Did you know? One blockchain token may display six decimal places while another displays eighteen, yet the second token is not automatically more valuable, more secure or more scarce. The number of decimal places is primarily a way to interpret underlying integer units, not an economic price.
Imagine a fictional school science club with ART credits. Alicia owns 1.25 credits. A computer register cannot store her balance as “one credit plus a quarter” under an integer-only design, so the club chooses to store every credit as 100 smaller units. Alicia’s recorded balance is 125 raw units, while the user interface displays 1.25 ART. The register has not manufactured wealth by dividing each credit into pieces.
ERC-20-style blockchain tokens use a similar convention. A token’s decimals metadata tells compatible interfaces how to scale its integer amounts for human display. Ethereum’s execution environment uses integer arithmetic, and smart contracts transfer raw units. The display convention helps people read fractions.
Quick answer: token decimals indicate how raw integer amounts are presented. For a token with six decimals, 1,250,000 raw units display as 1.25 tokens. For a token with eighteen decimals, 1,250,000,000,000,000,000 raw units display as 1.25 tokens. Neither setting determines the price of one token, its circulating supply, the issuer’s honesty, its redemption rights or whether it can be sold.
This reader-friendly guide explains how blockchain technology works through decimal scaling, wallet balances, fees, token swaps, cross-chain transfers, rounding and common marketing misunderstandings. All numerical examples are hypothetical and not real asset-price quotes.
Choose your reading route
- A credit can have smaller units without becoming more valuable
- What ERC-20 decimals means
- Worked example: six decimals versus eighteen
- Why Solidity and the EVM use integer amounts
- Decimals do not set the market price
- Decimals and total supply must be read together
- What a wallet actually needs to show a balance correctly
- Frequently asked questions
- Primary sources and related reading
A credit can have smaller units without becoming more valuable
A familiar example is the dollar and cent. One dollar can be written as 100 cents. The number 100 looks larger than one, but the underlying amount has not grown. It is expressed using a smaller unit.
Suppose the classroom art club gives Alicia 1.25 ART credits. If the accounting rule says one ART equals 100 raw units, the computer can store the integer 125. To display the result, the interface divides 125 by 100.
Now suppose Kai Kai designs a second register where each ART credit equals 1,000,000 raw units. Alicia’s 1.25 credits are represented by 1,250,000 raw units. The displayed balance remains 1.25 ART after the correct scaling.
Tricia asks which scheme makes the ART credit more expensive. Neither. The club’s reward and redemption policy determines what the credits can do, while the smallest-unit convention determines the precision in which they are counted.
This is the distinction between measurement units and economic value. Measuring a distance in millimetres rather than metres does not change how long the distance is.
A school maths lesson can use this principle to teach place value and dimensional analysis without involving real cryptocurrency. The blockchain merely adds a standard interface and cryptographic record of those integer quantities.
What ERC-20 decimals means
ERC-20 is a common token interface on Ethereum. It defines required functions for balances, transfers and approvals and describes optional human-facing metadata such as token name, symbol and number of decimal places.
The decimals function is an optional metadata extension. A compatible application uses its result to present balances in a readable unit. For example, when decimals equals 2, a raw balance of 505 is displayed as 5.05 tokens.
OpenZeppelin’s official ERC-20 implementation reference makes the distinction explicit: decimals are used for display and do not change the contract’s underlying integer arithmetic, including balance and transfer calculations.
That does not mean every program handles decimals correctly. Wallets, explorers and trading systems must know which token contract they are viewing and how its unit scale applies. An application can misdisplay an amount when it assumes the wrong decimals.
The token’s exact contract identity matters. Two different contracts may share a name and symbol while returning different decimal values. The name alone cannot tell a wallet how to interpret its raw balance.
The foundational source is EIP-20, which defines the token interface and its optional metadata. A student should learn the metadata’s purpose before drawing conclusions about scarcity or price.
Worked example: six decimals versus eighteen
Take a hypothetical raw balance of 1,250,000 units. In a token that uses six decimal places, divide by 1,000,000. The human-readable result is 1.25 tokens.
