Roulette Wheel Physics
A moving ball, a rotating wheel and collisions that make the final pocket difficult to predict.

The ball slows on the track, moves inward and collides with deflectorsequipmentDeflectorAn obstacle on the apron that redirects the falling ball. Names include diamond, canoe and pin.Read full definition → and pocket dividers. Those stages follow mechanics, but useful prediction requires measurements and a model that works under the actual conditions.
Three phases
First the ballequipmentBallThe small ball that settles in a numbered pocket to determine a physical roulette result.Read full definition → travels around the track. It then leaves the rim and moves toward the rotorequipmentRotorThe rotating part of a physical roulette wheel, carrying the numbered pockets.Read full definition →. Finally, collisions and rebounds determine where it settles. Splitting a spin into these stages helps explain it, but does not make the last stage physically random or wholly unpredictable.
The ball travels around the outer track.
Energy is lost through resistance.
The ball moves inward from the rim.
Deflectors and dividers alter its motion.
The ball comes to rest in a pocket.
The track
A circular path requires inward acceleration v²/r. The contact forces depend on the track geometry and the ball’s motion. In an ideal level conical-track model, the departure threshold is v² = rg tan(α), with slope angle α measured from horizontal. Real geometry, tilt and losses can complicate that approximation.

The rotor
If a 37-pocketequipmentPocketA compartment on the rotor in which the ball can settle. The number of pockets depends on the wheel.Read full definition → rotor turns at 0.4 revolutions per second, 14.8 pockets pass a fixed point each second. A 0.1-second timing error then corresponds to 1.48 pocket widths of rotor motion. This is an illustrative conversion, not a measured accuracy claim.
Deflectors and frets
Collisions can amplify small differences in position, velocity or contact angle. The resulting scatter should be measured for the equipment and conditions of interest. Neither a universal 10-to-15-pocket spread nor a symmetric distribution can be assumed without data.

What the prediction evidence shows
The size of any predictive advantage is not fixed. Small and Tse’s 2012 experiment reports 13 successful target-half predictions in 22 trials, corresponding to an 18% sample return under their stated calculation. That small experiment is not a universal casino edge or a guarantee that a player can reproduce it.
Where to go next
The related pages separate ball motion, rotor speed and simplified deceleration models. Timing-based prediction is also different from detecting a persistent imbalance in recorded pocket counts.
Questions
No. Uniform pocket probabilities are a statistical claim; complicated motion alone does not establish them.
Do not assume that count or a particular geometry without checking the model of wheel.
Yes. A fitted model applies to the conditions under which it was measured; changes need investigation rather than a universal correction factor.
Comic relief

A complicated bounce does not, by itself, prove fairness.
Sources
- Small and Tse (2012), Predicting the outcome of roulette
- Illustrative kinematics