When importing an IMDb ratings export (ratings.csv), Simkl currently appears to assign the import date as the watch date for every imported movie and TV show.

This causes several issues:

Monthly and yearly statistics become meaningless.
All imported titles appear as if they were watched on a single day.
"Most active day", "Longest binge", "Average per day", activity heatmaps and similar statistics become completely inaccurate.

However, the IMDb ratings export already contains a "Date Rated" field.

I understand that IMDb does not store the actual watch date, but using Date Rated as the watch date would be a much better approximation than assigning everything to the import date.

Suggested improvements:

Use Date Rated as the watch date during import.
Alternatively, provide an import option:
Use import date (current behavior)
Use IMDb "Date Rated" as watch date (recommended)

This would preserve historical activity and produce much more accurate statistics for users migrating from IMDb.

Even if Date Rated is not the actual viewing date, it is significantly more useful than assigning every title to the import date, which completely breaks all time-based statistics.