After tensioning a rope with an overhand knot in it, it is often very hard if not impossible to untie it; a figure-8 knot, on the other hand, still releases easily.
Why is that so? Most "knot and physics" literature seems more concerned with the topology of knots (see e.g. here), but I suppose friction and elasticity of the rope play a role in knowing whether a knot jams or not.
I found this reference that explains why the overhand knot breaks earlier than the figure-8 — the former has a higher maximum curvature than the latter. Why doesn't this larger curvature make the overhand knot easier to untie? The knot would have more stored elastic energy that would aid overcoming friction when picking the knot apart.