For a different token using eighteen decimal places, the same raw integer would display as 0.00000000000125 tokens. The raw number was unchanged, but the two tokens interpret that integer using different unit scales.
Now make the human-readable balance the same in both systems: 1.25 tokens. The six-decimal token needs 1,250,000 raw units. The eighteen-decimal token needs 1,250,000,000,000,000,000 raw units.
Neither representation says anything about market price. One credit could have a classroom redemption value of five pencils, while another token’s value might be unrelated. Both could correctly display 1.25 units.
The calculation uses the formula displayed tokens = raw integer ÷ 10^decimals. To prepare a raw token amount from a valid human-readable quantity, the corresponding scaling is by the same factor, subject to precision and rounding constraints.
Writing the powers of ten explicitly makes the units understandable. It also helps readers see why merely comparing raw integer values across contracts is meaningless without their decimal conventions.
Why Solidity and the EVM use integer amounts
Ethereum smart contracts work with defined numerical types, including unsigned and signed integers. In common ERC-20 implementations, token balances and transfer amounts are stored and processed as integers representing the smallest supported token unit.
This avoids treating fractional token amounts as loosely defined floating-point values during consensus-critical arithmetic. A transfer of 1.25 user-facing tokens can be represented by an appropriate whole-number quantity of raw units.
For a six-decimal token, transferring 1.25 units means moving 1,250,000 raw units. The contract processes the integer. A wallet handles the translation that makes the amount readable to a person.
This is why code such as transfer(destination, amount) generally expects the contract’s raw amount rather than a display string saying “1.25.” An interface must prepare the correct units before the network evaluates the instruction.
An incorrect scaling choice can create dramatic mistakes. Sending 125 raw units in a six-decimal token means 0.000125 displayed tokens, not 1.25. Conversely, multiplying twice by a power of ten can produce a far larger requested amount than intended.
Developers need appropriate exact-number arithmetic, input validation and unit tests. Readers should simply remember that the number people see and the integer smart contracts process are related by a declared scale.
Decimals do not set the market price
A token’s price arises from what market participants or redeeming institutions are willing and obligated to exchange for it under particular conditions. The decimal setting only changes the representation of fractions.
Imagine the school issues 100 ART credits redeemable for one classroom notebook each. Whether the register uses two, six or eighteen decimal places, the right to one notebook per whole credit does not change automatically.
If demand for notebooks disappears, ART credits may become less desirable even with eighteen decimal places. If the club promises valuable educational activities, students may value the credits more even with only two decimals. The relevant value comes from benefits, rights, obligations and market demand.
An internet post saying “This token has eighteen decimals, so its price can increase eighteen times” misunderstands the purpose of decimals. A higher decimals number increases the granularity of representable amounts; it is not a multiplying factor applied to every holder’s wealth.
Likewise, a token with zero decimal places can represent whole units without fractional transfer amounts under that interface. That does not make it scarce or expensive by itself.
A beginner can assess value only after identifying the actual asset, its supply, rights and economic environment. Decimal precision belongs on the technical identity card, not in a promise of profits.
Decimals and total supply must be read together
A token contract’s totalSupply method generally returns a raw integer quantity under its model. To display that amount in whole-token units, a reader needs the correct decimal convention.
Suppose a fictional token has raw total supply of 1,000,000,000 and six decimals. Its displayed total is 1,000 whole tokens. A second token with eighteen decimals and the same raw total would show 0.000000001 whole token under the normal scaling formula.
The raw supply numbers look identical but represent different quantities in their respective user-facing units. That is why comparing total supply screenshots without decimals and contract identity can be misleading.
The token may also have locked balances or issuance rules that affect circulating supply. Decimals do not tell readers how many units are liquid, who owns them or whether additional units can be minted.
A project may use a very large raw supply in code for precision. That should not be treated as a prediction of market value, just as the number of cents in a bank account is not a prediction of how much money someone will pay for each cent.
