[b3lha]

I think the percentage error idea goes back to the days of cable-driven speedos.

Our speedo is driven by a digital electronic signal and could potentially measure your speed accurate to a fraction of a mph. IMHO the only reason the speedo reads optimistically is because it has been designed to do so. I think that a constant error would be easier to engineer than a percentage error because it's just a matter of moving all the numbers back a little on the scale.

When I drive through those signs I find that my reads 5mph over and it's a constant 5mph, not a percentage.