Generally I agree with your points, but the outcome of the election will still be determined south of the border, whether or not the tories reestablish their representation in the junior partners. The question remains how large any majority will be: 20-60? we shall see.

So I wonder how you’d call the forthcoming Glasgow East by-election where it stands Lab 60.7%, SNP 17.0%, LD 11.8%, Con 6.9%, SSP 3.5%… this sounds like an ideal indicator for how the picture will change.

“In many areas I’ve been on the ground there is an increasing trend for divergence where support for one party at one level doesn’t carry across at other levels (at least for the winning side, opposition parties seem to count on entrenched support).”

With up to four parties in contention in different rankings in different parts of Scotland, and an electorate now accustomed to the potential of split votes for the SP, complexity is an understatement.

I’m not so sure that volatility is the right word, unpredictability and inconsistency would be better. There are entrenched LibDems and huge Labour majorities. There are also three party marginals.

If the Labour vote crumbles where there is a narrow majority over the LibDems, and the others are well behind, the LibDems can gain a seat even if they too lose votes. On the other hand, whether Labour are a poor fourth, or have an unassailable majority it doesn’t really make any difference what they lose or who benefits.

In my own LibDem held constituency, ex-Labour voters aren’t going to vote Conservative, and if the LibDem loses, (which I doubt) it is more likely to be to the third placed SNP than to the Conservative.

I think the Scottish results will demonstrate that apart from the last redoubt of Old Labour in the West, the class war tribalism of the past is gone; that very few vote for any party’s philosophy; and that most results are determined by regional opinion of the least-worst choice.

FPTP is like a gyroscope. It works just as well upside down. The GROT’s will win.

In many areas I’ve been on the ground there is an increasing trend for divergence where support for one party at one level doesn’t carry across at other levels (at least for the winning side, opposition parties seem to count on entrenched support). This naturally confuses the national (sorry, UK-wide) picture and indicates the greater complexity which can lead to unpredictably volatile results, which I can’t emphasise enough.

Different parties are taking different strategies which play on their strengths – ie wealthy Tories using national and regional media to reach the widest possible audience, Labour unionists using their official status to build themselves into the social framework, LibDems using the strength of local organisation to develop targetted campaigns, SNP/PC mixing and matching. What this means for the result of GE outcome is that none can be treated equally, only compared on like-for-like terms where they are in competition.

In the run-up to the GE I fully expect each side to play games of subtefuge in order to maximise their capacity to take seats without engaging in full-on battle (partly to keep costs down), so the idea of ‘key battles’ is particularly up in the air at this time.

Trying to make a complete or completely accurate prediction from this far out is futile guesswork, and is probably more helpful to predict ‘probable’ victories with an additional column for ‘uncertain’ outcomes (as distinct from ‘marginal’ or ‘swing’ seats).

I’ve tried calculating my prediction using this formula and the best I’ve come up with, taking all these factors into account is currently Con 230, Lab 150, LD 45, uncertain 186, based on a May2010 date. Not wildly helpful, I know, but it is still indicative that Labour will struggle not to be overwhelmed by hosting the majority of battles on their turf and seemingly positive for the LDs.

Of course an earlier date for the GE reduces the uncertainty factor (significantly in the favour of the Conservatives, which mitigates against the likelihood). Additionally the outcome should become more certain the closer we get as polling trends become clearer and more defined, but, equally, continued poll volatility works against this.

Adding this information into spending calculations suggests Labour will struggle to reach 200, while LDs could realistically make some gains. Under this model the Conservatives are still only heading for a small overall majority, though I think that is a message they will hope to exploit to their advantage and create additional forced choices to further encourage an increase in tactical voting levels, thus boosting future volatility again.

“One easy way to see how this works is by identifying and splitting the safe, marginal and three-way seats from each other …”

That is true overall, but even more necessary in four-party Scotland. If you think Labour will lose 10% of their vote it won’t make any difference at all where they hava a huge majority in Glasgow, nor where the are a poor forth in a LibDem highland constituency.

I am still developing my spreadsheet to apply a “What if?” percentage loss or gain of the previous election result in each constituency. So far I get the impression that the number of seats changing hands could be far fewer than some imagine; that all three parties will gain at Labour’s expense; and the most important change in many constituencies will be the positioning of the SNP in a winnable second place for next time.

The speculating commentary which accompanies any new poll essentially mitigates for the failure of pure numbers to describe a rounded picture of real opinion and the consequential shift which can only be guessed at.

So, for example, while Conservatives may crow at their current 20pt lead the distribution factor may well come into play and leave them racking up huge majorities in their heartlands and failing to pick up seats elsewhere, or alternatively we may see a continuation of the current propensity towards landslide elections.

Either way the conservatives have much to be concerned about because the difference between the two opposing scenarios is paper thin and indicative of the current volatile nature of our politics. Assuming the 20pt lead is borne out at the next election a failure to gain a majority would show a failure of our democracy and highlight the divisions in society, while a large majority above 100 seats will create an unstable parliamentary force liable to suffer from internecine splits.

Whether we choose to analyse the implications for strong leadership, healthy institutions or the expansion of well-being and prosperity across the nation we are completely ignoring the threat to the stability of our affairs by the growth of our polity and the destabilising effects caused by growing power inequalities (economic, political etc).

One easy way to see how this works is by identifying and splitting the safe, marginal and three-way seats from each other to show how the fluctuations in support vary relative to each other, but balance out across the board. Traditional weighting factors used by YouGov are becoming too blunt to be more than generally informative, yet their results are afforded equal if not special consideration when it would be more helpful to integrate these figures in a more complementary manner.

This and other similar sites could easily lead the way in presenting interactive user-oriented methods of differentiation which simultaneously enable additional original feedback covering the detail of such variability beneath the headline figures (by comparison central bank inflation measures cover a breadth of data from m0 through to m4 variously according to the country covered – which can be argued is a development that has proved a formidible tool in smoothing the transition of growth and contraction cycles).

The diffusion of democracy can only be defused by recombining the proliferation of political measurement tools available to us – the electorate is getting more sophisticated at judging the range of choices available to us (or lack thereof), so pollsters need to develop new methods to compensate for the abstraction of their broad-brush approach inherent in traditional poll questioning.