The chart itself is fixed to one time, one direction and one city — 2000 GMT on 9 November, looking south from London. The counting rules are what turn that plotted square into the experiment: naked eye only, corners excluded, no phone, and a wait of at least 20 minutes in the dark.
The point of the exercise is a personal reading of sky darkness, made by eye, without instruments. The same patch of sky can look busy or nearly empty depending on the conditions overhead.
Leave out the four corner stars: Alpheratz, Scheat, Markab and Algenib. They frame the square and are excluded from the tally, so only the stars inside it are counted.
A phone will undo that wait. The instructions say not to check the screen to pass the time, because the bright display will ruin dark adaptation.
The diagram shows the view looking south at that hour. A square marked on the chart is the field for the count.
When the eyes are ready, look carefully inside the square and count every star you can see with the naked eye.
The number that remains is the measure. Four stars indicate an average sky.
A chart of the southern sky, drawn for London at 2000 GMT on 9 November, sets out a simple test of how dark the night is where you stand.
To try it, find a place away from direct lighting. Then wait at least 20 minutes so your eyes can adjust to the darkness before you begin.
Seven stars suggest good conditions. A count of about 35 means you are observing under an exceptionally dark sky.