Second, when you looked at the timing of the poll - known as the capture period - it could hardly have been better for Labour as it included Jacinda getting feted as a sensation during her first overseas trip of the year - including the Davos exposure with Prince William and Sir David Attenborough.
Equally, the timing was not so good for National spanning Paula Bennett's off-the-wall marijuana fancies.
Third, Judith Collins sneaked past Simon Bridges in the preferred PM stakes moving up 2.5 per cent to 6.2 per cent while Bridges lost close to half his support, dropping to just 5 per cent.
This is bound to create tension within the National Party caucus, but the poll was quickly followed by a Judith Collins press statement in which she denied any interest in Bridges' job.
Probably the most interesting aspect of this poll was the decline of minor parties and an apparent drift back to a two-party system.
Neither Act or the Māori Party were reported as scoring any support at all, leaving only the Green Party with enough support (and then only barely) to return MPs to Parliament and New Zealand First at just 3 per cent and supposedly on course to depart from Parliament.
Before National dumps Simon Bridges for Judith Collings, Labour plans a big celebration, and various minor party denizens renew their CVs, a few observations should be made.
The polling industry is passing through a period of dramatic change which must, by definition, make their results at least questionable right now.
About 30 years ago telephone polling replaced face-to face surveying over a very short period, and a similar watershed period is occurring right now.
The switch to telephone polling in the 1980s was largely driven by cost, but it was only made possible by the fact that landlines had, by then, been installed in almost every household.
The communications landscape has changed dramatically since the 1980s with the growth of cellphone usage, the now very high levels of internet penetration and the slow decline of landlines.
We know that Colmar Brunton has changed its capture methodology so that 50 per cent of its respondents are landline captured and the other 50 per cent from randomly generated cellphone numbers.
The TV3 Newshub poll by Reed Research uses yet another formula with 75 per cent of its sample coming from landline responses and 25 per cent from the internet.
There are yet further methodologies in the field including pure online surveying and even robo-polling where responses are via the keypad on telephones ("press one to vote Labour, etc").
Robo-polling for politics is now used in Australia where the state and federal system of government means elections are much more common than in New Zealand.
It appears to generate reliable predictions, though certain strengths of human to human based polling like open ended questions are obviously not possible or difficult.
Having launched the first successful telephone-based polling company in New Zealand many years ago with an Australian partner and Peter Beaven, now a Hawke's Bay Regional councillor, I have taken a close interest in political polling and kept in touch with some key players in the industry.
I always look at poll results with a somewhat jaundiced eye and I'll share the following prejudices developed over many years.
Apart from its first post-Jacinda survey, the Colmar Brunton poll seems to have a slight bias in National's favour. I found that if I took a point off National and put it on Labour, that usually brought it into line with the aggregate of the other polls.
The Newshub poll usually gets the big parties about right but can't seem to catch the minor party voters.
All polls underestimate the New Zealand First vote.
The only big message out of this poll is that National still has no friends.
* Mike Williams grew up in Hawke's Bay. He is CEO of the NZ Howard League and a former Labour Party president. All opinions are his and not those of Hawke's Bay Today.
