I'm just going to toss this out there. A lot of people think the MPG's on a vehicle computer is wrong and that hand calculating your MPG's is best. I contend hand calc-ing your MPG's is not correct, statistically speaking. When you hand calc your MPG, you are using one data point... miles you drove over gallons used. Simple and is certainly an average of something.

The computer in a vehicle is calculating MPG's over hundreds or more likely thousands of data points. Some (most?) probably do it constantly while running. The "lie-o-meter" will seem like it's giving you better MPG's because it's taking into account data points when you're actually getting "better" MPG's so those readings will pull the actual average up with its larger sample data size. The more data points the more accurate the actual average is. Basic statistical math.

Every physics lab I ever had, you wanted as many data points you could get so when you ran a linear regression line, it more closely represented the trend. 2 points can give you a trend, 3 points gets you a curve, but 100 points of data will start to eliminate errors and will more closely represent whats happening.

I'm guessing most vehicle ECM's calculate fuel flow by volume using some sort of algorithm. In the boating world, most modern outboards (and inboards) do this and with the proper NEMA 2000 connections, you can get to that info and display it on a sonar screen display. Most mid to high end displays can calculate fuel flow (GPH), gallons burned, and MPG's. In my experience, these flow rates are dead on. Computer says I burned 83.6gal of fuel and it takes 83.6gal to fill it. Over the years, I've learned to trust my instruments and don't see how a car or truck is any different than a boat. It does have to be calibrated so if you've changed tire sizes and not adjusted the ECM to reflect that, will then it won't be giving you correct numbers. I've paced my speedo and it seems to be right on based on GPS and freeway speedo check markers.

Anyway, just something to ponder...

 

..
its the DRIVER...not the truck...

ask any real truck driver..

better mpg is the driver..... the truck just does what it does.

by the way.. my "XL" truck meter... is within 0.2 mpg every time. 166,997 miles. as of this morning.

 

River Wild -
I do statistical process quality control for a living; stats are my thing.

While I go agree that more data is typically better, there are things we have to consider in your assertion that may or may not be true.

While it's good that the "lie-o-meter" takes many samples, it's also basing those calculations on a calibration factor. Whereas the variation of many thousands of points of data will give a very crisp sigma (standard deviation), that does not mean that the equipment used to take the readings is necessarily accurate. I don't know how the fuel consumption is measured in our SD trucks. Is it based on a computer generated fueling rate that commands the injection events? Loading factors? Does it include readings from the fuel level sending unit? Does it account for fuel density based on atmospheric pressure delta P? Etc etc ..... The point I make is that the accuracy of the value is a separate topic from that of variation. One system could be off by 2mpg, but be very consistent, whereas another could be spot-on, but have more variation. Just because the vehicle computer takes many, many readings as data points, does not mean the output is accurate; it only means it's more refined. It can be very refined, and still wrong.

Bottom line is that using the pump reading in gallons is an effective way to get an "average". While it's only one reading from the pump per fill-up, you can do this repeatedly and start to average out small variations, especially if you return to the same pump most of the time. If you took 30 pump readings from 30 consecutive fill-ups, and ran the AVG and STDEV values, you'd get a really fair, reasonably accurate idea of your mpg's. After all, if you drive 300 miles on one fill-up, and do that 30 times, you've averaged your data over 9000 miles of use!

And, my "lie-o-meter" is actually quite accurate; typically spot-on or only 1/10th mpg shy of my hand calculated values.


BTW - sample sets of less than 30 are woefully inaccurate for predicting standard deviation. 30 is the absolute minimum sample. 50 samples is better. And, while more is typically better, you don't need thousands to be confident in your sigma value. Once you approach 100 samples, there is precious little change in the confidence interval for any assume value. Sample sets between 30 and 100 are best. Too few gives a very poor look at deviation. More than that and you're just wasting time/money because you're not refining the data in a meaningful manner relative to the cost inputs.

 

 

.
remember that the mpg meter.. is a small and cheap addition.. for an electronic package. My Opinion.

not calibrated.. or posted to WHAT it has been calibrated to.. ?? !!!

in the 90's I did work in the electronics field to the NIST standard.... to zero point zero 1 percent., a/d and d/a converters..

