I believe it was a huge error of Elizabeth May to choose Saanich Gulf Islands as the riding to run in as the numbers are not there for her to win it.
In her two previous runs for office as Green Party leader she has not managed to get past 1/3 of the vote. In the November 27th 2006 London North Central by-election she achieved 25.84% of the vote. In the 2008 election she ran in Central Nova and managed 32.24%, she benefited from the lack of a Liberal candidate in Central Nova.
In Saanich Gulf Islands she is running against Gary Lunn, a five time winner in the riding. He has ranged from 34.57% to 43.43% of the cote in the riding. He has achieved 43% of the vote three times out of the five elections. It is not realistic to think his vote will drop below 38% and the safe bet is to look at him getting 43% of the vote.
In the last election the Liberals did everything they could to try and win the riding with Briony Penn. They managed to get the NDP candidate to resign, though his name was still on the ballot. The pressured the Greens not to run. With all this the Liberals could not get to 40% of the vote.
So what is the theoretical amount of votes Elizabeth May could get? Let us assume that the NDP and Liberals finish as they have with their lowest percentage when running an active candidate - that is 18% for the Liberals in 1988 and 8% for the NDP in 2000. This means 26% of the vote is not available to Elizabeth May.
This leaves 74% of the vote for Gary Lunn and Elizabeth May. At 43% of the vote for Gary Lunn, Elizabeth May is at 31%. For Elizabeth May to win, she needs to 37.5% of the vote, meaning Gary Lunn has to fall to 36.5% of the vote. I am not all convinced that Gary Lunn will lose that much of his vote in this election. It also means Elizabeth May, a parachute candidate, has to better than she has ever done before and this time she has a Liberal candidate as well that feels like she is within range of winning.
Realistically the Liberals will do better than 18% and the NDP better than 8%. If takes only a small increase on either parties part to make it mathematically impossible for Elizabeth May to win.
The NDP and the Liberals are going to everything can to make sure Elizabeth May does not win, it is not in their interest to have her win a seat. I expect Renee Hetherington of the Liberals to spend the maximum in the campaign.
I expect Gary Lunn to win by more than 10 percentage points and for Elizabeth May and the Liberals to finish roughly even.
8 comments:
These are not two polls it would seem, the one is a correction of the data on one of the polls.
The poll is meaningless unless there is full data - who did the polling, when it was done and how many people answered. Given the 4.8% stated margin of statistical error, which in the case of SGI is not right for a host of reasons, the sample was very small.
If this poll is to be believed, one in four Conservative voters is voting for the Greens.
The result of the poll is an indication of methodological errors, not of an accurate representation of the public
There are approximately four polls for Saanich-Gulf Islands Riding. I believe there is also a dirty tricks campaign attempting to invalidate those results. I shall attempt to provide a complete list of the four as soon is we can locate archive copies of those that have been removed. Let me explain what appear to be deliberately misleading dirty tricks.
The "oversampling correction" was issued by GreenPartyStrategies, an organisation competing with the Green Party, or more specifically opposing Elizabeth May. Furthermore the correction does not refer to the Green Party's poll, but to GreenPartyStrategies' own. They were dissatisfied with it because it confirmed that Greens were neck and neck with Cons. They failed to identify which poll they are correcting, and then added to the confusion by making their own poll disappear, thus implying that the corrections were to the poll commissioned by the Green Party. This is an important distinction, and highly misleading.
The poll commissioned by the Green Party, was conducted by McAllister Opinion Research, and their sampling has not been questioned.
I believe Green Strategies used a firm from Montreal, and while it is their right to question their own data, the point they make, is that they don't trust their own poll. They should have withdrawn it entirely, rather than trying to fudge the numbers, hiding the original, and then implying that they were doing this in the Green Party's name, or that they are referring to the Green Party's data.
The article on the Green site, citing the McAllister poll, is from September 2010... why on earth are people referencing that poll result? What is relevant is how people are actually going to vote now that there's an election campaign under way.
www.projectdemocracy.ca
lunn..............22,160
may...............19,710
heatherington.....15,816
loring-kuhanga.....4,957
This riding is not one the model predicts well. The Liberals have a far weaker candidate than Briony Penn this time out, and Elizabeth May has been campaigning hard for a year in the riding. It appears that anti-conservative vote is rallying around Ms. May who has broad appeal to NDP and Liberal voters. Progressive voters should rally around Elizabeth May as the best chance to defeat Conservative Gary Lunn.
To account for the May factor and weaker Liberal candiate we are applying an adjustment to the model to reflect these external realities. It is difficult to know for sure and this is a riding we intend to poll specifically in to make sure we have things right before election day - so check back with us
You are assuming that
a) people who did not vote in previous elections will continue not to vote in this one. Given May's emphasis during her campaigning on democratic reform, it seems entirely possible that she could bring out people who previously felt too frustrated with their options to vote.
b) there has been absolutely no change in demographics in this riding since 2008. That is not necessarily the case.
Honestly, I expect voter turn out to go down in this election. There is not sense of engagement of the public in the election. There is no indication that more people will be voting in Saanich Gulf Islands this time.
You also have to remember that Elizabeth May could only manage 32.2% in Central Nova last time around and she had no Liberal running. The Greens were also polling much better in that election.
For Elizabeth May to win she needs to exceed 40% of the vote. This means the NDP candidate has to take no more votes than last time when the NDP candidate quit and did not campaign. Also the Liberal candidate has to worse than the Liberals have ever done in this riding.
The electoral math simply does not work.
Elizabeth May needed to run in a riding with an existing three way race. Victoria would have been a good choice, all three parties have come 1st, 2nd and 3rd in this riding in the last 25 years. She could have won Victoria with 30% of the vote, an achievable number for her.
good predictions bro
You lose! I totally thought you were right, though. What do you have to say?
Post a Comment