Well, no, they are pretty bad still, but I take back what I said yesterday about AR - at least in terms that when viewed in the context of AR always giving Liberals the short end of the stick anyways, their most recent poll isn't as horrible as it seems.
The topline numbers paint an always-nasty picture: 39% Con, 26% Lib, 18% NDP, 9% Bloc, and 6% Green. If you follow the standard rule of Canadian politics that sticking around 40% means you're likely to get a majority, this should please the Conservatives. Indeed, according to 308.com, this poll would give them 150 seats - just 5 short of a majority, and well within any acceptable margin of error.
But looking at the regions, AR does not buck any trends that their friends two weeks were showing, with increased Liberals numbers in the West, Conservative solidification in Ontario, poor results in QC for everyone but the Bloc, and something of a horse race in Atlantic Canada.
I'm also more inclined to give AR the benefit of the doubt in the case of their numbers, given that they surveyed over 6,000 people, and then broke down the results into every single province - meaning Saskitoba finally becomes separate provinces! Same with the Atlantic provinces too. Those results give the Liberals a higher score than 2008 everywhere but Newfoundland, where we've apparently dropped (though not to the point that John Ivison touted). Of particular note is the 41% score in Nova Scotia (we got under 30% last time), while the NDP sit at only 21% in the province (they had 28% before), as well as the perennial third party being down in Manitoba also with 21% (they had nearly 25% before), showing us that Darrell Dexter and Greg Selinger are not possibly the greatest assets to the federal party right now.
All this taken together means that within context of what AR usually gives us, we're not doing too bad. It'd only be a 2-point drop from their last poll, while the Conservatives gobble up support from the other parties. Consider as well that back in 2008, AR polls that put the Cons between 38-40%, had the Liberals sitting farther back, from a low of 21% (!) to a high of 28%, the rest averaging out to 23-25%. That's something of an improvement.
This isn't to say that Angus Reid doesn't stir up the skeptical side within me. Their big cavaet is the 43-30-19 split in Ontario, which is the source of the strong Conservative vote, while relatively strong results for the Liberals everywhere but Central Canada keep us from total oblivion. These results are pretty much what every other poll showed during that one week, but the problem is that they were in the field last week, when Ekos had a "corrective" poll that showed the Conservatives dropping and the Liberals rising ever so slightly. Given that I definitely don't trust the Abacus numbers, it means we have a case of dueling polls showing two very different realities. Who is right on this one?