In my opinion this is a more meaningful way to compare the MPG of the cars because at the end of the day it really comes down to how you drive, so what is most important is the potential of the car. I don't think any single MPG number can best fit a car, since the car is not the only factor in real-world driving.

Prius I --- Most: 45-48 --- All: 41-57
Prius II -- Most: 45-54 --- All: 36-70
Civic ----- Most: 45-51 --- All: 34-60
Accord -- Most: 30-36 --- All: 18-38
Insight -- Most: 51-75 --- All: 50-104
Escape -- Most: 30-33 --- All: 26-41

No point with the Toyota SUVs, there's not enough data yet. "Most" is based on the graphs at the top of each car's page. Insight was hard to generalize because the (few) cars are spread out somewhat evenly. More in the database would make that range easier to determine. In the "All" range I did not include cars that have had less than 3 tanks, or have used the lifetime function.

 

I find it interesting that with the exception of the Prius 1 and the Escape, the high end of the range is approx double the low end of the range.

I agree that much of the difference comes from "how" you drive. But the other major contributors are the length of the trip and the temperature. In fact, in my experience with Prius II, temperature has the potential to trump driving style. The best example I can give you is a comparison between me and my wife. I drive for FE, my wife just drives to get there ASAP. On a warm summer day (80's) I might get 70 MPG driving to work. My wife would get 50 MPG driving that same route. On a cold winter day (teens) I would be hard pressed to reach 50 MPG. My wife would still get about 45 MPG. So my 20 MPG advantage drops to maybe 5 MPG in the winter. This is because the engine has to run longer and harder in the winter to keep itself and the catalytic converter warm.

So when comparing the ranges of FE, it might be interesting to look at the weighted ave temps posted for each tank. I would expect that the higher FE's would tend to have higher ave temps associated with them.

 

Yeah, the temperature does make a huge difference too, but since most of the cars (not all, of course) have data for all year, then that averages out over time. It would be nice to be able to do statistical filtering on the mileage database to compare all the cars driven under similar conditions, unfortunately I don't think I have the patience to crawl through all the data for that purpose.

Another big factor that can trump "how" you drive is the terrain that you drive on. Around here there's a lot of steep hills that are impossible to take a run at, so they tend to suck the gas a lot more than people driving around on the prairies.

 

Another way to say this is: what is the standard deviation for each model? For some cars like the Insight, the "outlier" cars are shockingly far apart from each other.

A consistently high mean with a lower standard deviation will give potential buyers comfort that the driving style and location is not going to result in a massive variance from the norm.

 

If you look at the database and hover over the graph for a particular model you can see that information as reported by members here.