For the related economics distinction, see Why Burning Tokens Does Not Guarantee a Higher Price. Supply changes and display-unit changes are different events.
What a wallet actually needs to show a balance correctly
A wallet viewing an ERC-20 asset needs the correct chain and contract identity, the account’s raw balance and the applicable token metadata or documented presentation rules.
If it queries the wrong network, it may show the balance of a different asset or no balance at all. If it associates the wrong token contract with a familiar name, the display can be misleading even when the integer arithmetic is correct.
If it assumes eighteen decimals for a token using six, a large raw balance can appear incredibly small. The reverse mistake can make a tiny amount appear enormous. Neither presentation error changes the underlying raw balance in the token contract.
A third-party portfolio tracker might also attach an external price estimate to the displayed amount. That adds another separate data source, which may be stale or unreliable. A correctly scaled token quantity does not certify the price feed.
For students, write a complete balance statement: “Account A on network N holds R raw units of contract C, whose defined display precision is D decimal places, corresponding to H human-readable units.”
This level of precision helps support teams distinguish an actual missing asset from a wallet’s formatting mistake. It also helps avoid accidentally comparing unrelated tokens sharing the same logo.
What rounding can do to small amounts
An application may need to show a balance with fewer digits than a token supports. A wallet might round a small amount for display, while the contract continues tracking the full integer amount.
Suppose a fictional six-decimal token has a raw balance of 1,234,567. It represents 1.234567 whole units. A simplified screen might display 1.23. That is a convenient rounded label, not necessarily the exact transferable amount.
A system that performs calculations using the rounded displayed value can produce inaccurate totals. For financial or accounting operations, the precise underlying units matter, especially when many transactions are aggregated.
The issue becomes obvious with very small fees and rewards. A screen displaying 0.00 might conceal a non-zero balance smaller than the chosen display precision. Conversely, a rounded price could make a tiny change appear larger than it really is.
A careful user interface distinguishes approximate presentation from exact amounts used when authorising a transaction. It should not imply that a rounding choice silently created or removed assets.
This is why developers often use integer arithmetic and explicit scaling rules rather than trusting ordinary floating-point display calculations for token accounting. The Smart-Contract Integer Accounting guide studies these implications in more depth.
What happens with zero-decimal tokens?
A token can define zero decimals under the ERC-20 metadata scheme. Then each raw unit corresponds to one displayed token unit. The ordinary interface does not represent fractional amounts between its smallest units.
This may be useful for a particular class of digital ticket, badge or reward whose operator wants units to be handled as whole counts. But zero decimals does not by itself enforce non-transferability, uniqueness of individual tokens or other rights.
A token with zero decimals can still have a large supply. Another with eighteen decimals can have a small displayed supply. The precision choice does not tell you how many distinct holders exist or whether the issuer can mint more.
Some applications assume a fungible token always has eighteen decimals and may handle a zero-decimal asset poorly. Reliable software should query the actual metadata rather than blindly using its favourite default.
For the science club, a membership ticket might be recorded as whole units. A separate payment credit might use six decimals for more flexible amounts. Neither choice decides who may hold or redeem it.
The right question is what the asset represents and which smallest permitted quantity suits the application. Decimals are a design parameter—not a quality rating.
Decimals do not prove authenticity or rights
Token names, ticker symbols and logos are generally human-facing metadata. A malicious or unrelated contract can imitate those fields and sometimes the decimal convention of a genuine asset.
Suppose an official classroom token uses six decimals. An impostor deploys a lookalike with the same name, symbol and six-decimal setting. The metadata match does not create a legal or economic relationship between the contracts.
The actual chain and contract address remain important for identifying the asset. A rights or redemption claim also needs a credible issuer and applicable terms, rather than merely a correctly formatted balance.
A valid token transfer can establish that raw units moved under a specified contract’s rules. It does not prove that the token is authorised by a particular school or backed by a bank deposit.
This makes decimal checking useful but insufficient. A complete asset description includes unit scale, source identity, transfer rules, issuer authority and any external obligation.
