Making a good spinner is actually a fairly significant design challenge. Spin too much and it's boring, spin too much and it's a vomit comet. I have a few observations about controlling the spin from my own tinkerings with them.
What causes the spinning is the track turning. Straightaways, whether level, diving, or climbing, don't cause spinning and any spinning carried into a straightaway from a previous turn soon slows and stops. As to turns, the spin is in the opposite direction as the turn (IOW, if the turn is clockwise relative to the car, the spin will be counterclockwise). Thus, a succession of turns in the same direction cause an increased rate of spinning while a series of alternating turns causes the car to sway back and forth, often not completing 1 revolution before going back the other way.
Also, spinning is fastest on unbanked turns and very slow on highly banked turns, and spin rate on unbanked turns is very dependent on the car's forward speed (more speed = faster spinning). However, speed doesn't make much difference to banked turns. For instance, on a 180^ flat turn of the shortest radius, the car might spin 2 or even 3 times depending on speed, and it doesn't take that much speed to really increase the spin rate. But on a 360^ diving helix banked 45^, where the car might perhaps hit its top speed throughout the track, the car might only spin 270^ very slowly all through the helix.
