The Folly of Seat Projections

The recent surge of the NDP into Liberal and Bloc territory seems to continue without end. Despite the doomsdayers the lines will stabilize (if they haven't already.)

In the mean time we're being treated to a spate of seat projections showing such varying dizzying results. Some Tories, uncomprehendingly, have come out claiming that a Tory Majority is now out of reach. Some believe the Bloc is dead. Others that the Liberals are on life support.

All of this could turn out to be true. But much of it probably won't.

What puzzles me the most about this situation is the almost complete certainty of some with regards to just what these polling numbers actually mean.

How could anyone know what they mean? What are they gods of election prediction? Do they have some sort of crystal ball that allows them to see magically where these new NDP votes are going? Seat projections are based on certain theories on how votes distribute themselves. They have a history of working well when parties poll within a few percentage points of where they did the last time around.

With NDP support not doubling, not tripling, not quadrupling, but increasing by factors of ten or more in Quebec and lesser so elsewhere who is crazy enough to create seat projections at all?

The truth is that no one really knows (or should claim to know) what the NDP surge in support will mean. These new votes could be efficiently distributed for the NDP - or advantageously distributed for others.

Given this uncertainty, the insanity of all this news coverage is apparent. I expect (or maybe I hope) that many will be surprised come May 2nd.

If not it will stroke the already over-bloated egos of some, and turn others into blind pollster worshipers.