Why do average MPG readings 'jump'?

[monthefish]

I've noticed this particularly on BMW's, where the average MPG jumps between, for example, 55.4 and 56.5 and never reading anywhere in between.

No matter how much distance is attained, it never seems to be at any increment between the steps which, mathematically speaking, isn't possible.

Anyone know why they do this?

[5lab]

might be working out l/100km and then calculating it from there?

[stackmonkey]

I don't know.
I've had 2 BMWs, a Ford, a Skoda and a Seat with average readings and none of them has ever jumped in the way you describe.
May be you fell asleep for a while?

[matthias73]

My peugeot does this. From a computing point of view, I think its easier if you have predetermined miles per gallon for a certain speed with a certain amount of throttle/brake ect, rather than have a computer work out exactly how much fuel you are burning at that moment.

So the ecu will think: "Ooo, were doing 40mph in 2nd gear with 20 percent throttle on, i've been told this gives me 29 miles per gallon"

Or downhill: "Ooo this is fun, I'm in fith gear, doing 50 miles per hour, with 0 percent throttle. This gives me a million miles per gallon"

Shoot me if i'm wrong

[matthias73]

Sorry need to add some more. In my car it gives you a consistent readout of how many miles per gallon you are currently getting, and the average would simply be a mean average of the miles per gallon since i reset it.

So therefore its not completely accurate and also not made to such a high degree of accuracy that it gives every possible mile per gallon, but rather has preset figures it can give you based on the variables.

[Maxym]

It's just the way BMW group computers work, and the pre-set increments just happen to suggest a spurious accuracy. In my experience the consumption readouts on cars are optimistic anyway. My 911 gives mpg values to one decimal place but it's still about 2 mpg over reality.

Anyway, you wouldn't expect an outside temp readout to have finer increments than 0.5 deg would you? Why have mpg any different?

[gp1699]

if you havnt reset it in a while it wont be too accurate, i used to get lots of people saying that there 320d wasnt doing what it should be etc etc...

[175gt]

Just the way your particular computer's set up, though I think the unit change from l/kms is definitely a factor. My '04 Megane had precise readings but my '07 Laguna has an older computer/software for this (and a couple of other things) and moves in set increments (drove me mad at first but I've got used to it).

[monthefish]

For example:

You reset the computer, and then after 4998 miles you've used 198 gallons, so you've averaged 25.5mpg.

You then add a further 1000 miles using 40 gallons, which would bring your average down to 25.4 mpg (5998 miles on 236 gallons).

The point is, the bigger the numbers concerned, the smaller the increments should be but this doesn't seem to happen.

stackmonkey said:

I don't know.
I've had 2 BMWs, a Ford, a Skoda and a Seat with average readings and none of them has ever jumped in the way you describe.
May be you fell asleep for a while?

I've had 3 BMW's and a Porsche 996. They all do this.

[martin84]

The funny thing is when people stop at traffic lights and see their instant MPG go down to 0 and then they panic. Of course you're doing 0mpg, the car isnt moving!!!