Opinion polling has ostensibly 'become more sophisticated' year on year yet it failed to predict accurately any of the major earthquakes of 2016.
If the polls get it right this time, there may be a suspicion of mere coincidence. They are all over the place in any case with different methodologies yielding different results. We are reduced - since we are not psephological nerds - to taking an interest only in trends that are based on the averaging of polls - trackers. These do not predict the final result with any certainty but they tell us some useful things nevertheless.
The first thing they tell us is that public opinion (which may not be the same thing as actual voter turnout) has done something quite remarkable since the election was called. Whilst the Tory vote has risen and then fallen back (though not falling below the level at which it started on April 18th when the election was called), the Labour vote seems to have risen steadily and incrementally to the point where, if projected further with the latest downward trend of the Tories since May 15th, could theoretically find the two parties neck and neck on voting day.
Some of this is down to the other phenomenon to be seen in the tracking - the collapse of the two national minority parties who appeared to throw all their cards down on the one issue that initially seemed to define the Election but which has proved to be secondary, Brexit. Tactical voting against the Tories seems to have driven Liberal Democrat votes towards Labour (or middle class Liberals turning back to the Tories on tax are being offset by a return of Labour people to their Party). UKIP voters seem to have moved towards May and then drifted back to Labour on (perhaps) social care.
Of course, we cannot know precisely what the dynamics are but May 15th-18th seems to be the key period when Labour started to take off. And this is the remarkable bit. All our reports from the Northern and Midlands front line up to that date were crystal clear - Jeremy Corbyn was poison to the working class. These reports were unequivocal. But May 15th saw Labour deliver a Socialist Manifesto as radical as that of Michael Foot in 1983 while Prime Minister May seemed to blunder on May 18th with what Labour cleverly labelled a 'dementia tax' related to social care.
The received narrative about British politics since the Thatcher era imploded within these four days. Well over a third of the electorate now seem to be relaxed about hard-line socialism. The security crisis that emerged after the suicide bombing in Manchester on May 22nd seems not to have affected this trend. If anything, former Home Secretary May may be seen as partly responsible, fairly or not, and Corbyn's peaceniik interpretation of the causes of terrorism may have resonance with many, pulling the rug from under the Tory attempt to denigrate him as its appeaser.
On balance, we still would not expect a Labour victory because those more likely to vote are more likely to vote conservatively but we have to consider the possibility of no increase in the Tory majority or even a hung Parliament, either of which result would be a personal disaster for Prime Minister May and may result in her losing her position. Equally important for the future, it is probable that the Left can now point to the potential for radical and socialist ideas becoming the basis for future electoral success under certain circumstances. She needs a 50-80 majority, not 25-30, to be fully credible.