Northland voters carry big responsibility

The Northland by-election happens on 28 March. Under normal conditions, the National candidate, Mark Osborne, could sleep-walk to victory, given the 9,300 majority for former National MP, Mike Sabin, at the 2014 General Election. But, now that Winston Peters has put his hat in the ring, things have changed. If Mr Peters were to win, and were then to resign as a list MP, he would re-enter Parliament as the MP for Northland, and his list seat would be taken by the next willing candidate on NZ First's party list. Hence, NZ First's seats would increase by one, and National's decrease by one.
That would change the balance of power in the House, leaving National more heavily reliant on United Future and Maori Party for passing laws. (I assume that ACT is pretty much National's lapdog.)
But what are Mr Peters' chances of winning? He has the charisma and the name recognition, and the advantage of being a Whananaki boy.
NZ First didn't run a candidate in Northland at the last election, but they picked up 4,500 party votes anyway. That's not a bad start. But it looks like most of those NZF voters gave their candidate vote to Labour's Willow-Jean Prime (who is re-contesting at the by-election).
Assuming Northland gets the same level of voter turn-out at the by-election, then we could generously assume that Winston starts with 4,500 voters in his pocket. (And that's also assuming that none of those will vote for the Labour candidate). To win, then, Peters has to take roughly 7,000 voters away from the National candidate, reducing National's vote from 18,200 to 11,200, and increasing his own from 4,500 to 11,500.
Once Northland's more conservative, National-supporting voters realise the consequences of voting for Winston, it seems highly unlikely that 7,000 of them will follow him like some Pied Piper. There is not enough popular discontent with the Key government at this stage for those otherwise loyal to Key to want to stymie him, or 'to send him a message', at a by-election.
Of course, Peters could pick up a good number of those who otherwise vote Labour or Green, seeing in him an opportunity to undermine the Key government. But, at the last election, the Labour and Green candidates added together got about 12,600 votes. That's considerably fewer than Sabin's 18,200. Even if we assume that Mr Peters' 'handicap' starts at (a very generous) 8,000 voters taken from the left, he still has a huge job winning more than 5,000 from the right at the same time.
One learns from the past not to underestimate Mr Peters. But, on this occasion, I doubt that he can pull it off. The safe money would be on a National win. IPredict says it all.
If he loses, Mr Peters at least gains a chance to attract attention to himself. That may help him to build up some support for the 2017 general election, by which time his embarrassing defeat will have been forgotten anyway. It will be interesting to see whether he pitches to the left or to the right of the Northland electorate. He can lean either way.