5. Simplify
The simpler your plot the better.
Use the simplest and most basic plot possible to tell your story
A lot of visual clutter and plot complexity comes from not having a solid idea of the message of the plot before you start plotting, and attempting to fit too much in.
Remove clutter whenever possible
If your plot is confusing, try to clarify it by removing something, not adding extra labels and annotations. Beware of chartjunk, like filled backgrounds and uneccessary grids.
Sometimes this can be as simple as removing grid lines.
Sometimes, it means splitting your plot into multiple panels and making each panel more simple. Don't pick a complicated and unusual statistical plot just because it looks interesting: does it actually serve the dataset and the message or result you want to convey? Might it confuse or mislead the reader? Similar questions should be asked before building interactive or 3D visualisations. Does this actually help the reader to understand the message?