Charging losses are not a rounding error
The default assumes 10% losses between the wall and the battery, the figure fueleconomy.gov uses, and it is doing real work. AC charging loses energy in the onboard charger and to battery thermal management: about 5–10% on a Level 2 wallbox in independent tests, and more on a 120 V Level 1 outlet, where ADAC measured 12–24%. So the kWh your meter records is meaningfully more than the kWh the battery gains.
Which efficiency figure you enter matters here. The car's own trip computer reports miles per kWh out of the battery, so losses need adding on top. An EPA rating is measured at the wall and already includes them; if you use one, set losses to 0 or they are counted twice.
Cold weather makes it worse, sometimes considerably: preconditioning the pack before it can accept charge draws energy that never reaches the battery at all. A winter cost per mile noticeably above the summer figure is normal rather than a fault.
Comparing against petrol honestly
The comparison uses your efficiency in miles per kWh against a petrol vehicle's mpg and fuel price. The defaults — 3.5 mi/kWh and $0.18/kWh (the July 2026 US residential average, from the EIA) against 30 mpg at $4.35 (AAA's national average in October 2026) — produce the large gap people expect, and both sides are sensitive to inputs that vary hugely.
The electricity rate is the biggest lever and the one least likely to be a single number. On time-of-use tariffs, charging overnight and charging at dinner time can differ by a factor of three, and public rapid charging is frequently two to three times the home rate, which at the upper end is enough to erase the advantage against petrol entirely: at the default 3.5 mi/kWh and 10% losses, electricity above about $0.46/kWh costs more per mile than a 30-mpg car on $4.35 petrol.
Use your actual observed mi/kWh rather than the manufacturer figure. Real-world efficiency falls with speed, cold and load, and the gap between rated and observed is usually larger for EVs than for petrol cars.