Monday, May 5, 2014

Ontario Mathematical Prediction

This is a "Teddy Projection" - or a Blunt Objects Predictions; as Kyle handles the official Blunt Objects Projections.

This was built using my new and basic ElectoMatic for Ontario. You can download it for yourself here.
The map it produces is this:

There are a few things to note from past comments, maps, and predictions.

1 - This takes into account By-Elections; but does not presume they represent the electorate as a whole. London West, for example, had a very low turnout compared to the General Election (and to some of the other by-elections, frankly) and I interpret it more as a vote against Scandal over a vote against the Liberals. I expect some of these voters will return to the Liberals to stop the Tories.

2 - This takes into account the General election and history; but does not use this as a blunt object. Etobicoke-Lakeshore, for example, is a very liberal riding provincially. While the MPP is certainly a popular Conservative, despite two high-profile candidates, the turnout, again, was rather low.

3 - This takes into account local forces; but does not totally over-ride province-wide forces. Kitchener is an excellent example. The turnout in the by-election there was very high, in addition, the PC candidate for Centre not only made disparaging remarks against Women, but is facing female candidates in both the Liberals and NDP.

4 - This takes into account the last election; but not the one before that. That may sound odd but consider a popular MPP for the Liberals ran for re-election in Riding #70 on the map. A popular MPP for the Liberals did not run for re-election in riding #4. What this means is I need to make an adjustment (downwards) in the Liberal vote in the former but not the latter.

These 4 explain why this map looks a bit different than my previous maps, which had 0, 1, 2, or 3 of these taken into account.

A few things do pop out at me.

The Greens really do have a strong base of support in Orangeville (Dufferin/Caledon) I've thus removed some of what I thought was Guelph support.

The Greens really do not have a strong showing in SouthWest Ontario. That was one poll that quickly vanished during poll averaging. Despite that, they are trending towards 30% in Dufferin-Caledon

Etobicoke-Lakeshore will likely not turn Blue unless the PC Party wins a government.

Both the NDP and Tories have some chances north of the 401 within Toronto.

The PC vote is starting to look more and more like the CPC vote; while the OLP vote still does not look like the LPC vote.


  1. Using the ElectoMatic is simple. The only cells you are changing are the 4 cells at the top representing the share of vote taken by the PC Party, NDP, Liberals, and Greens. The rest of the cells, including the "others" are calculated for you automatically.

    Note that the ridings are coloured in the text of the colour the are in on the map above. This is to help you spot which ridings have changed, as you'll end up with a 1 in the PC column, for example, of a red-coloured Riding; meaning the Tories have now taken that riding on this new projection.

  2. Also note I find the Greens sitting at near 30% in Dufferin Caledon to be as unbelievable as you do. Change the Green number from 6.62 to 10.00 and see what happens.

    I've tried to come up with a good math based reason why the projection is wrong and I can not do so.
    Like the Tories winning 10 seats in Quebec in 2006
    Like the NDP sweeping Northern Ontario in 2008
    Like the NDP sweeping Quebec in 2011
    This is something the math is trying to tell me and I just can not ignore it.

    This really is a strong Green riding.

  3. The file just downloaded a bunch of god awful programs on my computer.

    1. I've uploaded it to a different website.

  4. The file is a simple XLS file.