Layer 2 networks: what they are and when to use one
Arbitrum, Optimism, Base and the rest settle to Ethereum but cost far less. How that works, and the one mistake that loses funds.
Ethereum is secure and expensive. Layer 2 networks exist to keep the security while removing most of the cost, by doing the work elsewhere and posting the results back to Ethereum.
How the saving happens
A layer 2 batches thousands of transactions, executes them on its own network, and submits a compressed summary to Ethereum along with a proof that it was done correctly.
Ethereum verifies the summary rather than each transaction, so the expensive part is paid once and shared. That is why a transfer on a layer 2 can cost cents where the same transfer on Ethereum costs dollars.
What you give up
Withdrawing back to Ethereum can be slow on some designs — up to a week on optimistic rollups, because that is the window during which a faulty batch can be challenged.
You also depend on the layer 2's own operators for liveness. The funds remain recoverable through Ethereum, but the convenience can stop.
The mistake that costs money
Layer 2 addresses look exactly like Ethereum addresses — the same 0x and the same 42 characters. Many people even control the same address on both.
So sending USDT on Arbitrum to an address that only credits Ethereum mainnet fails every intuition: the address is right, the format is right, and the funds are on a chain the recipient is not watching.
This is the single most common layer 2 loss, and the only defence is reading the network field rather than trusting the address. BYCRYPTO supports Ethereum mainnet, not layer 2s — so for a deposit here, mainnet is the network to select.