Measurement you can cite.
Every figure Lahza publishes is reproducible: the procedure, the formula and the evidence are public. Lahza sends nothing to any chain. It listens, from inside the network, and times what users do. This page is the specification.
Independence
An observer cannot influence what it measures. Lahza holds no wallet, sends no transaction and runs no relayer on any chain. Observer nodes never run on infrastructure of an observed chain. No chain pays to be observed, and payment never touches a number.
Reproducibility
The score is a deterministic formula over the observed record. Auditors and licensees recompute any score from the raw dataset; every incident is published with the on-chain transaction that proves it.
Proof
Every observation names a real transaction on a real chain, and every hour the Merkle root of the raw record is committed on-chain. What happened is provable; that the record was never altered is provable too.
How a measurement is made
Each observatory runs its own full-capability node inside every covered chain's network. For every user transaction it records three timestamps from its own clock (NTP-disciplined, sub-millisecond):
Finality is t₂ − t₀: what a user waits, measured on real users. Per chain and per observatory we report p50, p95 and p99 over the window. There is no cadence: every transaction is measured as it arrives.
What each interval measures
Finality as a user feels it mixes network propagation, block production and consensus. The method separates them and says which is which.
inclusion — first seen to first inclusion: mempool and block production.consensus — inclusion to irreversible acceptance.finality — what a user waits. This is end-to-end user experience, not pure consensus latency; the score uses it deliberately, and the two components are published beside it so consensus can be read on its own. t₀ is the moment our node saw the transaction, which follows the user's send by the network's own propagation time.Observatories
Frankfurt, Virginia and Singapore today; the roadmap is seven continent-scale internet regions. Each figure is reported per observatory and as a cross-observatory median, so a single vantage point never defines a number, and a chain that is fast from one continent only is visible as such.
The score
Four components, each normalised to 0–1 against a fixed public scale, weighted and summed. Weights are versioned; a weight change is a new methodology version, never a silent edit.
finality — 1 at p99 ≤ 1 s, 0 at p99 ≥ 10 s, log-scaled betweenuptime — blocks produced on schedule, and transactions that arrive are included; 90 daysmessage delivery — share of observed cross-chain messages delivered within 60 s over 24 hvalidator health — set size, stake concentration (HHI) and churn against public thresholdsCalibrate it yourself
The formula above, live. Move an input; the dial follows. No number on this site asks to be trusted — it asks to be checked.
Why it cannot be gamed
A chain can favour a transaction it recognises. Lahza has none to recognise.
Confidence
A difference smaller than the noise is not a ranking.
Coverage
Every mainnet L1 is in scope. A chain is either observed from inside its network — one class of measurement, the same for all — or listed until it is.
Three-leg ICM timing
the message is finalised on the source chain
a relayer observes the signed message and submits it
the message executes on the destination chain
Every cross-chain message our nodes observe is timed leg by leg. A stalled message is attributed to the leg that broke — the alert says where, not just that. Lahza never operates a relayer; where no relayer serves a route, delivery is not applicable, M is omitted and the remaining weights are renormalised, with the label shown.
Region gap
For each chain: (max − min) / median of p50 finality across observatories. Below 0.10 is consistent; above 0.25 is flagged. It reveals chains whose validator geography favours one part of the world.
Incidents
An incident opens when block production stalls, when transactions arrive but are not included for 60 seconds as seen from at least two observatories, or when p99 finality exceeds 5× its 30-day baseline. It closes after five clean minutes. Duration, affected metric and evidence transactions are published.
Validator sets
Permissioned validator sets are small by design; V measures them as they are and publishes the set size next to the score (e.g. PoA · 5 validators). The reader sees the design choice; the number is not softened.
Retention, verification, versions
Raw observations are kept for 90 days, then reduced to hourly aggregates kept indefinitely. The public layer shows aggregates and every incident's evidence transaction; the raw dataset is available to licensees and auditors in full. The observer and scoring engine are proprietary; this specification, the evidence transaction behind every incident and the hourly on-chain commitment are what make them checkable. This is v1.0.