… is the excellent title of an article in the December 2014 edition of the Harvard Business Review. Two examples that apply to support metrics:
Truncating the Y (vertical) axis. By only showing a slice of the data universe (usually by truncating the low values), it’s easy to convince eyes and brains that declines or increases are way more dramatic than they really are. Start axes at 0!
Using cumulative growth rates. They fool the mind into thinking that growth is continuous and smooth — and hide stagnation. Instead, show period-over-period changes. Compare the graphs below:
Time to re-read The Visual Display of Quantitative Information. It’s a classic!