilsyth thinks a Yes vote is worthy only of humour. More worrying for the Nats is that I suspect so does Cumbernauld. #itsover"

Hmmm. Speaking as a former resident of Kilsyth, and a current resident of Cumbernauld, "I suspect" I have a fair idea whose name will be the punchline to a good few jokes told in these parts on September the 19th. I don't want to be unkind, but frankly we're still trying to recover our composure after listening to Ian's plucky attempts to explain away his repeated predictions that "there will be no referendum".

"Mind you, there is no room for complacency. Off to bed."

Bingo!

Barking mad though he is, a man with Smart's Labour background presumably does have some genuine connections, so the specificity of the comments about "phone banks" and "28%" led me to wonder if we were supposed to assume that he was hinting at private polling or canvassing data. Luckily, I wasn't left in the dark for long, as it turn out his witterings tonight were just a rehash of something he wrote on his blog a few weeks back -

"Firstly, we have the polls. Here they are. Look at the orange line (that's us) and then look at the blue line (that's them). What do you notice? The orange line is comfortably above the blue line in every case but one and even that one is a poll now accepted on both sides to be flawed in it's methodology. And what else do you notice? That this hasn't changed in a year."

Hang on - I thought the polls were nonsense?  (Mind you, Ian wrote this a month ago, so it's possible they've only become nonsense since then.)  For reasons I've rehearsed umpteen times, the averaging method I use on this blog is actually much more generous to the No campaign at present than pretty much any other averaging method would be (it's certainly more generous than the one the FT is using for its own Poll of Polls), and yet even I'm showing that the No lead has collapsed from 21.6% to 11.4% since September. What do you notice most about those two numbers, Ian? That's right - the second one is only half as big as the first.

"Then we have real elections. The SNP won a local government by-election, retaining a safe seat, in Kilmarnock on Thursday. Congratulations to them. Particularly since it was the first by-election of any sort that they had won in the last eleven. And even then their vote dropped 9%. While the unwritten story of Scottish politics continued. That there was a major swing to the Tories. Now, it might be the case, as the SNP losing candidate in Cowdenbeath maintained with a, just about, straight face, that voters are embracing Nationalism while simultaneously abandoning the Nationalist Party. Somehow, however, I doubt it."

To be fair, this is (just for once) a point with a small glimmer of reasonableness about it, and if you dilute it by a factor of about 20 it's similar to one I've made myself a couple of times. If we believe the Yes-friendly pollsters, the SNP still have a whopping Holyrood lead, and the pro-Labour swings in the real elections in Cowdenbeath and Kilmarnock do contradict that to some extent. But here's the thing - those were only two localised elections. Anyone would think Ian was talking about dozens of elections that had taken place over the last few months, all showing a uniform picture. The Scotland-wide European election next month will resolve the issue - if Scottish Labour do much better than the likes of ICM and Survation are predicting, then Ian has a point and those companies' referendum polls should probably be regarded as suspect. But if the SNP do as well or better than predicted, Ian will naturally eat his words with the good grace to which we have become so accustomed. Oh wait...

"Then we have the school polls. School polls are interesting since they don't rely on adjustment by pollsters."

You mean they are restricted to a very narrow demographic, and are in any case self-selecting, thus robbing them of the scientific validity that the polls enjoy? That's an unusual definition of 'interesting'.

"There were another slew on Friday and again Nationalism was rejected by a margin of between to (sic) and three to one. Indeed, to the best of my knowledge, no secondary school of any proper size in the whole of Scotland has yet to vote Yes in such an exercise."

In other words, several secondary schools have voted Yes (Plockton High and Dornoch Academy, for example), but SmartDroid Z-3000 has deleted those from its memory-banks due to arbitrary post-programmed threshold criteria relating to 'properness'. The reality is that school mock votes have occurred in many wildly different formats, and the claim that the No campaign has generally come out on top depends on conveniently denying the validity of some of those formats.

As an aside, Better Together ought to be absolutely petrified by the result of the Strathclyde University mock referendum the other day. As we know, the more prestigious universities in Scotland tend to have a disproportionately high number of affluent students, and a very significant minority of English students - both of which are No-heavy demographics. There is no way on Earth the Yes vote in Strathclyde should have been anything like as high as 45% - and yet it was.

"Then we have the 'feel' of the ground troops. Those who are doing the telephone and street canvassing. Obviously I speak to more of ours than theirs. But I do speak to theirs. Ours admit to genuine bewilderment as to where this supposed, even significant minority, Yes vote is."

There may be a grain of truth in that - we've been told, for example, that the Labour and SNP canvassing on the independence question in Cowdenbeath produced utterly irreconcilable results. Unless one side or other was lying about their numbers, it's possible that an undecided electorate is giving different answers depending on how the question is posed, and who poses it.

"At best, theirs insist they have 'still got a chance' even while bemoaning the various factors (apathy, fear, the BBC) conspiring against them."

This claim, on the other hand, appears to be utterly delusional. Admittedly I've no knowledge about phone canvassing, but we've all heard feedback from street canvassers for Yes, and the vibe we're getting from them is, to put it mildly, somewhat better than "well, maybe we've still got a chance". But I suppose it's possible that SmartDroid Z-3000 is programmed to interpret legitimate complaints about an obscenely biased media as some kind of admission of defeat.

"Then we have the argument, not on the policy but on its prospects, of our opponents. Chief among these is Mr Stephen Noon, who informs his observations with hints of secret polling known only to an elect of which he is one. I have no idea why this guy is taken seriously. Here is what he wrote on 27th April 2012, a week out from the local government elections in 2012. Either his "inside information" was seriously flawed or more likely he was just making it up. For we know the actual results of these elections."

Yes, we do - and as Stephen Noon correctly anticipated, the SNP won a historic victory. Ian Smart on the other hand, claimed that the Scottish Government had no intention of holding an independence referendum. He claimed that Labour had won a stunning triumph in the 2012 local elections even after they had lost. He tells us on an ongoing basis there is no room for complacency, but adds that a crushing referendum victory is already predetermined. And then he questions (of all things) the credibility of others, based on their track record.

For a human being, that would take some gumption. But for a malfunctioning SmartDroid Z-3000, it's just part of the daily routine.

* * *

As of this moment, it looks like this is going to be a poll-free weekend, but thanks to Calum Findlay for pointing out that ICM asked something very close to the actual referendum question in a GB-wide poll conducted between the 19th and 20th of April. That's a few days later than the fieldwork for the sensational ICM poll that put Yes just 3% behind.  The results among the Scottish subsample of 166 people are eerily similar to that poll -

There will be a referendum in Scotland in September 2014 on Scottish independence. Should Scotland become an independent country?

Yes 39%
No 43%

Even if the subsample figures were properly weighted (and the chances are they weren't), the margin of error would be a very high 7.6%.  So the fact that the numbers are so similar to the full-scale Scottish poll may simply be coincidence, but it's an encouraging straw in the wind all the same.