The Globe and Mail has launched an election forecaster on its website that traders on the Sauder School of Business Prediction Markets may find useful. The main idea is quite similar to my own Seats Distribution Election Forecaster. The main innovation is to use a Monte Carlo simulation technique to generate a distribution of outcomes, calibrated to match the most recent polling data. As polling data has a margin of error, this margin is taken into account to generate 1000 random draws based on the underlying statistical distribution. The result is that this method can put upper and lower bounds on predicted outcomes and give a better sense about the robustness of the results. The downside is that the tool only uses "uniform swing" across the country rather than a more nuanced provincial breakdown. The tool also does not take into account swings between non-voters and voters (as my tool does). The current prediction: "There is a 52 per cent chance the NDP will win the most seats, but the Conservatives are close behind". You can try the election forecaster at tgam.ca/election-forecast.