By Gideon Vance, longtime Dungeon Master and gemstone dice collector. Last reviewed July 2026.
The salt-water float test is the cheapest way to check whether a d20 secretly favors a side, and you can run it with nothing but a glass, warm water, and ordinary table salt. The short version: dissolve salt into the water until no more will dissolve, float your d20 in the brine, and spin it a dozen or more times. If the numbers that surface look random, the die passes. If the same face keeps bobbing to the top, the die's weight is off-center and it will favor certain results on the table. The one big catch is that the test only works on dice light enough to float — most acrylic and resin dice qualify, while stone, glass, and metal dice sink no matter how much salt you add. This guide walks through the exact procedure, how to read the result, and what the test can and cannot tell you.
Why salt water makes a d20 float
Whether anything floats comes down to density: by Archimedes' principle, an object floats when it is less dense than the liquid it sits in. Fresh water has a density of about 1.0 g/cm³, and typical dice plastics sit just above that — acrylic, for example, is about 1.18 g/cm³ — which is why a d20 dropped in plain tap water sinks. Dissolving salt raises the water's density. Keep adding table salt until the water is saturated — at room temperature that is roughly 26 percent salt by mass, the point where extra salt just settles on the bottom — and the resulting brine reaches a density of about 1.2 g/cm³. That small margin is enough: the brine is now just denser than the plastic, so the die pops up and floats with a sliver of one face above the surface. The physics of buoyancy then does the work for you. A perfectly balanced die has its center of mass at its geometric center, so no face is "heavier" than another and every spin settles at random. A die with an internal air bubble or a dense spot has an off-center mass, and the heavy side rolls down while the light side surfaces — the same face, float after float.
The salt-water float test, step by step
- Pick a clear container. A drinking glass or jar works; clear sides let you watch the die spin and settle instead of peering from above.
- Start with warm water. Warm (not boiling) water dissolves salt much faster. Fill the container deep enough that the die can rotate freely.
- Stir in table salt a spoonful at a time. Keep stirring until each new spoonful stops dissolving and grains start collecting on the bottom. That leftover salt is the sign the brine is saturated — you cannot usefully add more.
- Let the water go still. Give the glass a minute so the swirl dies down and the temperature evens out; a moving current will fake random results.
- Lower the d20 in. If it floats, you are in business. If it sinks, stir to confirm the brine really is saturated and try once more — if it still sinks, your die is too dense for this test (see the table below for what to do instead).
- Spin the die underwater. Push the die down with a finger or spoon and give it a gentle twirl so it tumbles before it floats back up. Note which number ends up facing the ceiling.
- Repeat a dozen or more times. One float means nothing. Re-spin the die again and again, writing down the top face each time. More floats give the pattern more chances to show itself.
- Read the pattern. Numbers scattered all over the die: balanced. The same face — or a tight cluster of neighboring faces — surfacing most of the time: the die carries off-center weight on the opposite side.
Reading the result fairly
Be careful not to convict an innocent die. Even a perfectly balanced d20 will repeat a number occasionally, exactly the way a fair coin sometimes lands heads three times running. What you are looking for is not a repeat but a strong, boring pattern: the same face or its immediate neighbors surfacing float after float while most of the other seventeen faces never appear. Two practical tips improve the read. First, if tiny air bubbles cling to the die's surface, they can steer the float — lift the die out, wipe it, and re-dunk it; a single drop of dish soap in the brine also helps the water wet the surface evenly. Second, spin the die properly under the surface each time rather than letting it bob up from the same grip, or you are testing your fingers, not the die.
What the float test can and cannot detect
The float test is a balance test, not a complete fairness audit. It is superb at catching the single most common defect in mass-produced plastic dice — an air bubble trapped inside the resin during casting — but it is blind to problems of shape. The table below is the honest scorecard.
What about dice that sink?
Stone, glass, and metal dice cannot take this test: at densities from roughly 2.5 g/cm³ for glass up past 6 g/cm³ for zinc alloy, no kitchen brine will float them, and that is fine — solid mineral and metal dice cannot trap the casting bubbles that the float test exists to find. For those materials the practical checks are different: look at the machining instead. Our guide to telling real gemstone dice from fakes covers the giveaways for stone, and if you are choosing a workhorse die for important rolls, our breakdown of what makes a d20 worth trusting goes deeper on materials and manufacturing quality. If you are not sure what your suspiciously light "resin" die actually is, start with the difference between acrylic and resin dice — both float in saturated brine, so both are fair game for the test.
The bottom line: for a few cents of salt and ten minutes, the float test answers one sharp question — is this die's weight centered? — with a clear yes or no. It will not certify a die tournament-fair, but it will absolutely catch the bubble-cursed d20 that keeps rolling low, and knowing is better than wondering.
| Situation | Can the float test catch it? | Why |
|---|---|---|
| Air bubble trapped inside a resin or acrylic die | Yes | The bubble side is lighter and keeps floating up, so one face repeats |
| Dense spot or off-center core inside a plastic die | Yes | The heavy side keeps rotating down, so the opposite face repeats |
| Slightly misshapen, shaved, or uneven faces | No | That is a geometry problem; the die can float randomly and still roll unfairly |
| Worn or rounded edges | No | Edge wear changes how a die tumbles on the table, not how it floats |
| Stone, glass, or metal dice | No | At about 2.5 g/cm³ and up they sink even in fully saturated brine |
| Very small imbalances | Not reliably | A tiny offset may not overcome water drag and surface effects |
Frequently Asked Questions
How much salt do I need for the dice float test?
There is no recipe to measure: stir table salt into warm water a spoonful at a time until it stops dissolving and grains settle on the bottom. That is saturation — roughly 26 percent salt by mass — and it is the densest brine plain salt water can reach, about 1.2 g/cm³.
Does the float test work on metal or gemstone dice?
No. Glass sits around 2.5 g/cm³, quartz-family gemstones around 2.65, and zinc-alloy metal well above 6 — all far denser than saturated brine at about 1.2, so they sink. The test only works on plastics such as acrylic and resin.
Is my d20 unbalanced if the same number floats up twice?
Not by itself. Balanced dice repeat numbers occasionally, just as fair coins land the same side a few times in a row. Suspect the die only when one face or a tight cluster of neighbors dominates across a dozen or more spins.
Can the float test prove my die rolls fair?
No — it only checks that the die's weight is centered. A die can pass the float test and still roll unfairly if its faces are misshapen or its edges unevenly worn. Passing rules out the most common defect, an internal air bubble, not every defect.
Why does my die sink even in fully saturated salt water?
Its density is above the brine's ceiling of about 1.2 g/cm³. Plastics loaded with mineral fillers, metal-powder blends, glass, and stone all exceed that. Sinking says nothing about balance — it just means this test does not apply to that die.
Does soaking in salt water damage dice?
A brief float in salt water is harmless to acrylic and resin dice. Rinse them in fresh water and dry them afterwards so no salt film dulls the finish, and keep painted or inked numbers out of long soaks.