Saturday, September 19, 2015

#Elxn42....Attempting To Quantify The Number of Voters Willing To Vote To Stop The Con In Lotusland.

That's from a recent Metro Vancouver-focussed poll of decided voters in Metro Vancouver by Justason.

My opinion?

Nineteen percent of decided voters willing to vote to Stop-A-Con, with a 12 percent differential, is enough to make a real difference.

But only 'if'.....

And it's a really big 'if'....

Which is...

'If' they know who to vote for to stop the Con.

Which is why we need more riding-specific polling.




scotty on denman said...

I wonder if it's locally affordable to hire polling expertise---bunch of SV proponents in a given riding scrape up some funds and do it. Perhaps voters from solid, foregone conclusive ridings could chip in to help focus on ridings where SV is feasible. We are, after all, asking Greens to embrace SV in most ridings---just this once---and send their donations instead to ensure Ms May's reelection in Saanich-Gulf Islands; we might just as well ask progressive voters in Con-locked ridings to send donations to SV operatives in pertinent ridings. It would behove Greens, say, in most Alberta ridings---i.e., where Cons regularly win outright majorities---to assist in Harper's defeat; donating to SV-specific polling in pertinent ridings is another way of helping. Same goes for Libs or Dippers in these out-of-reach Con ridings.

I can tell you I'd trust a poll of specific design that I approve of a lot more than some bigger polls---we've seen how inaccurate, or, potentially, manipulated they can be. I think a lot of respondents give their preferred candidate or party without any intention of actually voting, or, more likely, without revealing that they might switch at last moment; some pointed, focused polling in tight ridings would be useful.

But I'm really out of my water here. Can anybody tell me if or how this might be done?

RossK said...

The LeadNow folks are attempting to do it scotty.


Steve Cooley said... has a page that attempts to estimate how each riding is leaning.

RossK said...



And I use it a lot (see this for example).

Even the author, Eric Grenier admits it is only good for predicting winners when things are pretty clear cut.

However, when you only have 'leaners' AND (most importantly) you are trying to figure out who is second, which is critical in Strategic Voting, you need riding-specific polls.