The mpg figures are highly dependent on how and where you drive your car, manufacturers figures are only a guide, driving with a heavy foot in a lot of traffic may only give you half the mpg that driving gently on country roads will.

Remember that they are not manufacturers figures, they are independantly produced figures that comply with strict EU rules. They are intended for calculating VED and company car tax, they should only be used when comparing cars, don't expect to get what you see written, you will only be disapointed, best to be realistic.

The responses so far, whilst accurate, do not answer the OP’s query.

When the Real MPG Register was started, I thought “What a great idea!”. Like the OP, however, I was dissappointed to find that the spreadsheet doesn’t indicate the sample size. At the moment it only needs 2 people to log their mpg for a figure to be given for a particular car. That’s not really fair on the manufacturer nor particularly useful for a buyer.

Is it too late to put that right?

I compared the HJ values against the VCA test values (for around 1000 models) and the overall relationship was close but for around a 13% shortfall in MPG compared to VCA but there also appears to be a bias - higher VCA MPG values yield a larger shortfall and newer car models yield a larger shortfall. If true this is very important result in relation to car emission estimates and forecasts but the lack of sample size info reduces confidence in the results. Would be good if the number of correspondents was simply logged next to the range values.