I have noticed that the average MPG calculated by the dash computer is not as accurate as I hoped it would be. For the last 5 tanks it's been reading 1.5 to 2.0 MPG high (e.g. 35.5 MPG when the actual turns out to be 33.7 MPG). I've noticed that as I accumulate more miles on my MMH and the actual fuel economy is improving, the descrepancy is getting worse. The 1st few tanks it was reading more in the range of 0.75 to 1.25 MPG high.

In previous threads here I thought that I read that the computer calculated MPG was supposed to be pretty accurate. I don't think that being off 5% to 7% is very accurate. Should I be taking my MMH back to the dealer and complaining?

The dash read out is not per tank. It is an average the spans tank fulls.

If you went back, and calculated by hand, your mileage back starting from your first tank, up thru to your last tank, how does that compare to the read out? Basically, adding up all the gas you ever bought, total, over the miles shown on your odometer.

Best single LD trip: 35.3mpg

Walt -

Your statement is true - assuming that I didn't reset the computer calculated avg MPG field each time I filled up the tank - but I did reset it at each fill up! So the computer should be giving me an accurate reading for each tank.

This may vary from vehicle to vehicle. I find that my display is optimistic also, but only by a little over 0.5 mpg.

What did it display immediately after you did a reset? 0 mpg? 99 mpg?

I have never did a reset.

The reason I ask, is that I wonder just how much data is actually reset, and what data is then used to calculate an average, when there is little-to-no data still available.

Best single LD trip: 35.3mpg

It will always vary a little because there are so many variables involved. The only way to really limit them is to always go to the same pump at the same station in the same weather, and to somehow figure out how to always fill the tank to the same level. It's nearly impossible.

People with the Nav / Hybrid radio have MPG in the Radio display.

This display holds "2000 data points". When it gets full, the oldest data is overwritten.

We do not know the sample rate.
If it saves 1 per minute, then 2000 minutes = 33 hours of driving data.
This is usually the 2-3 most recent tanks.

If it samples once per second, this holds 33 minutes of data.
This is almost certainly too short.

33 hours of data is most likely, however it looks like it is a weighted average. The most recent 1000 records ( 1000 minutes ) get more weight than the oldest 1000 minutes.

Why? No one knows. But the calculation method is one of Ford's best kept secrets.
-John
P.S. A manual reset wipes out ALL 2000 samples.

So, it averages in 1,999 "zero mpg" ????

That's what I am wondering. After a reset, what exactly does it use for those 2,000 samples points in its calculation? 1,999 "0" data points and 1 "40" data point, would result in an average of 0.02 mpg (which would drive a hypermiller to suicide ).

Best single LD trip: 35.3mpg

After a reset:
Sample 1 = 33 MPG
Sample 2 = 35 MPG
Sample 3 = 37 MPG
Sample 4 = 22 MPG
Sample 5 = 28 MPG

Running Ave = 33+35+37+22+28 / 5

Running Ave of 5 points is 31.0 MPG.

Since all five are in the "recent" group of samples, all carry the same weight.

So, in summary, would everyone say that the ave. mpg reading is accurate within .5 mpg for the tank, if you reset it at each tank refill? My first tank computer reading was off by about 1 mpg but I figured that was due to the differences in gas/tank/temp, and how full I actually got it. I don't have the nav system and just use the dash reading.

At this point I could fool myself into believing the computer, which makes me feel good (even more so) about the car

2008 Ford Escape Hybrid - Adopted 7-7-07