The overall message of what is now just a hobby is this - very small initiatives that are not obvious can be crucial election-winning strategies.
When I started this practice, it was usually necessary to laboriously type in large quantities of data but these days that step is removed by web-based publication of all of the necessary figures and with the final voting statistics for the September 23rd now available, some interesting results are beginning to emerge.
Like nearly every MMP election, the 2017 General Election produced an extremely close result.
National was always going to be the single biggest party so its key objective was to win enough votes so that New Zealand First would have no choice but to join it in a post election coalition.
This was very nearly achieved but two shortcomings, one of which could have been easily countered, stymied this outcome.
So what were the little differences that might have worked for National and got the party a fourth term?
It is arguable that on election night, with a one seat advantage to the opposition parties (including NZ First), National would have been the only possible choice for New Zealand First as a one-seat majority would have been just too fragile.
This situation changed when the special votes delivered an extra seat to both the Green and Labour Parties at the expense of National.
This was unusual. National has lost one seat on the special votes in the past but never two.
Not surprisingly, National's deficit in the special votes popped up as a key factor in the final outcome.
Success in winning special votes is related to the strength of a party's on-the-ground activities and as this aspect of campaigning has never been important to Crosby Textor, National's Australia-based campaign strategists, it's not too surprising that the party remains weak in this aspect of its organisation.
In Australia where Crosby Textor is based and cut its teeth, compulsory voting means that rushing about getting voters to cast special votes hardly features in campaigns.
It will be interesting to see if Crosby Textor is engaged by the National Party for the 2020 general election.
This company was closely involved in the Conservative Party's loss of its majority in the recent British general election so its magic is fading badly.
After chewing its way through the election numbers, the software aggregates the party vote, then electorate vote in all electorates, finds a median and a mean and then looks for electorates that fall outside those boundaries.
In the past this exercise has pointed at particularly attractive or hard-working candidates, effective new campaigning techniques, key local issues or some reason the have a close look at what went on.
Predictably, this process picked the now Prime Minister Jacinda Ardern as a serious electoral asset for the Labour Party very early in her political career.
This analysis has demonstrated over all of the MMP elections something that politicians instinctively know: that there is a relationship between fielding a candidate in a given electorate and the party vote result.
It stands to reason that if a party manages to get some to carry their banner as a candidate in an electorate that the party is likely to benefit from this activity by way of an improved party vote. This is the main reason that the Labour Party makes every possible effort to field a full slate of candidates.
National fielded candidates in all of the general seats and scored an average party vote of about 18,000 votes.
However, with no candidates in the Maori electorates, National's party vote average in those seven seats was a pathetic sixteen hundred votes.
With the defeat of the Maori Party and the wastage of that party's party vote, National will be ruefully reviewing its clearly misguided strategy.
When National had a policy of abolishing the Maori electorates, not fielding candidates made some sense.
Now it certainly doesn't.
It helps lose elections.
