This is the fourth party leadership contest in a row that Mainstreet has attempted to poll and got totally wrong...though not quite as totally wrong as last weeks Sask Party leadership. Honestly why do they even bother trying to poll leadership contests? They get it dead wrong every single time and it’s bad for their already damaged reputation
It's impossible to poll leadership elections unless you have what you did in the last federal Liberal leadership race where someone runs away with it. What you can do is poll how each candidate would fair in a general election, otherwise whom would be best for the party to choose but that doesn't mean they will follow. The problem is in most leadership races, less than half of the members actually vote. Also you have to ask about second and third choices as those play a major role (Just think if Michael Lee only got a few more second choices of Stone and De Jong he would have knocked off Wilkinson and probably won it). Also every riding has equal weighting so a riding with only 10 members has the same weighting as one with 1,000 members so pretty much you would have to survey every member to even get a slight indication. Otherwise unless there is a clear favourite, its a crapshoot. Now I think in the last federal Liberal leadership race where Justin Trudeau got 80% on the first ballot, a poll would pick that up, but that was a blowout not close like the recent races have been.
FWIW, Mainstreet says they poll leadership races based off of past federal and provincial donor lists. It could be that the federal CPC component of the BC Liberals were over-emphasized in polling. It's very possible that a similar thing happened in the Sask Party Leadership where Mainstream had the more conservative Cheveldayoff ahead.