What a DeFi route check actually measures
A route check scores a supported swap route on the dimensions that quietly erode execution quality: modeled MEV exposure, expected route leak in basis points, slippage versus the mid quote, gas, and the venue/route assumptions behind the number. Routescore turns those into a single route-quality grade plus a modeled expected-leak figure for the trade size you enter.
The point is comparison, not prediction. You see how a quoted route compares to supported alternatives at the same notional, so you can tell whether the route you were about to use is roughly in line, or whether a lower-leak route exists for the same pair.
- Route-quality grade (A–F) with a modeled base-leak band, not a quote.
- Modeled expected leak in USD for your trade size.
- Side-by-side comparison against supported route scenarios.
- An optional live quote check (1inch · 0x · CoW) for provider timing — the displayed grade stays the modeled comparison.
How to read the route-quality grade
The grade summarizes modeled route quality; the expected-leak figure is what that grade implies in dollars at your notional. A route with a lower modeled leak than the canonical Uniswap V3 baseline is surfaced as a cleaner alternative — but Routescore shows you the comparison and the caveats and lets you decide. It never picks a route for you or sends a transaction.
Larger trade sizes change the picture: leak and slippage scale with notional, so the same route can grade differently at $1k versus $100k. Move the trade-size slider to see the modeled band shift before you commit to anything in your own wallet.
When the modeled score is enough — and when it is not
The score is decision support for supported routes at a point in time. It is useful for comparing route scenarios and keeping a record of why you leaned one way. It is not a guarantee of execution quality, and it cannot see everything: unsupported tokens and chains, contract or bridge risk, and future MEV behavior are explicitly out of scope and named on every artifact.
Treat the route check as the pre-decision step. When you act, you act in your own venue with your own settings — Routescore stays read-only and non-custodial throughout.
From a route check to a reviewable decision record
The check is more valuable when it leaves a trail. Save the scenario and you get a decision record that captures the route, notional, model and feature versions, source freshness, caveats, and confidence state. Later you can attach an outcome label and review whether the modeled call held up.
That loop — run a check, save the record, review the outcome — is the part that compounds. It turns one route check into calibration evidence that improves the methodology over time.