How do you document outliers without distorting your weekly statistics?

When tracking daily, I always notice that individual outliers - such as illness, travel or extremely little sleep - severely distort my weekly values. I'm interested in how you mark such days in the Daily Data Tracker: leave them in completely, mark them separately or exclude them from certain evaluations?

My thought on this: Outliers should not be deleted, but rather visibly contextualized. Otherwise the statistics will appear cleaner than everyday life really was. Maybe a simple rule like:

• normal day
• special day with marker
• non-comparable day

In this way trends remain recognizable without individual extreme days covering everything up.