I love my Fuelly and I try to "sell" it to friends when I can, but I have found it less useful than it could be in terms of using it an alternative to the EPA's formulaic (flawed) MPG ratings while shopping for vehicles. Sure it's less scientific but it's more real.
Questions that the Fuelly Database knows but isn't telling:
*What is the average MPG for a given model across all users
*What is the average MPG for a given model and year across all users
*What is the average MPG for two very similar models across all users (say the Mini and the Mini S)
*How do various options within a model effect the average MPG
For instance... If I went looking into a Toyota Tacoma pickup all the very different builds are lumped in together so I don't have a way to sort out the big '01 4 door 4wd with a v6 from the basic little 2 Door 2WD 4 banger on the lot this year. These two truck have an EPA estimate that is like 6 mpg apart so lumping them together seems to dilute the usefulness of all the Tacoma data. There are other such models with similar problems on a smaller scale but this is the worst I have found.
I'm not a web site or a database pro but I have some experience with both and I think this type of thing is well within the realm of possibility.
Matt and Paul, thank you for making us an awesome Free web tool! Maybe you have your reasons but this is my 2c... How about taking it up a level?
We have a fine line to walk between making data available in an easy-to-consume way, and not providing incentives to game the system once that data is easy-to-consume. I completely agree with you that for the purposes of shopping, Fuelly could be much, much better. But we also have to look at Fuelly as a method of gathering data. Sometimes those two ways of using the site are at odds with each other. We refer to this FAQ entry more than just about any other because listing things by fuel economy is our top request.
We also often say that if everyone reporting via Fuelly had to enter 20 different data points about their vehicle before they started tracking, we wouldn't have any data to look at. We definitely made some hard choices erring on the side of keeping things extremely simple when we built the site. I'd like to be able to compare trim levels of various models to see how it affects fuel economy, but we assume a large swath of people reporting at Fuelly don't know what a "trim level" is, and we think that's spectacular. We didn't build Fuelly for car enthusiasts, we built it for anyone to use, and that's a hard balance to strike.
These are things that we are constantly thinking about, and there are probably ways to improve "pre-sale browsing" without compromising reporting, but we need to think carefully about it.
People may not know a trim level but everyone knows whether they have a manual, automatic or continuously variable transmission. Perhaps only putting in only one parameter asking which transmission someone has wouldn't be too complicated or sour them to the site.
__________________ 2008 Toyota Prius Touring purchased 04/2016 at 101k miles, currently at 186k miles 2009 Acura MDX purchased 12/2019 at 159k miles, currently at 194k miles