ANY independence supporter who remembers how the polls suddenly tightened in the run-up to polling day five years ago will have had a sickening sense of familiarity when they saw the new Savanta ComRes poll, which suggests the SNP could once again fall slightly short of an overall majority, and that the combined advantage for pro-independence parties is not as comfortable as we’d like.

But any worries will have been at least partly eased by a YouGov poll, conducted over exactly the same period, which shows the SNP on course for an overall majority and pro-indy parties heading for a very large combined majority. One might almost be tempted to call it a “supermajority”.

So why the contradiction? The voting shares reported in the two polls are actually not light-years apart, but the impact of the differences is magnified due to the quirks of the voting system. It must be remembered that the Scottish public are more or less evenly split between voting for Unionist and Yes parties, and the only reason a pro-indy majority in the new parliament is widely expected is that the SNP enjoy an in-built bonus due to their dominance in the first-past-the-post component of the election. But the smaller their lead over the second-placed party on the constituency ballot, the weaker that bonus becomes.

READ MORE: Two new polls predict pro-independence majority at Holyrood election

YouGov suggest that the SNP’s constituency lead has grown from 26 percentage points to 28 points, courtesy of the good fortune of the combined Tory and Labour vote being exactly split down the middle in a way that wasn’t the case previously. In the seats projection, that’s enough to ensure that the SNP reach the 65 seats required for a majority on constituency seats alone, before a single list vote is even taken into account.

By contrast, Savanta ComRes suggest that both the Tory and Labour vote shares have risen simultaneously, allowing the Tories to maintain their clear second place and cutting the SNP’s lead to 21 points. That still sounds enormous, but it means the SNP are unable to quite reach 65 seats in the constituencies, and the list seats simply aren’t there to make up the shortfall. Despite taking 38% of the list vote, the SNP are projected to take just three of the 56 list seats across Scotland – a drop of one on last time.

The other key difference between Savanta ComRes and YouGov relates to the standing of the Greens. YouGov have them surging to 10% of the list vote, allowing them to take lots of list seats and put generous icing on the cake of the SNP’s overall majority. But ComRes have them dropping back to 7%, leaving less scope for SNP disappointments on the constituency ballot to be offset by Green list successes.

The net result is a pro-independence majority in the parliament, but a much more modest one.

Both polls agree that the new Alba Party is falling short of the 5% or 6% that would be required to take at least one list seat in most regions. However, it must be remembered that the two Panelbase polls that have been published in this campaign so far have both shown Alba on 6%.

The disparity can probably be explained mostly by methodological differences between the polling firms, and it remains to be seen who is getting it right.

Intuitively, the story YouGov are telling seems to "fit" better than the ComRes version. The general consensus is that Patrick Harvie, Lorna Slater and Anas Sarwar performed better in the TV leaders’ debates than Douglas Ross, so a Green surge and a swing of Unionist votes from Tory to Labour would appear to make perfect sense. However, what seems intuitively likely often bears little resemblance to what is happening on the ground, so we’ll have to wait for more polls to give us a definitive answer.