Dice Shaker Motor Randomness: Can Be Biased?
A motor-driven dice shaker is only as random as its weakest mechanical tolerance, and I have pulled apart enough of these rigs to tell you that most of them leak bias like a cracked seal. The core problem is never the dice themselves. It is the motor torque curve, the cup geometry, and the dwell time between shakes. Get any one of those wrong and your "random" throw becomes a predictable pulse you could set a watch to.
Let me be blunt: a dice shaker motor running on a cheap brushed DC drive delivers a rotational velocity that is anything but uniform. Voltage sag, brush wear, and bearing friction all conspire to skew the final resting orientation of the dice. Over a million spins, that skew shows up as a measurable deviation from the theoretical 1-in-6 distribution. I have seen units where a single face landed 11.3 percent more often than expected because the motor's decaying wobble dumped energy into the cup at a consistent phase angle. That is not randomness. That is a predictable mechanical signature.
The Torque Curve Is Where Bias Is Born
Every motor has a startup transient. The first 80 to 120 milliseconds of a shake cycle are dominated by the rotor accelerating from rest, and during that window the cup experiences a whip-like impulse that is far from isotropic. If your control board fires the motor with a fixed pulse-width and a fixed ramp, the dice get the same directional kick every single time. The result is a biased dice shaker that looks legit on the outside but behaves like a card mechanic on the inside. The fix is not a software patch. It is a randomized drive profile, with jitter injected into both the ramp duration and the peak current, so no two shakes share the same energy envelope.
I have audited rigs where the firmware used a simple lookup table of shake intensities. Ten values, cycling in order. Ten. That is a pattern a player can feel in their gut after a few hundred rolls, and a sharp auditor can prove it with a chi-square test in under an hour. The motor was fine. The randomness source was the problem. 🎲
Cup Geometry and the Dwell Trap
The cup itself is a resonator. Its interior ridges, its lid clearance, and the hardness of the baffle all determine how energy dissipates after the motor cuts out. A tight cup with rigid baffles produces a short, sharp settle that favors dice sliding to the lowest-energy corner. A loose cup with soft baffles produces a long, chaotic tumble that looks random but can actually over-rotate the dice into a narrow band of outcomes. Neither is neutral. The dwell time between the motor stopping and the dice fully settling is the silent variable that most testers ignore, and it is exactly where a randomness bias hides. If the settle window is too short, you are reading dice that are still in motion. Too long, and you invite micro-vibrations from the table to nudge a nearly-settled die.
What a Real Audit Looks Like
You do not trust the spec sheet. You instrument the thing. Mount an accelerometer on the cup, log the motor current waveform, and run ten thousand automated shakes while a high-speed camera records every landing. Then you run the numbers. Look at the autocorrelation of the outcome sequence, not just the aggregate frequency. A fair shaker shows no lag correlation above noise. A biased one shows a spike at the shake-period frequency, and that spike is your smoking gun. ⚙️
My hard rule after years of this work: any motorized shaker that does not randomize its drive waveform, its dwell timing, and its internal cup pressure profile is suspect until proven clean. The motor is not a random number generator. It is a mechanical system with memory, and memory is the enemy of fairness.