To clarify how the mileage database averages:
For each model, the total fuel economy values for each car are added up and divided by the number of cars in that category. This ensures that each car is given equal consideration. If there is a 20-year-old in Maine who gets 30 MPG over 10,000 miles and a grandpa in California who gets 50 MPG over 100,000 miles, the average is 40 MPG.

 

NO! NO! NO! NO!

 

lol! I'm working on a month since my last fill-up also! I'm getting such great mileage now that the weather is warm I think I ought to drive more to take advantage of it!

 

Hmm, I would be interested in seeing the difference if the database were adjusted for miles driven. I woudl suspcet that woud tend to exaggerate the higher-MPG values more because people who drive longer will tend to have the engine warmed up more etc. which is condusive to good fuel economy. Someone with a 100mi each way drive could average 60 and outweight a bunch of drivers with 5 mi each way drives getting 45 on a per-mile basis.

Donig it based on driver average to driver average, weighed equally makes the most sense, as what you are really interested is not how many miles driven, but the actual usefulness of the car (transporting a person to work and back, to run errands, etc. however far that is). Having some sort of cutoff though probably is still a good idea.

 

I, for one, am more interested in what CGameProgrammer is calling "total mileage," which would be more correctly called a "miles travelled weighted average" (or something like that). That figure can be thought of as combining all vehicles of a given type in the database into one vehicle of that type:

606 Prius II's become just ONE "big" Prius II which has travelled a distance equal to all of those cars. Unfortunately, the way the database is constructed (or at least displayed), users can't get that info.

I think that is more useful precisely because it DOES give more weight to cars with more miles driven. Those cars give you more information than cars which have not been driven as many miles, and that information should be used.

For comparison, we can look at fueleconomy.gov's database: The average of the mpgs of the 95 Prius II's (2005 model year) in their database is 48.2 mpg, but the displayed average for those cars is 47.7. While some of that difference could be due to rounding, I think it is more likely that they're using {total miles driven by all in database}/{total gallons consumed by all in database}. Note that you are required to to put in both miles travelled and gallons consumed in their database or it doesn't accept the entry.

 

Just my 2 cents.

Even though individuals do not continue to input results, it does not diminsh the validity of the data they have entered. Over time the frequent posters data will overshadow the less frequent. Unless there is some reason to suspect fraudulent entries, I recommend that all data entered be considered valid and counted. This will give us the most accurate real world data.