Project Democracy is a new website designed to help those inclined to vote strategically to prevent the Conservatives from winning a majority in the current federal election. In that sense it may not be the most aptly named web initiative out there — many, of course, will argue that strategic voting is undemocratic — but I suspect its founders are aware of this tension and hope to capitalize on it. For better or worse, this method of voting is often exercised in the election of representative democracies — though one suspects not always with the requisite information to make it altogether effective.

The way it all works is pretty straightforward. By providing historical statistics and predictions on how each riding will ultimately vote, the website helps those considering the possibility of casting a strategic ballot to decide on the non-Conservative candidate in their riding (which can be located on a map or by typing in a postal code) who has the best chance to win. There's nothing particularly revolutionary here, but assuming the predicative information is mostly accurate, Project Democracy should help users to avoid casting misguided or completely ineffectual strategic votes.

"[Project Democracy] takes the spin out of co-operative voting. It let's the numbers do the talking. All we have to do is listen," argues Alice Klein, co-founder of the project and Editor at NOW Magazine. As to whether or not one can trust the talking these numbers do rests on your faith in the method by which Project Democracy arrives at its riding predictions.

Here's how it's explained on the website: "The provincial average for any party on election day 2008 resulted in a certain number of votes for that party in any given riding and a defined ratio of provincial average support to votes in the riding. That ratio is applied to an average of polling results for this election to calculate the number of votes the current support level would result in for a given party. This allows you to see how the votes would split today and who would likely win each riding."

I'm not a math expert, but it strikes me that any number of variable go unaccounted for in this system, including a change in the candidates themselves, something the website accounts for via what is described as a "correction" process. Despite the potential for inaccuracy, however, Project Democracy is obviously a hell of a lot better than basing a decision to vote strategically based on national polling results that haven't be tailored to one's specific riding. And even if it's not your desire to block the Conservative party in this election, the website's worth a look if only to get a better idea of how your area tends to vote.