This is probably a stupid question, but. . .

We are not supposed to run our gas tanks dry.
We are not supposed to top off the tank.
So, can we really get accurate MPG calculations?

Consider this scenario:

Let's say my tank is getting low.
But I buy only, say, 7.21 gallons of gas on January 1.
(I don't fill it up.)

Then I drive a while, but the tank is NOT getting low.
Then I buy, say 3.12 gallons of gas On January 10.
(Again I don't fill it up.)

Then drive till I have 1/4 a tank left (or whatever).
Then I buy, say 6.53 gallons of gas On January 18. . . . etc..

As long as I document how many many gallons I bought - and how many miles I drove since the last gas purchase - it doesn't matter if I filled the tank or if I topped it off.

Also, I think it does not matter that we are not running our tanks dry, and not filling up with any particular amount of gas left.

I will still get accurate MPG calculation.

Right?

In other words, I guess my question is, do we need an exact repeatable physical reference for low (such as a bony-dry empty tank, or the instant only one bar is left on the fuel gague) or an exact repeatable physical reference for full, such as in topping off the tank.

Actually I think it IS necessary to run bone-dry to topped off to get accurate MPG calculations, I think I just pushed the scenario to the extreems to make the inaccuracy more obvious.

Need fewer troops to support.
Drive a hybrid.
Best tank 71.65 MPG.

Before this car I spent two years learning hypermiling on my 2004 HCH1.