The retarded vertical motion of a short cylindrical magnet falling inside a non-magnetic conductive pipe is an appealing phenomenon that students, teachers and the public at large enjoy as in a magician performance. This content rich phenomenon has been recently addressed a number of times in physics teaching journals. The retarded fall of the magnet is seen to consist of an initial transient accelerated regime followed by motion at a terminal speed. After the phenomenon is demonstrated many interesting questions and useful suggestions quickly arise during the ensuing discussion. What would happen if the magnet falls off the pipe axis? Why don’t you try with two magnets at the time? In this work we present theoretical models and experimental results for this retarded fall, considering first the case of conductive pipes of different materials and different wall thickness, and then the case of two magnets falling together. We here predict with good accuracy the experimental results later obtained by us in the laboratory. The effect of varying the pipe wall thickness on the magnetic drag is studied for pipes of two different materials, copper and aluminium. The case of magnetic braking in aluminium pipes is interesting since copper is diamagnetic while aluminium is paramagnetic We have performed experiments with the two magnets falling together either with parallel or opposite magnetic moments, with variable separation in between, and also experiments with a single magnet falling at different distances from the pipe axis. We develop successful analytical models for these cases of magnetic braking. The experimental setups are inexpensive and can be readily assembled in a teaching laboratory.