i need to do my undecided analysis, as i always do.

is this actually in play?

i picked this one out, but the numbers across the board are closer than you might expect, if still just outside of the margin of error. at a 7% spread, perfect distribution and a 3% margin, a minority is still the best possible outcome.

is there any evidence, though, that the undecideds might lean ndp?

you'd need some kind of shy dipper effect, and the article i just posted may suggest one. albertans may not realize the speed that the province is transforming at, as they seem to think that they're more conservative than they actually are. that is the kind of social pressure that may lead somebody to tell a pollster that they're undecided, when they really expect to vote ndp.

i would certainly not expect a shy tory effect, anyways.

if the redistribution is slightly tipped towards the ndp, you start getting into a blurry error bar. if it's a bit bigger, they start running neck to neck. and, if it's overwhelming, you're looking at an upset.

we'll find out soon enough.

but i wouldn't expect it split evenly, anyways - they will beat the polling, even if it's not enough.