At the outset of every election, the commentariat often make claims about this being the most exciting and unpredictable election ever.

In part, these claims are to justify an existence in the self-preservation stakes among journalists. After all, no-one wants to write about a boring election with a predictable outcome.

This time, however, it really does smack of the most unpredictable election ever. Not since the election of February 1974 with perhaps a passing nod to 1992, has an election induced the jitters when it comes to the prediction stakes.

In February 1974 Ted Heath tried to turn a general election into a referendum on the issue of 'who governs Britain?'. This was after two miners' strikes, a three-day working week and a protracted state of industrial paralysis. As the late Anthony Howard observed, the answer to his question was "not you, old fruit".

Not since 1974 has a general election served as a referendum by proxy. Or should that read three referenda with implications for Brexit, Scottish independence and a potential border poll in Ireland.

December 12, 2019 is fraught with risk for every political party, holding as it does the tantalising prospect of a breakthrough or an unwelcome date with the slippery slope of defeat.

Boris Johnson in a sense has most to lose with the debit side of an election defeat reading out of office with his Brexit deal in tatters and personal authority shot to pieces.

Conservative strategists who are angst ridden about Nigel Farage are nervous with good reason. A populist guerrilla campaign by the Brexit Party could deny Mr Johnson the majority he so badly craves.

Jeremy Corbyn will have a sense of déjà vu. Written off in 2017 he managed to take Labour's vote share to over 40% and proved to be the author of a tale called Tory Troubles. He knows that failure to deliver a hung parliament at the very least will probably spell the end of his leadership. This is the election where Labour wants the agenda to be more about bread and butter issues and less about Brexit.

The Lib Dems, even by the standards of a party desperate to talk up their chances, seem to be hovering in an orbit of the delusional. Jo Swinson parrots that she wants to be Prime Minister with the apparent conviction that she actually believes it might happen.

Back in 1983, the then SDP-Liberal Alliance polled 25% of the vote, far higher than current Lib Dem poll ratings and delivered just 23 MPs. She could literally put on millions of votes from 2017 and still be a million miles from Downing Street, such is our electoral system. Her best shot is that her party can wield influence in a hung parliament with a majority of MPs who back a Remain position.

For the SNP this is the election they hope is a springboard to IndyRef2. If they progress by winning more seats and votes than in 2017, then any Westminster government will have to revisit the issue of a second referendum, despite what the leaders of the UK parties currently say on the issue. Holding a line in an election is one thing, dealing with a post-election reality is quite another.

In Northern Ireland we have witnessed unprecedented manoeuvres by all parties to maximise votes for Remain and Leave positions. More so than any other part of the UK we are reminded that this is an election with one huge constitutional question dominating.

Election night will be unpredictable and I do not envy the brilliant Professor John Curtice his task in overseeing the exit poll and more importantly predicting what it all means in terms of seats won.

Most exit polls are pretty accurate (with the exception of 1992 when it predicted Labour would be the largest party) but the dynamics of this election make firm predictions really difficult. In Professor Sir John, however, I trust, even if the exit poll predictions are likely to come with more than the usual caveats.