Dice & Gear
How to Tell If Dice Are Balanced (Salt Water Test)
Test a die's balance with the salt water float test: dissolve salt into warm water until the die floats, then flick it to spin a dozen times. A balanced die surfaces with different faces up; a biased one shows the same face again and again, meaning the opposite side is heavy. Most dice pass. Your d20 that "always rolls low" is almost certainly innocent, and we'll get to why your memory says otherwise.
The salt water float test, properly
You need a glass, warm water, and a lot more salt than intuition suggests.
- Fill a glass with warm water and stir in salt until no more dissolves. For most acrylic and resin dice this is a lot, a third of a cup or more; epsom salt dissolves further and floats denser dice more easily.
- Drop the die in. If it sinks, add salt and stir until it hovers or floats. If it simply won't float (dense resin does this), see the alternatives below.
- Flick the floating die to spin it. Note the face that ends up. Repeat 10-15 times.
- Read the verdict: different faces every time means balanced enough to play. The same face surfacing repeatedly means the die's center of mass is off, pulled toward the opposite side, and that face will genuinely come up more on the table.
Two caveats the internet versions skip. Surface tension can stick a die to one orientation regardless of balance, so poke it fully under between spins. And the test only detects internal weight problems; it says nothing about warped faces or out-of-true geometry, which brings us to what actually causes bad dice.
What makes a die biased
Air bubbles are the main villain: a void trapped near one face during molding makes that side light, and hand-poured or cheap resin carries this risk far more than mass-market acrylic. Big inclusions are the second, since a flower or die-cut foil sitting off-center shifts mass permanently; this is a known trade-off of the prettiest sharp edge sets. After that come out-of-round tumbling (faces of slightly different sizes), and, for softer materials, wear: years of rolling can round the corners a favorite die unevenly.
Material matters mostly through manufacturing. Machined metal dice are the most consistently true; name-brand acrylic is boringly reliable; budget resin and artisan pours are where the float test earns its keep. The full material rundown lives in our dice materials guide.
If a die won't float
Dense resin, stone, and metal laugh at salt water. For those: the pivot test (balance the die on a fingertip or pencil point at various orientations; a heavy side falls the same way repeatedly) is crude but revealing for metal. The honest tool is a roll log: 100 rolls tallied.
It sounds tedious and takes ten minutes with a friend, and for a d20 you're looking for any face showing up dramatically more than the expected 5 times per hundred. Statistics needs bigger samples for proof, but a die that rolls 1 fifteen times in a hundred doesn't need a bigger sample. It needs a jar to live in as a warning to the others.
The part where your d20 was innocent
Here's the uncomfortable truth behind most "cursed dice": humans remember failed saves and forget routine hits. A balanced d20 rolls a 1 five percent of the time, which is once every twenty rolls, which is a couple of times per session, and every one of them is memorable. Confirmation bias does the rest, and an entire folklore of dice jail is built on it. Casinos, who care about this more than anyone alive, solve it with machined precision dice and still assume randomness will produce brutal streaks, because it does.
So run the float test for peace of mind, retire the rare genuine offender, and then accept the liberating truth: the dice were never the reason the barbarian missed three times. The dice are just dice. That's the whole magic of them.
Frequently asked questions
How does the salt water test for dice work?
Saturate warm water with salt until the die floats, then spin it repeatedly and watch which face surfaces. Varied faces mean balanced; the same face over and over means the die is heavier on the opposite side and genuinely biased.
How much salt do I need to float a die?
More than seems reasonable: keep dissolving until the water is saturated, often a third of a cup per glass. Epsom salt reaches higher density than table salt and floats stubborn dice. True dense resin, stone, and metal may never float; use a roll tally instead.
Are cheap dice unbalanced?
Mass-produced acrylic from reputable brands tests fine almost every time; the risk zone is bargain-bin resin and handmade pours where bubbles and off-center inclusions hide. Test anything with visible inclusions if fairness matters to your table.
Do casinos really use perfectly balanced dice?
Effectively, yes: casino dice are machined cellulose acetate with sharp edges, drilled and back-filled pips, and tolerances of a fraction of a millimeter, replaced after hours of use. It's the standard hobby dice are loosely imitating, at one percent of the rigor and price.
Should I retire a die that fails the float test?
For any game where fairness matters, yes, or demote it to a prop. A die that consistently surfaces one face is measurably loaded, and no amount of dice jail rehabilitates physics.