Earlier this week the Political Studies Association provided a General Election briefing to journalists in London. Scotland, Wales and Northern Ireland were all covered (the lattermost by your correspondent). Professor Sir John Curtice - who leads the admirable team which has produced such accurate exit polls at recent elections - gave an overview of the contest.
His overall conclusion was that Labour's chances of achieving an overall majority are close to zero - but that does not necessarily preclude the possibility of Jeremy Corbyn being your next Prime Minister, via a (no doubt unstable) 'rainbow' alliance.
This would be an informal government by pact, not coalition, of many parties, operating as a minority administration.
Regarding the exit poll, Sir John and his colleagues must be hoping that the dark December nights will make people vote earlier in the day. Election results can alter with evening voting. The Conservatives had an overall majority in 2010 until deep into the day but a late surge of Labour voters meant a no overall control outcome.
The exit poll team uses the same polling stations each time where possible. It gets voters to complete a ballot paper identical to the one they have filled inside the polling station and measures change from the previous election. Obviously the team will be hoping for mild and clement weather come December 12, as voters already annoyed at having an election for a Christmas present might not want to be hanging around to vote again for the exit pollsters.
Only a fool has taken on the exit poll in recent contests, such has been its accuracy. Only a complete and utter idiot would take on the task of predicting Northern Ireland's constituency results, given the likely closeness of several and that we are weeks from the contest. So here goes.
West Belfast: Sinn Fein. Strangford: DUP. That do? Okay, let's look at those constituencies which might conceivably change hands. North Belfast is perhaps the most interesting but hardest of all to predict, between two prominent and personable candidates.
The bookmakers cannot split the chances of Nigel Dodds and John Finucane, calling it a dead heat. If two-thirds of former SDLP and Green votes go to Finucane in the absence of those parties, Sinn Fein take the seat. I'd still call it for Dodds but without the slightest confidence.
Perhaps equally hard to call is Foyle, where I'll be spending the weekend at Sinn Fein's ard fheis. Not according to the bookmakers though. They now make Colum Eastwood a long odds-on favourite to capture the seat. Given the SDLP's proud history there, Eastwood's status as party leader and encouraging recent polling for his party, this is understandable. Sinn Fein could be a value bet to hold on though.
The SDLP should enjoy success in South Belfast anyway, unless absentee Sinn Fein's vote surprisingly heads over to Alliance. Meanwhile, Naomi Long's candidature in East Belfast is capable of springing a surprise victory, not on the scale of her remarkable 2010 win there, but a real coup on a minimum 10% swing nonetheless. Where else could surprise? The UUP desperately needs Westminster representation and while much of the focus will be upon Fermanagh and South Tyrone, Sinn Fein look strong favourites and it is just possible that South Antrim - scene of joy in 2015 - represents a better chance.
DUP difficulties in Belfast may possibly be assuaged in the previously unlikely domain of North Down.
Wriggle room is obligatory for all of the above and we've only just sorted out who is standing, let alone discussed any policies. At the 2017 election, the Northern Irish electorate identified the most important issues as the constitution, Brexit, the NHS and education. We've already had a DUP election broadcast which didn't even mention one of those words (it began with B) but I suspect the other parties won't be making a similar omission.
Overall, I'd make it DUP 8, Sinn Fein 7, SDLP 1, Alliance 1, UUP 1. Everyone happyish, perhaps. But I reserve the right to revise all predictions on December 13 and to ask the Belfast Telegraph to remove all traces of this article. Political science is really about measuring what has just happened to make predictions - the exit poll - or explaining why voting patterns are occurring - not making rash forecasts.
Even explanations of contemporary voting behaviour can be tricky.
Wish I'd been a historian.