i'll reiterate: the tories usually poll at the bottom of their error bars. this generally leads to exaggerated predictions that fail to materialize. tonight, they polled in the middle of it. this is an exception to the rule.

the tories have never won a majority government with less than 40% of the vote - and have ended up with leads in the 30s many times, always ending up with minorities. this has also repeatedly produced many exaggerated predictions. but, tonight was the exception to the rule.

while red/blue switchers were the dominant voter in the twentieth century, they have not existed in large numbers for the course of my conscious lifetime. liberal voters simply haven't voted for the conservatives in large numbers in decades. but, tonight they did.

& what the data suggested, when adjusted for all of these known historical errors and biases, was that we were looking at a small pc bump (not much more than the margin of error) that was distributed equally across the province, as well as exaggerated bumps in ndp support in urban ridings, which should have produced big ndp gains in urban seats and not much for the conservatives, who had a less concentrated vote. while we saw these big ndp bumps in the very downtown cores, we also saw large conservative bumps in suburban areas that i maintain was not something that was predictable from the data, besides through projection. but, it happened, nonetheless, whether via projection or via something else.

i'm not going to adjust my methods over this. i'm going to shrug it off as weird.

and, as mentioned repeatedly, both before and after the vote, as very fishy.