Read Why Token Names and Symbols Can Be Copied and Why a Blockchain Address Is Not a Verified Identity for the other sides of the identification problem.
Why changing decimals can confuse existing holders
Some token implementations have decimal metadata fixed or effectively fixed by their deployed code. Others may have upgradeable logic or different compatibility behaviour. A project should not assume changing the displayed scale is harmless merely because decimals are presentation metadata.
Suppose the school records raw balance 125 for Alicia when its convention is two decimals. She sees 1.25 credits. If an authorised software update changes the display convention to three decimals without migrating the relevant underlying accounting, that same raw value now displays as 0.125 credits.
The raw storage has not changed, but the human meaning of every value has. Wallets, exchange routes, prices and external reports may interpret the old and new values differently. A change intended as cosmetic can therefore become a serious accounting or consumer-communication problem.
A responsible upgrade plan preserves the intended relationship between raw amounts and user entitlements, documents any conversion and verifies that all integrated systems use the correct scale.
This is not an instruction to change token decimals after launch. In fact, the example demonstrates why such a change requires careful governance, compatibility analysis and real-world disclosure.
Read Why a Contract Address Can Keep Working After an Upgrade for why an address can remain stable while the meaning of stored values changes.
Cross-chain representations can use different scales
A bridge may create a token representation on another chain. The source and destination assets may use different token contracts and potentially different decimal conventions.
A correct bridge must define how amounts are translated and how rounding or residual small units are handled. The displayed numerical balance on each network should be understood under the relevant asset and contract rules.
Imagine the science club records credits in hundredths on chain Blue and millionths on chain Green. A transfer of 1.25 whole credits should correspond to 125 raw units on Blue and 1,250,000 raw units on Green if the bridge intends one-to-one human-readable units and implements conversion correctly.
The integer values differ dramatically but the intended economic quantity can remain equivalent under the bridge’s agreement. Conversely, an incorrectly implemented conversion could change how much the recipient is credited.
A bridge’s token might also represent a distinct claim rather than the exact native asset. Correct decimal scaling does not prove the bridge’s backing, custody or redemption conditions.
For further reading, see What a Blockchain Bridge Does and Does Not Guarantee. Unit precision is one piece of cross-chain accounting, not the entire settlement guarantee.
Decimals and gas prices are not the same thing
Ethereum users may encounter words such as ETH, wei and gwei alongside an ERC-20 token’s decimals setting. These concepts are related by arithmetic but refer to different assets and purposes.
ETH is Ethereum’s native asset, with wei as its smallest commonly named unit. Gwei is a denomination of ETH often used to quote gas prices. ERC-20 tokens can define their own display scale independently.
A token with six decimals does not automatically pay Ethereum mainnet gas in its own smallest units. Ordinary network transaction fees are charged under Ethereum’s native fee mechanism, though sponsored arrangements can change who bears the cost.
An interface might show “1.25 TOKEN” and “0.0003 ETH gas.” These are different quantities, with different denominators and economic roles. Adding them directly without converting and identifying units would be as wrong as adding metres to kilograms.
This distinction also matters when comparing quoted national-currency values. Token price changes and ETH fee changes can affect a combined cost independently.
Read Why Network Fees Change When It Gets Busy for the difference between gas used, price per gas and native-asset payment.
A classroom worksheet with exact units
Create a table with columns for token, raw units, decimals and displayed tokens. Give ART-A 1,250,000 raw units and six decimals. The displayed balance is 1.25. Give ART-B 125 raw units and two decimals. Its displayed balance is also 1.25.
Now give ART-C 1,250,000 raw units and eighteen decimals. Its displayed balance is 0.00000000000125. Ask students why ART-C does not become more valuable merely because its displayed amount is smaller.
Next add a market-price column. Let the fictional issuer promise that each whole ART-A unit can be redeemed for one pencil while ART-B has no redemption entitlement. Both have the same displayed balance, but different rights. The source of value is not the decimal convention.
Introduce a wallet error. Suppose the app shows ART-A using eighteen decimals instead of six. Students should identify the incorrect presentation without assuming the contract’s raw balance changed.
