Methodology
How the explorer turns messy listings into a usable market view
The core promise of evrange.info is not that every listing is perfect. It is that the site should still produce a defensible, decision-useful market picture when marketplace data is incomplete, mislabeled, or inconsistent. That means being explicit about source quality, fallback logic, what the range numbers do and do not mean, and where the product deliberately depends on the original listing instead of copying it.
Quoted range first
Current-range modes built on top
Conservative row filtering
Source data and cleanup
The explorer starts with active classified listing exports. Those rows are merged into one dataset and then normalised so the site can compare like with like: make and model names, seller type, asking price, year, county-level location, seat information, and outbound listing links all need to line up before plotting happens.
The long-term store is intentionally thin. Source images, seller names, exact addresses, and full raw listing text are not meant to live in the durable index. Instead, the workflow derives the EV-specific signals it needs, keeps the source link, and leaves full descriptive detail to the original marketplace page.
The filter is intentionally conservative about what gets kept. If a row still looks like a mislabeled non-BEV or a vague future-order placeholder after cleanup, the safer choice is to drop it rather than pretend it belongs in a used EV market chart.
How range references are chosen
The site picks a reference range in a strict order so the user can understand where confidence comes from:
- Use a plausible quoted listing range when the listing itself provides one.
- If the range field looks broken but the title or description contains a clear WLTP or km claim, recover that text quote instead.
- If the listing still has no defendable quote, borrow the median quoted range from the same make, model, and year when there is enough supporting data.
- If that is still not possible, fall back to a static model estimate.
This matters because some marketplace rows put battery size, trim codes, or obviously wrong low numbers into the range field. The site would rather admit it is estimating than quietly reuse a bad source field as if it were trustworthy.
Important boundary: the explorer does not claim to know the exact real-world range of a specific used car. It estimates what a listing probably represents, then shows the level of confidence through the reference source and the selected range mode.
How the four display modes work
The explorer exposes four ways to look at range because buyers need different views at different stages of the shortlist:
| Mode |
What it is for |
How to use it |
| Minimum current estimate |
Conservative shortlist starting point. |
Best for buyers who want winter and motorway caution built in from the start. |
| Balanced current estimate |
Middle view of the likely everyday picture. |
Useful once you are comparing a tighter shortlist and want less pessimism. |
| Maximum current estimate |
Best realistic upper-end view, still not a guarantee. |
Useful for checking whether a model stays viable when conditions are more favourable. |
| Quoted WLTP |
The original quoted headline range. |
Useful for understanding what the seller is likely quoting before you move back to current-range planning. |
The current-range modes apply the site's degradation and realism logic to the chosen reference figure. That is why switching between modes can materially change where a listing sits in the plot.
How buyers should read the chart
The best way to use the explorer is usually:
- Start with Minimum current estimate and set your real price ceiling.
- Use seats, county, approximate distance, seller type, or make/model filters to narrow the market to the cars you could plausibly buy.
- Inspect the reference source tags before trusting a listing too much.
- Switch to Quoted WLTP only when you want to compare the seller's headline claim to the product's more cautious view.
If your biggest question is whether the range assumptions make sense for your own routine, pair this page with the used EV range guide. If the practical challenge is family layout or boot space, jump to the family EV buying guide.
The explorer is best used as a shortlist and sanity-check tool. The original listing should still be opened for photos, live status, seller identity, and the most current specification.
Dedupe, cross-posts, and outliers
When the same car appears across multiple marketplaces, the site tries to keep one plotted point with multiple outbound links. The matching stays conservative: if the identity is not strong enough, the rows stay separate rather than being merged too aggressively.
Rows that cannot be plotted on a price-versus-range chart are dropped before shipping to the browser. Extreme price outliers can be split into a rejected bucket so the main chart remains readable while still allowing investigation when needed.
Important caveats
- Any non-WLTP range number is still an estimate, not a battery-health report.
- Asking price is not the same as the final sold price.
- Seat count, county, approximate distance, and seller type depend on source quality and can be incomplete.
- Same-model fallback estimates can still blur short-range and long-range trims when the listing itself is vague.
- Live classified stock changes quickly. A useful shortlist today can disappear next week.
Related reading