But as we all know, some cars have very few entries. Some, the driverrs meticulously record odometer miles and litres put in, and calculate. Some, the drivers work off the onboard mpg. Some from people who drive like grannie, others who drive 'full-on'. Some figures are even probably false - there are m****s out there who just enjoy messing around and doing stupid things like that.

Some will have even been entered against the wrong car - I've seen it often enough with 'owner reviews' !

So all that user data is a very rough guide. Attempting to work off it for a purchase decision would be nonsense. Add in the fact that a single model in 'Realmpg' will cover a range of cars with differing specs, wheel sizes, etc. and it really should be taken as such, and not really relied upon much more than the 'official' mpg figure.

Here's an example. The entry for the F30 BMW 320d has an 'official' economy ranging from 61-67 mpg. Now, that covers 2 different engine outputs, everything from 'base SE with 16" wheels' spec to 'fully-loaded MSport with silly 20" wheels and BMW bodykit' variants. Those are obviously going to give very different fuel economy figures. The wheel/tyre alone will probably give a 5-10% difference in economy.

And that's before you've put different drivers in them, and one spends it's life doing 70 on the M6, whilst the other is hammering it round the home counties in constant traffic.

It's a really rough guide. Basically says "The official figures are nonsense. Here is what some other people have reported". But as much as that it's a tool for HJ (and others with similar databases) to use to campaign for the 'official' test to be reviewed (and to get publicity).

In addition to which, if the OP is going to tow a large caravan with it, then you can probably take another 20-30% off the 'Realmpg' figures.