I’ve been playing with my spreadsheet. It’s silly, fun, and a waste of time. That aside, I’ve taken the shift between polling numbers in 2011 (allowing for the current trends to continue until Monday) and voting numbers from the 2008 election for each party. I’ve applied the shift to the 2008 votes to predict the seat counts after Monday. Please do not take this prediction seriously because a) it is deterministic, not probabilistic; b) it’s based on national-level numbers, not regional nor riding-level; and c) it’s just for fun and d) it doesn’t redistribute any of the absentee voters, who seem to actually be motivated to vote this time out.
Because of the huge shift in national support for the NDP, and the blanket mathematical multiplication of votes used on the spreadsheet, some of the numbers for individual ridings predict insanely impossible NDP landslides, which means that the vote distributions are skewed heavily towards previously successful ridings for the NDP (because they are the party with the largest upwards shift of support). For instance, this bad-math approach has Linda Duncan winning by 20,000 votes in Edmonton-Strathcona. I think she’ll win, but not by such a silly margin.
Because of changes in candidates, it completely mispredicts some ridings (such as Central Nova, where no Liberal ran in 2008, and probably Saanich-Gulf Islands, and whatever riding Guergis is running in (Guelph?). Nonetheless, here are the seat counts from this prediction:
Please commence with the laughing now!