The Record So Far
The 338Canada / Qc125 model has thus far covered 19 general elections in Canada.
In total, 2,382 electoral districts were projected.
So far, the model has correctly identified the winner
in 2,120 districts, a success rate of 89.0%.
Among the 262 remaining districts, 161 of the winners
(6.8%) obtained a share of the vote that was within
the projection's margin of error (moe). Only 101 districts
(4.2%) were complete misses.
On this site, electoral districts are labelled according to the odds of winning by the projected favourite party.
If a party's odds of winning are at or above 99.9% (formerly 99.5%), the district is labelled as safe. Between 90% and 99.9%, the district is likely.
Between 70% ad 90%, we use the label leaning. Finally, if no party has odds above 70%, the district will be labelled as a toss up.
Here below is the complete breakdown of correct and incorrect winners from safe, likely, leaning and toss up districts:
Districts
Correct
winner
Incorrect
winner
Correct %
Safe
1096
1080
16
99%
Likely
576
536
40
93%
Leaning
313
258
55
82%
Toss up
397
246
151
62%
Total
2,382
2,120
262
89%
Here is the record of the 338Canada model in every election covered since its creation in 2017. Click on the links to see details.
So far, the model has correctly identified the winner in 2,120 districts, a success rate of 89.0%.
Among the 262 remaining districts, 161 of the winners (6.8%) obtained a share of the vote that was within the projection's margin of error (moe). Only 101 districts (4.2%) were complete misses.
Here below is the complete breakdown of correct and incorrect winners from safe, likely, leaning and toss up districts:
| Districts | Correct winner |
Incorrect winner |
Correct % | |
| Safe | 1096 | 1080 | 16 | 99% |
| Likely | 576 | 536 | 40 | 93% |
| Leaning | 313 | 258 | 55 | 82% |
| Toss up | 397 | 246 | 151 | 62% |
| Total | 2,382 | 2,120 | 262 | 89% |