 

If I'm doing some in town at 10mpg and some highway at 15mpg, the average on the "lie-o-meter" is different than hand calc'd. MPG's on the lie-o-meter can be off by 1 mpg, usually 1 mpg better. If I do all in town on the entire tank or do all highway on a tank, it's within .1 mpg. I use fuelly so I have hand calc'd data for every fill up mostly for entertainment and just to monitor what's going on. What many call the lie-o-meter is actually quite correct imo. I think too many people like to think there is some sort of conspiracy by manufacturers.

The mpg meter is nothing more than math, not an add on. Each injector delivers a given flow over time. Tires go around at a certain speed/distance (provided the tire size is input correctly). Simple math. The ECM knows all of this, tracks it, and does the math.

 

Yes, flow can be measured (GPH) and gallons used / MPG can be calculated from it but......

Is the flow calculated by tracking injector pulse frequency and duration along with fuel rail pressure? Or measured by the fuel flowing to the rail from the pump (which is inaccurate because every vehicle I've ever seen has a fuel return)?

Also take into account that every vehicle has some version of DFCO and SEM.

It's only my opinion and worth nothing but the latter is what I believe makes the MPG calculated by the vehicle incorrect (assuming distance traveled is accurately captured, blah, blah). Nothing can divide by zero so the gallons used calculation uses some obscurely small number. Also starting a cold engine fuels more to light the converter, shifting and tq management change fueling/timing strategies at each shift so the ride is smooth to you, keeping the converter lit during long decel periods uses its own fueling strategy. Lots of variable driving, lots of variable fueling = MPG displayed that isn't as "accurate" as hand calculating it because the gallons used aren't calculated accurately in all situations.

I just don't think it's possible to calculate MPG 100% accurately and display it on your dashboard taking all these scenarios into account. Modern vehicles with modern transmissions, especially with 6,9 and now 10 speed transmissions are operated under such varying conditions compared to say, a boat engine with one gear that runs at a relatively constant speed all the time where those parameters can be measured more accurately.

 

The so called "lie-o-meter" and hand calculations are only as good as the speedometer calibration in your truck. This is for both long and short travels.

Example: My old F150 started it's life with 265/70/17 (31.6") tires and the MPG meter was off compared to my hand calculations. I didn't think much of it. Then, I put 285/70/17 (32.7") tires on. While this wasn't a drastic change, it did change the speedometer to be off enough for me to want to calibrate it. Using a hypertech speedometer calibrator and a GPS, I got the speedometer dialed in exactly 60mph running at 60mph. And, in doing so, the "lie-o-meter" began to be more correct than it ever was.

To further prove this, the trip meter was extremely accurate as well. The guage showing the fuel usage to the nearest tenth was just that at the fuel pump too.

Jumping to my newer F250, the "lie-o-meter" is off by about .4 on a tank of fuel. But, my trip meter fuel usage is also off by about half a gallon. Testing with the GPS showed that my speedometer reads a bit fast. This also effects the odometer that is crucial to correct calculating.

I say, BLAME FORD!!!

 

If you take the computers miles driven by what the computer also says is the gallons used you get what the lie-o-meter says is your average mpg- at least on my mine. In my hand calcs I also use what the computer says is the miles driven but the gallons per fill up is determined by a “strict” scientific process😂. I go to the second click and to the nearest nickel.... usually add 3-4% more than truck says....🤔.But we are assuming the gas pump is accurate or at least more so than the truck itself which could be a point of contention lol. Definitely food for thought....

 

 

 

Press and hold the ok button on the wheel while turning the truck on. This will get you to the engineering mode, go to AFE bias to fine tune. Someone here has posted the formula.

 

https://www.ford-trucks.com/forums/1...l#post18092841

 

 

I normally don't care too much about MPG because I know I am driving a heavy brick. I use to get 13.2 or so with mostly highway miles. I switched to a bigger tire and my MPGs went down as expected. I am now at 12.6 or so. I read a bit in here about the "Lie-o-Meter" so I thought I would do a hand calculation and see how far it is off. I got 10.5 instead of the 12.6. I thought that is way off. I reset the Trip A and tried again and got the same results. But the Trip A has how many gallons you used. Some how I put more gallons in per the pump than what the truck said I used. I thought that is strange to be off by 2 gallons or so. I think the truck is setup to hold so much gas. If you don't pump any more than after the pump shuts off, I think the Lie-O-Meter is correct. If you pump more in to round up to the nearest dollar, it maybe off.