The Lie-O-Meter - how far off is yours?

My Prius v counts the first tank and lists my car's lifetime MPG as 43.8 It also includes 250 miles of THIS tank.
Fuelly.com omits my first tank as I did not pay for it, and lists my car's lifetime MPG as 40.9 It will not consider this tank until I pay for the next tank.
That is a 7% error but they are measuring different things at different times so I would not expect perfection.

My Gen 2 Prius was much closer to my per pump calculation. But then, the Gen 2 lied about how much gas was left in the tank, a much worse lie.

 

When i calculated mine last, reading was 27km/l but calculated was 29km/l, and a few times before it was never constant, the meter was either under reading or over reading, so I basically put the flaws in my method of calculation, cause honestly, I can't really tell when refuelling if the amount I put in is exactly the amount that was already used and to actually empty the tank everytime is not possible and of course i need to make sure all the fuel lines are empty, even fuel in the lines can account for 1L not in the tank when refuelling .

 

Mine is optimistic by around 2 MPG.

 

Ditto, about 4% seems to be the usual, though I haven't logged it like Old Wolf did in his thread.

 

Let's say I compute my MPG by looking at the odometer and keeping my gas receipts. Basically doing the same thing Fulley.com does. What is the likely error in this method? I think there are two sources of error (1) The odometer might be "off" by some amount. and (2) you never know if you have filled your tank to the exact same level as the last time. Let's assume the odometer is "good enough" and the gas tank might be over or under filled by 1/2 gallon.

The worst case error would happen right after a new owner filled the tank for the fist time. The amount in the first dealer supplied tank and the next on could be up to 1 gallon different. Their could be on the order of a 10% error in MPG in either direction. But after 10 tanks the worst case error is still 1 gallon but after ten tanks that is on the order of about 1% error worst case.

So,... Those people who say the trip computer "lies" need to consider the worst case error in MPG computation based on gas receipts. I don't think the odometer divided by gas receipt methods is very accurate until you have put in about 100 gallons total. After using 100 gallons I think you can assume the error is under 1% but not before.

It will be months before I use 100 gallons.

 

I just did this very comparison using Trip B reset at the pump, ran a whole tank (down to a couple of pips), then filled up at the same gas station, exact same gas pump. I even made sure the car was parked in the same relative spot so the tank was not out of level.

Results were 52.4 MPG indicated on Trip B to 50.9 MPG by manual calculation. <3% Error.

This was using Real Gas (a.k.a. "Pure Gas" a.k.a. "Ethanol Free") for other scientific purposes The difference in MPG between Ethanol blend last tank and Real Gas this tank was a staggering 13% increase in MPG! I suspect the car breaking in and me getting better at driving it may have artificially inflated that percentage, so I will run a tank of ethanol blend again next tank to verify the drop. Just under 2k miles at this point. Daily temperatures between 50F during the day and 25F at night. Mild grill blocking.

 

Mine has been off over 10 mpg's before. Dont really trust it anymore.