If there’s profit to be made from lying, assume the worst. In Tesla’s case, the rosy picture was to convince people who had range anxiety over EVs as well as look better against the competition.

Have you ever tried to reach advertised mpg (or l/100km) on gasoline car?

It’s not the software. Everyone is lying.

My car generally exceeds its efficiency rating by around 15%.

The sort of malfeasance Ars Technica alleges here is not normal.

For many years, I got better-than-advertised mpg with a Honda Civic. My more recent vehicles, though, don't quite reach their claimed mpg. Is this "everyone lying" a new phenomenon, or is Honda an exception from "everyone"?

I frequently exceed it on mine as well as on rental cars. It really depends on one's driving style.

Coasting to upcoming red lights/stop signs and accelerating at a gentler pace make a big difference, as well as driving the speed limit on the highway.

Most of my ICEs and EVs reached advertised efficiency without much effort.

Let’s not normalize amount of deception Tesla does in their marketing and user hostile behaviors.

I usually do a lot better than the advertised MPG on my Mazda

I have. Perhaps not everyone?

There is a difference between the range that a standardized test reports which allows for comparing apples to apples and a range that my car shows me based on my driving style.

When the missus drives the car and then fills up the tank I see range X. When I drive the car and fill up I see range Y. Different driving styles.

Uber greyball, VW emissions test profilers, Tesla range shenanigans.

Seems we should default to assuming it lies until otherwise proven.

Look into closed source breathalyzers