How to compare L2s without the marketing: five metrics instead of TPS

TPS from pitch decks is the sector's worst metric: lab numbers say nothing about users or money. L2s should be compared differently.

Five metrics that do not lie

• TVL and its composition. Not the total but the structure: how much sits in native protocols versus bridged stables, and whether TVL is inflated by a points program that will end together with the liquidity.
• Network fees (revenue). How much users actually pay for blockspace. A network with revenue is alive; a network where activity is subsidized by grants is a desert with props.
• Active addresses versus transactions. A million transactions from ten thousand addresses is bots. Watch the ratio.
• Native protocols. Are there apps born on this network and living only here - or is the entire DeFi a set of multi-deploy forks.
• Stablecoins. The volume of native (not bridged) stables is the most honest indicator of big money's trust.

By these metrics the sector stratifies quickly: two or three networks with a real economy, the rest are storefronts for a token. We pull TVL data from DeFiLlama in the TVL section, coin metrics - in the catalog.