I'm using the data at
https://covid-19.ontario.ca/data . It shows the inversion you've mentioned.
- Cases: vaccinated = 100/100k, unvaccinated = 80/100k
Vaccinations reduce the severity of the disease. If you get exposed, vaccination will prompt your body to deal with the infection more quickly when there's less of the virus. This can, but doesn't necessarily, clear the infection before it's detected. You can see the protective effect in Ontario by looking at different severities.
- Population 12+: vaccinated = 88%, unvaccinated = 9%
- Hospital non-ICU: vaccinated = 65%, unvaccinated = 25%
- Hospital ICU: vaccinated = 40%, unvaccinated = 50%
- Deaths aren't reported by vaccination status on this page