Finally ask how many smallest units a transfer of 0.000001 ART-A requires. With six decimals, it is exactly one raw unit. A smaller non-zero fractional amount cannot be represented under that particular scheme.
This simple exercise combines powers of ten, exact quantities and critical reading. It requires no real token, wallet or price quote.
Seven checks when token amounts look strange
Which contract and chain? Make sure the amount belongs to the intended token rather than a lookalike or another network.
What raw amount is recorded? Distinguish the underlying integer from a rounded portfolio label.
What decimals value applies? Use the actual metadata or documented asset convention rather than assuming every token has eighteen decimal places.
Did the interface round? A display of zero can sometimes mean “less than the chosen display precision,” not truly zero.
Is the price another source? The market quote or redemption right is independent of the decimal scale. A wallet’s national-currency estimate can be wrong even when raw token arithmetic is correct.
Did any implementation change? Upgrades or new token representations can affect interpretation. Confirm which scale the current system actually uses.
Are any cross-chain units being converted? Identify how source and destination representations reconcile, including rounding and residual quantities.
A careful conclusion could say: “This wallet displayed raw amount R under an incorrect token-decimal setting; the accepted on-chain balance remained R.” That is stronger than claiming a token suddenly gained or lost value because the screen changed.
Frequently asked questions
What does 18 decimals mean in a crypto token?
It means one human-readable token unit is represented by 10^18 raw units under the ordinary display convention. It does not say how much one token costs.
Are tokens with more decimals more valuable?
No. Decimal precision and market value are different properties. A price comes from market conditions or redemption arrangements, not directly from the decimals field.
Why do some tokens have six decimals?
A token issuer or implementation can select a precision appropriate for its design. Six decimals represent one whole token as one million raw units.
What is the formula for displaying a token balance?
In the ordinary ERC-20 model, displayed amount equals raw integer amount divided by ten raised to the number of decimal places.
Does changing decimals create more tokens?
Changing display interpretation does not automatically mint raw units. It can nevertheless alter what users think their balances represent, making it a sensitive design and communication issue.
Are ERC-20 balances stored as fractions?
Common ERC-20 implementations store integer raw units. Wallets convert those integers into human-readable fractional token amounts.
Can a wallet show the wrong balance because of decimals?
Yes. Incorrect metadata or wrong token identity can produce a misleading displayed quantity while the underlying contract’s raw amount is unchanged.
Is zero decimals a rare or invalid token setting?
A token can use zero decimals. It means whole raw units match whole displayed units under its convention, but it does not establish any particular scarcity or transfer rule.
Is gwei an ERC-20 decimal?
No. Gwei is an ETH denomination often used for gas-price quotes. A separate token’s display decimals do not define Ethereum gas units.
Can two tokens have identical symbols and different decimals?
Yes. Names and symbols are not globally unique, and distinct contracts may implement different decimal metadata.
Can a bridge convert between different decimals?
A bridge can be designed to reconcile differing raw-unit scales, subject to its exact rules and rounding behaviour. Proper conversion is not automatic proof of backing or settlement.
What should students remember?
Decimals tell you how to read the units, not what the units are worth. The actual contract, asset rights and market determine the rest.
Primary sources and related reading
The ERC-20 token standard describes balance, transfer and optional token metadata methods. The OpenZeppelin ERC-20 reference explains that decimals affect user presentation, not the underlying contract arithmetic, and that its default implementation uses eighteen decimal places. The OpenZeppelin ERC-20 guide gives further accessible examples using integer scaling.
For deeper eduKateSG reading, continue to Smart-Contract Integer Accounting: Scales, Rounding and Conserved Remainders, Why Token Names and Symbols Can Be Copied, Why a Wallet Balance Cannot Always Be Spent and What a Blockchain Bridge Does and Does Not Guarantee.
The useful habit is to write raw amount, token decimals, displayed quantity and actual economic rights as four distinct facts. Once you do that, large numbers and tiny fractions stop being mysterious.