Four different kinds of information
| Label | Meaning |
|---|---|
| Manufacturer / regulatory | A statement from a cited primary source, scoped to its date, market and vehicle configuration. |
| Reported by you | A measurement or diagnostic value you enter. EVVerity does not authenticate it. |
| Calculated | Arithmetic performed on the supplied values. Correct arithmetic cannot make uncertain inputs exact. |
| Estimated | A result that depends materially on imperfect observations or scenario assumptions. |
Estimating usable capacity
Range and consumption must describe the same conditions. Electricity from a wall meter includes charging losses and is not interchangeable with driving consumption. We compare the result with original usable capacity, never silently with gross capacity. Age, model year, mileage, climate and charging habits are context; we do not turn them into an invented degradation curve.
Sensitivity is not a confidence interval
The default ±10% capacity band is a transparent, editable input-sensitivity assumption. It is not fitted to an EV fleet or a laboratory validation study. For a central estimate of 52.8 kWh it produces 47.52–58.08 kWh. Degradation bounds are calculated against the original usable baseline, with negative loss bounded at zero. A central estimate above the original capacity is flagged; over 115% suppresses the estimate pending an input check. These are editorial plausibility checks, not diagnostic thresholds.
What our confidence labels mean
- Low: insufficient or conflicting evidence, an explicit scenario without vehicle calibration, or inputs requiring review. Age and mileage alone receive no numeric battery-health result.
- Medium: a range and consumption pair or a straightforward planning estimate, still subject to unverified inputs.
- High: in the battery tool, the user confirms a recent matching diagnostic report, and its reported SoH agrees with the range-derived percentage within five percentage points. The threshold is an editorial consistency check. High refers to supplied evidence completeness and agreement, not certified measurement accuracy.
A BMS value stays separate. We do not average it into the estimate or silently make the sensitivity band narrower. These labels are not calibrated probabilities. More relevant vehicle data can improve the estimate; more unrelated fields cannot.
Range and winter scenarios
Range equals usable energy divided by consumption. The driving-range tool explicitly deducts your arrival reserve. Its sensitivity band varies consumption by the chosen percentage. Temperature and profile are context; only the visible consumption-adjustment input changes the result. The winter tool applies the entered lower and upper range-loss scenarios to normal-weather range. The initial 15–35% band is illustrative, not a universal measured winter penalty.
Charging energy, cost and time
Losses are defined as a fraction of energy drawn from the supply: grid energy = battery energy ÷ (1 − loss fraction). Cost uses your tariff without an external price feed. Monthly cost is annual-distance energy cost divided by twelve. AC time uses the lower of charger output and vehicle limit, adjusted for losses. DC time uses the selected average-power factor below 80% SoC and half that factor above 80%. This deliberately simplified taper model is low confidence and is not a vehicle charge curve.
Ownership and replacement costs
True Cost of Ownership adds depreciation, interest paid during ownership and the entered running costs. It does not add loan principal to depreciation. The amortising loan assumes fixed APR, monthly payments and no balloon or fees. Remaining debt at sale is shown separately. Resale values are user scenarios. The EV-versus-ICE comparison instead adds purchase outlay and a narrower set of running costs; its exclusions are visible. Replacement costs come only from a user-entered quote until a properly scoped price source exists.
Vehicle data and sources
We prefer manufacturer documents, regulatory sources and official manuals. Each current model guide shows the source, publication context and review date. A historical specification is not silently promoted to all model years. Ambiguous gross/net values remain unverified. We do not copy a proprietary database or estimate missing warranty terms. The 19 detailed guides are an editorial selection; search-demand ranking has not yet been validated. The broader model directory verifies identity separately from technical data. A manufacturer listing does not establish usable capacity, charging capability, warranty eligibility or current availability for an individual car. Each directory record links to its source and review date. Historical and regional derivatives are included, but the directory is not an exhaustive register of every EV or variant. Only explicitly scoped, sourced references can fill a calculator; other models require your own verified inputs.
Limits and review schedule
EVVerity cannot inspect damage, verify faults, authenticate seller documents, determine warranty eligibility or forecast a failure date. No VIN or battery-passport lookup is connected. Review all high-impact specifications before adding a preset; target a quarterly review of active sources and an immediate review when a correction or manufacturer change is identified. This is an editorial schedule, not a claim that an automated update service is running.
Source and calculation review: 24 September 2026. Professional validation and broader real-vehicle calibration are still needed before making stronger accuracy claims. Read the practical guides or report a correction.