Measurement methodology
How to measure MEV execution routes
A neutral, reproducible template for recording endpoint timing, provider responses, transaction inclusion, and costs in your own environment.
Linkable summary
Scope of this page
This page explains what to measure and how to compare observations. It does not publish live FRB or provider results, rankings, or performance claims.
A repeatable measurement process
Define the test before running it and keep the same inputs when comparing routes.
- Pin the environment. Record network, region, client, endpoint, transport, software release, and test window.
- Separate each stage. Measure endpoint response, provider or relay acknowledgement, transaction inclusion, and final outcome independently.
- Use comparable samples. Keep transaction inputs and fee policy consistent; record sample size, retries, errors, and exclusions.
- Publish limitations. Include timestamps, units, conversion sources, and the exact method so others can reproduce the run.
What to record
| Measurement | Record | What it does not prove |
|---|---|---|
| Endpoint request timing | Region, endpoint, transport, timestamp, sample size, p50/p95 | Does not measure relay acceptance or transaction inclusion. |
| Submission response | Provider, route, response or error, retry policy, target block | An acknowledgement is not proof of inclusion. |
| Transaction outcome | Simulation, submitted transaction, confirmation and final status | Keep each stage separate; do not infer from simulation alone. |
| Execution costs | Gas, priority fees, failed attempts, refunds if applicable, conversion source/time | Report actual observations; do not imply savings or expected returns. |
Interpretation limits
A network or relay acknowledgement is not proof that a transaction was included or settled.
Endpoint latency alone does not establish route safety, execution quality, or profitability.
One test run is an observation, not a trend or a provider ranking.
Technical references
Verify route-specific behavior in current provider documentation before drawing conclusions.
Frequently asked questions
Does low endpoint latency guarantee transaction inclusion?
No. Endpoint response time, relay acceptance, and on-chain inclusion are separate events and should be recorded separately.
Can I compare providers with this template?
Yes, if you use comparable inputs and conditions, document the sample and method, and avoid treating a small sample as a universal ranking.
Does this page show FRB production benchmarks?
No. It is a measurement guide and contains no live FRB or provider result feed.
Measure from your own environment
Use the WSS latency test for client-observed endpoint timing; it does not measure transaction inclusion.
Open WSS latency test