Andy...as bad as I wanted to laugh at your post, I will hold it and explain what's going on.

We were talking about how much miles per gallon a car REALLY gives. To get this true value, we divide the numbers of miles travelled for the whole tank by number of gallons in the tank.

The MPG that the computer read-out gives is as bad as one can get. Most of us know, from the earliest days, BMW had the legacy MPG gauge. The higher you rev, the lower the number you get. This number is an instantaneous screen-shot of MPG which is never a steady number through out your drive.

Now...there is another type of MPG read-out/display. This type keeps track and remembers how many miles you have driven with what kind of tachometer behavior you have driven. In contrast to the BMW type that gives you the screen-shot MPG, this type gives you a computational averaging. It still is a very very bad number. Even dumb blonds should not refer this number for any practical use.

I hope you understand and could appreciate all the differences that I am trying to clarify here.

Thank you, Andy.

 

Hi fly-major-fly,

May I add further clarification.

Fuel consumption can be measured in a number of different ways dependent upon what the result is going to be used for.
The only "accurate" way (and even then there are some small inaccuracies in this method least anybody wants to point them out) is to fill the tank to BRIM...not auto cut off...and to then immediately start driving the car (to stop the fuel expanding and damaging the evaporative loss system) till the tank is nearly exhausted and then to REFIL to the BRIM (take care of fuel expansion once again).
MPG is then the number of miles covered divided by the number of gallons added to the tank FOR THE DRIVING STYLE AND CONDITIONS EXPERIENCED.
Its important to state wether the MPG is Imperial or US. The US gallon is 0.8 of an Imperial (UK) gallon.
Running the test over several tanks improves the accuracy of the reading because it averages out SOME of the variables.

Some notes on accuracy.

Odometer / Mileage covered.
Old mechanical type speedos could easily be up to 5% out.
Modern digital ones are most probably better than 1%. However remember the size (diameter) of the tyre changes with wear and the diameter of the tyre is directly proportional to the reported mileage.
The best way to confirm odometer reading is to use the mile markers on the motorway (16 markers/mile in the UK & Europe) over a 100 mile+ journey.
A modern car with OE wheels and tyres is MOST PROBABLY accurate enogh for casual testing.

Fuel used
Did you brim the tank?
Was the car on level ground?
What was the temperature of the fuel? Underground temperature will be relatively static but fuel already in the tank may be very warm. Dependent upon quantities in the car/added this can impact the amount of fuel used.
Can you be sure the fuel dispensing pump is accurate?
You MUST NOT ASSUME THE VOLUME OF GAS USED IS EQUAL TO THE TANK VOLUME (in the handbook)

DRIVING STYLE and CONDITIONS
What were the conditions, speeds, journey length, environmental conditions (rain, headwind) and your style (harsh acceleration etc) during the test. These factors can make up to 50% difference easily.
What equipment levels were being operated? Headlamps and air conditioning can worsen fuel consumption by up to 10% dependent upon engine size and speed etc.
Cold start and town driving can easily EXCEED 50% worsening so when you see 8mpgus on your readout don't totally dismiss this. That is POSSIBLY what the car is doing at that point in time.
Bottom line. When you quote MPG ensure you state the conditions and instruments used.

I agree MPG computers have improved and the current MB design CAN BE quite accurate. The "Since Start" and "Since Reset" is a nice convenient feature.
On my S Class I have determined that it is well within 1% on a regular long distance journey. However I only know this after doing such a controlled test. Another S Class might well give very different results.
On a previous 2000 m/y European Ford the reading gave a 10% better than real readout. I understand Ford have a policy of making the gauge "flatter" the results. Customers don't usually question the reading (if good!)and it obviously holds the car in a good light and avoids complaints of poor fuel consumption!!

MPG read out on cars BEFORE electronic injection were based on instantaneous manifold vacuum and the readout was only indicative. Good, normal, bad dependent on how much throttle relative to revs or speed, you applied.

I could continue but I hope I have made my point.

Please note that I advised the original poster to agree a procedure with MB for determining the real world overall fuel consumption. Normal everyday use including some town, some highway, driven in a responsible way should be better than 20mpgus. If it does not then MB should determine why. It may well be that the user only does short, cold engine journeys for instance.

In Europe all cars have to be sold with environmental information including fuel consumption. This consumption is determined in a laboratory and is therefore COMPARATIVE between makes and models. The journey is simulated on a climate controlled rolling road following a prescribed test plan. eg Start from cold accelerate to 30mph (rate of acceleration is controlled) hold 30 for 2 mins then decelerate to idle. Idle for 2 mins etc. etc etc.
However there is a lot of debate from customers who do not realise the published figures.
If we assume that the results have been obtained and verified professionally then this further goes to show just how much real world driving STYLE impacts results.
Should you wish to review the consumption figures of models available in the UK go to http://www.vca.gov.uk/.

 

Well put, Stu.

May I add that -- I dont want people to nick-pick on this line I wrote.
"We were talking about how much miles per gallon a car REALLY gives."

By "a car", I was referring to the car in question. And the test method I described is very practical and very accurate. It is the operator's job to accurately measure his/her total gas and total miles travelled. I did not and do not want to specify how. Stu did a great job on the details.

I did not pretend to suggest on how to get a standard MPG figure for a particular car model. So, terrain features, wind resistance, and tire conditions DO NOT MATTER.

I was giving a formula to get MPG for a particular car in question. It's brakes may be stuck; it may locate on a hilly area; it may have triangular wheels. In this topic, we just want to know how much MPG one particular car is giving under all the odd conditions around it.

 

Further more....

Most women assume that when the fuel warning light comes on, the tank is almost empty. Not true.

I assume Valerie is a woman. So I have to warn Val and the general audience that...usually there is(are) at least 2 gallons left when the light starts to come on. Do not assume gas tank is empty at that point. You have to use the procedure Stu described to get to the figure on the real number of gallons in your full tank. Fill it to BRIM, drive, and fill it to BRIM again. The total number of gas filled "at the second time" is your number of gas used for the number of miles travelled since last fill. (With this method you do not need to drive until the tank is empty, you drive any distance. You wont risk empty tank...dangling in the middle of nowhere.)