[Kerplunk]

Originally Posted by kirant

It is worth noting that, according to people who have it, it is apparently in line with the paywalled Mainstreet daily polling data (with the last Mainstreet polling apparently showing an NDP lead).

I'm not fully aware about how accurate Mainstreet is (they apparently did a review and made significant changes to their techniques after bungling the Calgary municipal election, with "proof of concept" results made when they accurately predicted Ontario's provincial election)...using the previous tracking between the last election an this one, they seem to be favouring the UCP a little when compared to similarly timed Abacus and Leger data while being less favouring of the UCP than Lethbridge College. But it could very well be a reasonable result and a realistic possibility. I think the probability of a close race being fact increases if we get more and more agencies reporting the same result.

On reddit, it sounds like the stats crunchers still have it in the UCP's favour with the NDP needing to make gains in UCP Calgary strongholds to win the election.

I still take Mainstreet with a block of salt as well. I seem to recall reading that they corked up with a recent BC byelection as well, so whatever is wrong isn't fixed.

I'm interested in the methods they used for the online polling, out of sheer curiosity as an IT guy, but I haven't had time to dig around for it yet.