How accurate it the on-board fuel usage display? Mine is reading a consistent 18.2 MPG. I do mostly city driving. If its accurate I’m impressed. Its not as bad as consumer reports listed it (15MPG city)

 

The MPG computer is notoriously inaccurate. And unfortunately it reads high. How much is up for debate, depending on who you ask on this (or any other) forum.

Regardless, I still find it useful for a "relative" value. If it reads 18 before a mod and 15 afterward, you know the mod cut 3 mpg.

I've been noting my mileage at the last 5 or 6 fill-ups, but so far I've been too lazy to actually do the manual mpg calculation!

Maybe subconsciously I really don't want to know...

 

In the city? I'm impressed.

As to accuracy, it is. However, you need to understand a few things first to know why there are seeming differences.

It is set up for the wheels and tires that it was shipped with from the factory. And changes and it will require recalibration.

Unless it is reset it reads the average MPG from the last reset. So that means total avg city, highway, offroad, etc.

If it is reset when you fill it up, then took a ride you should see the true avg MPG and for the first few miles it might show "instant" MPG readings.

Of course you could always dbl check it by the fill to top off level, drive it and fill to top off leve again and do your math. It won't be off that way by too much unless your gears, or tire size or both are changed and the speedometer has not been recalibrated.

 

Mine for the most part is 1.5MPG off. It reads high.

 

I see a lot of people here complain about how inaccurate the onboard mpg calculator is, but I don't quite understand their logic. They come to this conclusion after doing hand calculations between fillups, and come up with a different number than the computer, and somehow conclude that the computer must be wrong....hmmmm.

In order to know how much fuel you've consumed, you need to know EXACTLY how much you started with, and EXACTLY how much you ended up with. Most people try to do that by filling the tank until the pump shuts off, then driving a few hundred miles, then filling the tank again until the pump shuts off to see how much fuel was consumed. Sounds simple enough, but you are assuming that the pump shuts off at EXACTLY the same level both times. You are assuming that you started with EXACTLY 20.2 gallons in the tank, then filled it back to EXACTLY 20.2 gallons again to determine how much fuel was consumed. But I'll betcha the shutoff switch on the pump isn't all that precise. You might have had 20.1 gallons after the first fillup, and 19.4 gallons on the second fillup - guess what, you're off by 0.7 gallons, throwing off your mpg calculation by a lot. This hand calculation is terribly inaccurate. The only way to accurately determine how much fuel is in the tank would be to weigh your Jeep before and after you fillup and calculate the volume of fuel based on the weight of a gallon of gasoline (don't know that number off the top of my head). Either that, or start with a completely dry tank and rely on the pump to tell you how much fuel you put in, and then drive it until it's completely dry (but even then, you have to assume that the pump itself is accurate, which it probably isn't, and even when you run out of gas, there is still some in the tank, just not enough to feed the fuel pump - still not an accurate measure of consumption).

Now, the onboard computer, at least in theory, has access to very precise fuel usage information - the engine is controlled by a computer, and that computer knows EXACTLY how much fuel is being burned at any given time. Whether or not the mpg calculator has access to this precise data is open for debate, but the fact is the computer KNOWS how much fuel is being burned. The odometer is also controlled by the computer, and is the same odometer used in hand calculations done by humans, so we can eliminate its potential inaccuracy as a constant.

So in my mind, it is more likely that the computer's calculated mpg is much more accurate than any human calculated number, simply because it eliminates many variables. And I can think of no (good) reason why the engineers would choose not to use consumption data straight from the ECM do do its mpg calculations, so I assume it is accurate, and the manual methods used by humans are flawed.

I'm not trying to fault anyone for thinking they're smarter than the computer - we all know computers are dumb (especially when they're running Windows), but I think in this case, the computer has access to more accurate information than we do.

 

Yes, you've completely missed my point . The point is, it's impossible to know, with any precision, how much gas you've put into the tank using the "fill until the pump shuts off" method. It's inaccurate by definition, because every pump is different, and any given pump is likely not going to shut off consistently at the same level of "fullness" every time - you never REALLY know when your tank is "full", so you never REALLY know exactly how much gas you've put in. And if you're off by a little bit, your mpg calculation will be wrong.

 

Mine reads about .5 to 1 mpg lower than calculated by hand..

 

I disagree a bit. "With any precision" is a bit off. You've always got SOME precision! It just may be a rather large margin of error!

That said, if you were to use the SAME pump every time, you'll be more accurate. However, how much do you wager pumps shutting off automatically will differ? Surely not a full gallon. Let's say a half gallon. If one gets 15 MPG on the first tank of 20 gallons (for 300 miles on a tank), erring by a half gallon either way changes the MPG (again, assuming 300 miles on the tank) that the MPG changes only 0.4 MPG either direction. Or switching it the other way, say one gets 15 MPG again, that's only 7.5 miles different on a tank.

Either way, the calculation by hand would be averaged over time to something that is a bit more accurate than any one fill up. And if its consistently off by the same number, I'd assume that the computer was wrong.

And I've geeked enough about this now.

 

Mine's about 1 MPG optimistic. I've never reset it. I really didn't think I would use it, but it is actually useful for determining what type of driving kills your mileage. I'm expecting 19 on the current tank (computer is showing 20.0).

As stated above, hand-calc is quite accurate. If you did, in fact, fill more/less than usual on a particular fill-up...any discrepancy would be made up in subsequent fill-ups. One time in my old XJ, I filled up and got hand-calc of 21 MPG. I knew this was incorrect because I usually got around 19 MPG. Sure enough, next fill-up showed hand-calc of 17 MPG.

 

We didn't buy a Jeep for it's excellent fuel economy, did we?

If you want to be anal about it, do this:
1 - go to the gas station, fill up, and reset the mpg meter and the trip meter
2 - next fill-up, note the mileage and number of gallons, and reset the trip meter
3 - do the same for the next 5 (or so) fill-ups
4 - divide total miles by total gallons, and compare that to the mpg meter

That should give you a pretty reasonable indication of actual mpg to compare to the computer mpg.

Disclaimer: I won't guarantee this, as I haven't done it, and really don't give a rat's ass what my mileage is.