Prompted by the Angus-Reid poll which gave me a huge, and seemingly impossible, 17 NDP seats in British Columbia, I decided to put some 'ceilings' on how well each party can do in each region.
These ceilings were determined by looking at each of the parties' best performances in each of the regions, and then adding to that best performance the amount of ridings in which the party placed second with less than a 10% margin between them and the winner. I only looked back to 2004 because prior to that the Conservative Party did not exist. With the vote split of the Canadian Alliance/Reform and the Progressive Conservatives, the results prior to 2004 are distorted.
So, for example, the best performance of the Conservatives in British Columbia was in 2008, where they won 22 seats. In another eight ridings they placed second, within 10-points of the winner. That means that the Conservative ceiling in British Columbia is 30.
The Liberal ceiling in British Columbia is 12, as they won 9 ridings in 2006 and placed second within 10% in three other ridings. The NDP ceiling is 16, so apparently my projected result of 17 wasn't off the mark by much. In 2006, the NDP won 10 seats and placed second within 10% of the winner in another six ridings.
Taking all of these ceilings together, we get an idea of the maximum probable result for each of the parties:
Liberals - 177 seats
Conservatives - 173 seats
Bloc Quebecois - 65 seats
New Democrats - 52 seats
What is striking about this result is that the Liberals and Conservatives only have a 20-seat margin to work with in order to get a majority. That doesn't leave a lot of room for error.
It's also surprising to me that the Bloc can go as high as 65 seats (in 2004 they won 54 seats and placed second within 10% in another 11). That would be a virtual sweep of the province. The NDP, at 52, can only realistically hope to place third among the parties.
Now, of course, my choice of 10% is completely arbitrary. We could go up to 12%, 15%, 20%. We've certainly seen swings like this in the past. But I feel 10% is a realistic mark to set, and it is extremely unlikely that any party would be able to do so fantastically well that, not only do they overcome the 10% gap in all of those extra ridings, they overcome even larger gaps elsewhere.
Within the realm of the rational, I think these are good ceilings to set - at least until after the 2009 election, when everything will have to be re-jigged again.