Originally Posted by Zeeper
However, you cannot draw the conclusions you want to draw from that graph. If more miles are driven in the city by the drivers on Fuely, the Combined Average mpg will be lower to reflect that. They are not robots driving on a closed course, 50% highway, 50% city.
Since you don't know the mix of driving, it is pretty meaningless data for the conclusion you are reaching for. Try clicking on individual cars, and see if the owner reported the driving mix, and draw your conclusions from that.
You have the data - if you want to refute it you get the "mixes" of highway vs. city and plug them into the 4,000 data points. What's "highway" anyway? There are state highways running through the middle of major cities - usually referred to as "Central" or "Main Street" etc. No stop lights? Freeway speeds? Everybody's idea of "highway" is different. City drivers vs. enthusiasts, CVT (maybe) vs. 5MT (definitely).
Nope, this matches up quite well with the Consumer Reports data. Give it up.