AS I sit down to attempt to forecast the outcome of the Holyrood election, I'm haunted by the knowledge that almost everyone who engaged in the same exercise just before the 2011 and 2016 elections got it wrong. 

That includes myself - despite issuing warnings in the closing days of the 2016 campaign that the SNP were in more danger of losing their overall majority than most people realised, I hoped and believed that they would just about make it. 

This time the situation is a little different, because it's widely accepted that a hung parliament is a significant possibility, although the most recent batch of opinion polls has raised hopes that the scandals surrounding Boris Johnson may be obligingly handing the SNP a majority on a plate at the very last moment.

Nevertheless, given the recent history of electoral volatility and unpredictability, it's always worth asking the question: "What's the surprise going to be?"  There will almost inevitably be one. 

The nastiest possibility, which too few people are taking seriously, is that we could lose the pro-independence majority altogether - in other words, the SNP, the Greens and Alba in combination could fall short of 65 seats.  That may happen if the SNP lose a significant number of constituency seats, which given the first-past-the-post nature of the constituency ballot wouldn't necessarily take all that big a drop in the SNP's vote.  Remember that the public are close to being split 50/50 between pro-indy and Unionist parties, so if the SNP lose their in-built advantage on the constituency vote, there'd no longer be any strong reason to expect the overall outcome to be a majority for Yes parties.  However, with the latest Opinium poll showing the SNP on 51% of the constituency vote, hopefully the danger is relatively slim.

READ MORE: Opinium puts SNP on track for a majority government

For some people, Alba winning any seats at all would be a major surprise, but it really shouldn't be.  No party in the history of the Scottish Parliament has failed to take at least one seat on 3% or more of the national list vote - and the majority of polls so far have shown Alba on at least 3%. 

As many as three polls have shown them as high as 6%, which is the sort of level at which they'd be taking a seat in most regions.  Even if the polls are accurate, there must be a 50/50 chance or better that there will be an Alba presence of some description in the new parliament.

The Greens are virtually assured of a presence, but they have their hopes set much higher, and are looking to significantly exceed the record return of seven seats they bagged way back in 2003.  Many polls suggest they'll do exactly that, although their reliance on the list means that a relatively small error in the polls, or a modest late swing against them, or a disproportionate failure of their supporters to turn out, could cost them dear. 

We've been here before - in both 2007 and 2011 the Greens were confidently predicting that they would end up with their biggest ever haul of seats, and in both cases ended up with only two.  For what it's worth, though, my gut instinct is that they won't be disappointed this time.

POLL: What would be the Yes movement's best outcome on May 6?

The other big question is which Unionist party will end up in second place - will it be Labour or the Tories?  Most polls suggest the latter, but there's been the odd exception - a Survation poll a couple of weeks ago had Labour two points ahead of the Conservatives. 

The National:

I suspect most people in the SNP will be secretly hoping that Douglas Ross does emerge as the leader of the largest opposition party, because having him as the public face of any future No campaign would be the dream scenario. 

Polls showing the Tories with a lead over Labour give a misleading impression of the two parties' underlying popularity, because Labour are far more people's second choice.  The Tory brand remains hopefully tainted for the majority of the public.

READ MORE: Wee Ginger Dug: This is the big risk in how the Tories played this election

Usually, at least one party leader resigns after an election.  Logically, looking at the polls, the most likely candidate for the chop ought to be Anas Sarwar.  In practice, it's hard to imagine that happening, because so many of his colleagues and his admirers in the media are convinced he's the right person for the job.  But if he loses seats, and does worse than Labour managed under Jeremy Corbyn and Kezia Dugdale in 2016, what exactly is his alibi?

Here's my best guess for the result: SNP borderline for a majority (perhaps somewhere in the 63-67 seat range), the Tories in second place, the Greens making modest gains, and Alba with a presence.  And based on past history, I also confidently predict that my guess will be completely wrong.