Actually, it would have a rotational symmetry of order 1 because there is only one orientation in which its outline matches…

'Luthon64

Well, that is my whole problem. In the textbook's "teacher's guide," which contains answers to the exercises, such a shape is claimed to have a symmetry order of zero. Now I still can't work out whether that is correct or not, because your smiley face indicates that your reply above may or may not be entirely serious.

If something like a propeller has an order of 2, i.e. we count the original position (or the 360 degree rotation, which comes to the same thing), then logic would suggest that in a shape lacking rotational symmetry, we should still count the original position as 1. I can't work out whether this is what is in fact done because of conflicting reports from various authorities. :-)