Almost every die in the world is slightly biased, and for almost every purpose that is completely fine. Knowing where the bias comes from is what lets you decide whether it matters for what you are making.
Source one: hollow pips
A standard die has its pips drilled out. The six face loses six pips of material; the one face loses one. The die is therefore very slightly lighter on the six side, which means the six side comes up slightly more often.
The effect is small but real and it is entirely systematic. Precision dice fill every pip flush with material of matched density, which removes the source rather than compensating for it.
Source two: rounded corners
Moulded dice are tumbled to remove flash, which rounds the corners and edges. Rounded dice roll longer and settle more randomly, which sounds good but actually makes the die more sensitive to tiny asymmetries in shape.
Casino dice have razor edges: square, sharp and machined. They tumble less predictably in the air and settle decisively, which combined with flush pips is what makes them fair.
Source three: manufacturing tolerance
Injection moulding holds a tolerance of a few hundredths of an inch, which is invisible and still enough to bias a die slightly. Machined precision dice hold a few thousandths — an order of magnitude tighter.
This is the part that costs money, and it is why precision dice are priced where they are.
When it genuinely matters
Casinos, card rooms and sanctioned tournaments, where the fairness of the die is part of the product and someone may reasonably ask. Also statistical or educational use, where the distribution is the point.
For a board game, a giveaway or a retail set, the bias is far smaller than the variance any player will ever notice, and paying for precision is paying for a claim rather than an experience.
What you can honestly say
If you have precision dice, say the tolerance and say the pips are flush filled. Those are specific, checkable claims.
If you have standard dice, do not call them casino grade, balanced or fair. Those words mean something to the people who care, and using them loosely is how a supplier loses that audience permanently.
Testing a die yourself
The saltwater float test — suspending a die in heavily salted water and spinning it to see whether the same face keeps surfacing — catches gross bias and nothing subtle. It is a useful sanity check and not a measurement.
Real testing is thousands of rolls or a density measurement. If you need that, you need precision dice, and we will supply them with the tolerance stated.