How We Score Electric Scooters

The real-world range correction

This is the most important number on the site. Manufacturers measure range under conditions almost nobody rides in: a rider of roughly 165 lb, a low constant speed, flat ground, mild temperature and no wind. Independent measurement consistently lands well below the claim.

We therefore display every claimed range alongside a computed band:

estimated_real_range_low  = claimed_range_mi × 0.55
estimated_real_range_high = claimed_range_mi × 0.70

A scooter claiming 62 miles is shown as “Claimed: 62 mi · Est. real-world: 34–43 mi”. Push towards the lower end if you are heavier than 165 lb, ride hills, ride in the cold, or ride near the scooter's top speed. This band is a computed estimate derived from the manufacturer's own claim — it is not a test result, and we do not present it as one.

The scoring formula

Every score is deterministic and computed from stored specifications. Nothing is hand-tuned, and no retailer can influence a position. Each component is percentile-ranked within a scooter's own category, so a lightweight commuter is not punished for failing to be a 50 mph performance machine.

Components

• Range score — percentile of estimated real-world range (claimed × 0.60) within category.
• Power score — percentile of peak watts, with a 15% bonus for dual-motor drive and a contribution from stated incline capability.
• Build score — IP rating × 0.4 + suspension × 0.3 + brake tier × 0.3. IPX4 = 40, IPX5 = 60, IPX6 = 80, IP66 = 100. No suspension = 0, front or rear = 50, dual = 100. Hydraulic = 100, disc = 75, drum = 55, regenerative only = 35.
• Portability score — inverse weight percentile within category, plus a bonus for a folding frame.
• Value score — the average of the four scores above divided by current tracked price, percentile-ranked within category.

Weighting

overall = range × 0.25
        + power × 0.25
        + build × 0.20
        + portability × 0.15
        + value × 0.15

Exclude and renormalise — never zero, never guess

Where a manufacturer has not published the input a component needs, that component is dropped entirely and the remaining weights are renormalised to sum to 1.0. A missing figure is never scored as zero and never imputed. So a scooter whose maker does not publish peak wattage is judged on the components it does document, not punished for the gap. Each model page lists exactly which components were scored, and names any that were dropped and why — a data gap made visible rather than hidden.

Power fallback. Power is ranked on peak watts where a manufacturer publishes them. Where only nominal (rated) watts are published, the model is ranked against other models also reporting nominal watts — never nominal-against-peak, which would systematically punish the honest manufacturers who quote the conservative figure. Only when neither is published is the power component dropped.

Within-category comparability

Every percentile is computed inside a scooter's own category — budget, commuter, premium or performance. This has an important consequence, which we state plainly: a budget scooter's 90 and a performance scooter's 90 are not the same claim and are not directly comparable. A 90 in budget means “among the best budget scooters”; a 90 in performance means “among the best performance scooters.” When a category has fewer than five members — too few for a meaningful percentile — we score its models against the union of that category and its adjacent tiers, and record which population was used.

Value

The value component is quality-per-dollar — the average of a model's other computed components divided by its current price — then percentile-ranked within its category. A $3,000 performance scooter is judged on value against other performance scooters, never against $400 commuters, so price cannot structurally lock the top of the market out of a good value score. Where no current price is tracked, value is not computable and is dropped per the rule above.

Total cost of ownership

The two-year figure on each model page is purchase price plus estimated electricity plus estimated consumables. Electricity assumes about three full charges a week over two years at the US average residential rate of $0.17/kWh. Consumables assume tyres, tubes and brake pads across two years, lower for solid tyres than pneumatic ones. It is an estimate for comparison between models, not a quote.

Price tracking

Amazon prices are refreshed by a scheduled job that runs once a day and writes one append-only snapshot per model per vendor. Snapshots are never overwritten, so the price history and the “lowest price we've tracked” badges build up honestly over time. The job aborts and writes nothing if the retail API errors for any model in a run — a partial update would make cross-model comparisons inconsistent — and it rejects any price that moves more than 35% from the last recorded figure, or that falls outside $80–$8,000, logging those for manual review rather than publishing them. If a model's newest snapshot is more than a week old, its price block says so rather than presenting a stale figure as current.

Official-store prices are a known limitation. We do not scrape manufacturer websites on a schedule — doing so reliably and respectfully is a project in itself — so “official” prices are seeded manually and refreshed less often than Amazon figures. Where only an official price exists, treat it as indicative and confirm on the manufacturer's own page before buying.

The calculators

The range calculator starts from the manufacturer's claimed range and the 55–70% correction above, then applies five documented factors: rider weight (about −4% per 20 lb over a 165 lb baseline), average speed (about −3% per mph over 15 mph, because air resistance rises with the square of speed), terrain (flat 0%, rolling −10%, hilly −25%), temperature (below 40°F −20%, 40–55°F −10%, 60–85°F none, above 85°F −5%) and tyre pressure (−7% if not maintained). It always returns a band, never a single figure, and it is an estimate derived from the claim — not a test result. It assumes a healthy battery; lithium packs lose roughly 20–30% of capacity after about 500 cycles, so older scooters fall short.

The cost-of-ownership calculator sums purchase price, electricity (about 25 Wh per mile at the wall, at your editable electricity rate) and consumables over your chosen period. Consumables differ by tyre type — about $70 a year for pneumatic tyres against $25 for solid — because that is a real cost difference. Optional comparisons weigh the total against driving at a documented $0.25 per mile (fuel and running costs, excluding depreciation and insurance) or a $70-per-month transit pass. It is an estimate for comparing models, not a personal finance projection, and it does not model battery replacement, resale or insurance.

Source policy

Specifications come only from manufacturers' official websites. News sites, aggregators, retailer listings and review blogs are not used as sources of record, because they routinely repeat each other's errors and mix regional variants of the same model.

Regional variants matter more than most buyers realise. Several manufacturers publish EU-configured figures on their global sites — a 25 km/h speed limiter and EU rated power that do not describe the US-market scooter. Where a spec sheet is explicitly labelled for another region, we store the US field as unavailable rather than converting a number that describes a different product.

Every model page links to the exact manufacturer page its specifications were taken from, so you can check any figure yourself.

Affiliate disclosure

We may earn commissions from purchases made through links on this site, including as an Amazon Associate. Rankings are computed from specifications and are never sold. A retailer cannot pay to move a scooter up a list, and no manufacturer sees a page before it is published. Where a scooter has no affiliate link at all, it is ranked on exactly the same basis as one that does.