Ontario General Election Prediction thread

[adma]

Exactly.  If you look at the results of Vaughan in 2006 and 2008 vs. 2011, it swung much harder towards the Tories than most ridings did.  I describe it as essentially a Blue Liberal/Red Tory with a weak hard right base, but also weak progressive base too thus why you can see these strong swings.  In many ways more akin to Winnipeg South or Saint Boniface in Manitoba which see similar swings although very different demographics or perhaps even the North Shore of BC is somewhat comparable.

Though you're not acknowledging a major factor behind that hard swing--Liberal MP Maurizio Bevilacqua stepping down to run for Vaughan Mayor, and his replacement by Conservative Julian Fantino in a 2010 byelection squeaker.  Had Bevilacqua been defending the seat in 2011 (rather than Fantino defending his byelection win), I suspect the race would have been much tighter.

Oh, and re "weak progressive base", I think it might have been the only federal Ontario seat in 2015 where the NDP scored less than 5%.

[adma]

One thing I'm wondering: why does "favourable distribution" favour the PCs in Ontario, while it favours Labour in the UK?  Or is it the built-in flaw of Ontario's projection models...

[adma]

100% agreed. The NDP have really solid strongholds which is helpful when you're in the low 20s like the NDP normally are, but it hurts you when you're tied or marginally ahead like the NDP are.

Except that in "wave elections" like the present, the gains can be more emphatic in the less-than-strongholds than in the status-quo strongholds--something which conventional projection models may have difficulty picking up.

To take an extreme case, the Mainstreet poll for St. Paul's showing the NDP nudging the Libs for 1st with a third of the vote; last time, they barely crossed the 10% threshold there with what was basically a paper candidacy.  By comparison, I can see the share remaining relatively stable or possibly even declining in "2014 overperformance" seats like Essex...

[adma]

What does "running up the margins" even mean for the NDP?

I don't know; Taras Natyshak 2014 vs 2011?  Certainly wasn't Rosario Marchese in those respective years.

And besides, we've already debunked running-up-the-margins logic relative to the 2015 Alberta election (i.e. the presumption of an inflated NDP vote all wasted in Edmonton and Lethbridge, because 2012 tallies suggested the same)

[adma]

the NDP wins a number of affluent Toronto ridings, including Don Valley West (this is a very unhinged prediction but I'll stick with it).

I think the NDP winning DVW would require the Libs sinking to single digits provincewide.

[adma]

I'll throw out another prediction: the Libertarian Party will break 1.5%

With Trillium, Freedom, and Ford also in the picture, I'd reckon that a touch high--though they do now have enough candidates to make such a share "doable"; and who knows if there'll be "Renatagate" blowback...

[adma]

And speaking in terms of before whatever Renatagate may portend: I don't like to *predict* outcomes, but an outcome I was *preparing* for was the PCs going back into the low 40s sharewise, and the NDP sinking back to the low 30s--something which'd accord more with seat count projections combined with Mainstreet-style riding polls.  Which'd be a bit of an Alberta 2012/BC 2013 "surprise" result, yes, I know, but...

[adma]

Funny.  I've been saying all day that I'm *preparing for* a 90 PC/30 NDP kind of result.  But that's a preparation for, not a prediction--I'd rather be open-ended (even on behalf of a worst-case scenario), than be skunked with a woefully incorrect guess.  Likewise a week or so ago, I was "preparing for" the PCs to get back past the 40% threshold and the NDP to settle into a "normalized" low 30s--all of that is plausible given ground-level evidence and, as it turns out, late polls.  But I'm also prepared for all of that *not* being the case...

[adma]

I was quite a bit off on both predictions, but I don't think anyone was expecting such a huge majority for the PCs..

In the end, it wasn't *that* huge or unexpected, especially considering the final (herded?) polls.

But I do find that a lot of polling predictors seemed to be pointing at a 76-to-40 type of result out of a much smaller vote differential than the 7-point margin that actually transpired...