I am finding the car is about 2 mpg more optimistic than reality

I don't have any reason to think that SpeedTracker is a particularly accurate app. It's just the first one I came to reading down through the App Store listings that looked as though it was designed for use with cars and had an odometer feature. There were lots of bicycling and running apps, but I wasn't sure if they might have slower sampling rates and therefore be less accurate for a faster moving object like a car.

I would expect a properly done GPS measurement to be more accurate than a car odometer because the odometer is subject to variations in tires' circumference (especially due to wear for example) and squirming/slight slippage of the contact patch where it meets the road surface. GPS is accurate enough to guide aircraft around, after all.

I'm seeing a very similar difference. A trip Tuesday gave 168.2 on the odo and 170.1 on SpeedTracker, a difference of 1.13%. Part of the trip was on slightly snowy/icy roads, so the tires might have spun a bit here and there and so increased the odo reading slightly. A previous trip on dry pavement gave 71.5 on the odo vs 72.6 on the app, or 1.53%.

 

My odo also reads 1.5% low and the speed at 60 mph is 1 mph fast compared to a GPS.

 

Keep in mind that GPS is measuring straight line distance at some polling interval. I doubt that any are compensating for change in elevation which makes the road slightly longer than the straight line between polls. Also the earth isn't a sphere, so really calculating true distance changes slightly with where on the earth you are.

Odometer measurements on the other hand depend on circumference of a tire whose length changes with temperature, wear, air pressure (both internal and external), brand, size, etc as well as slight inaccuracies in the sensor or measuring methodology.

Also as noted above, gas pumps are checked (regularly?) to ensure the consumer isn't getting cheated. But the inspection process doesn't fail a pump if it is "within tolerance". It's no wonder that such deviations exist in calculating MPG.

 

Yes GPS' polling interval is a concern and especially on a curving road could make the measured distance shorter than the actual. Elevation changes are also a concern, especially in mountainous terrain. I'm less worried about the curvature of the earth: after all its circumference is roughly 25,000 miles and our wanderings are over such a small potion of it that the surface is effectively flat.

In any case, all these errors should make the GPS-measured distance SHORTER than actual. And what we are seeing is odometer readings that are EVEN SHORTER than the GPS results.

The only error that I can think of that might make the GPS readings longer than actual is jitter: if the first reading is "off" to the left of your actual track and the next is to off the right, the line between them is longer than the real distance between your actual positions. So you might end up with a zig zag GPS track that is longer than the actual path taken.

It's always a decision how much effort to make to get an accurate measurement of something. And when two or more values are combined (e.g., miles traveled and gallons used) there is no point in knowing one value very accurately if the other's accuracy is less certain. For MPG, least effort would be simply using the MFD readout. Next would be dividing odometer reading by gallons pumped. It seems to me that going the next step to GPS distance involves very little additional effort if one uses an app, and probably adds a bit of accuracy. Improving the accuracy of the gallons pumped number probably involves a lot of additional effort, at least with the methods readily available now.

That's where I'm at.

 

Opposite for my 2016 (Touring four). Showing longer GPS and shorter odometer on a long trip. GPS is pretty accurate if you make allowances for pulling off the road. I suspect the instrumentation on the Prius is pretty accurate and any MPG/fill up discrepancies are gallons dispensed errors from the pump and fill up environment conditions. Test sites use measured gallons for MPG. Very difficult to measure gallons from infinite environmental issues and pumps calibrations.

 

Actually what you are seeing is exactly the same as everybody else has reported: GPS>odometer reading. How large percentage-wise is the difference you see?

Measurement errors can be random or systematic. Random are equally likely to be off one way as the other and so will even out if you make enough measurements. Systematic are always (or preponderantly) off in one direction due to a problem with at least one of the measurement methods. A consistent difference such as we see suggests systematic error(s?) somewhere.

Since as you point out exact MPG data is difficult if not impossible to get, we don't know for sure which number (MFD number, odometer reading/gallons pumped, GPS/gallons pumped) is closest to the truth. So we have to choose which one seems likely to be most accurate and not too cumbersome to use.

 

That's not so bad. My gen 3 Liftback shows a consistent 7% over-report. I got used to it, but I still don't like it.

 

Pretty close to 5% for me. And consistent.

 

Have you compensated for the under-reading odometer? dkraus and I both find 1-1.5% error there, which would reduce the discrepancies you find by that %age if you have not factored that in.

I hate MFD error too, and hated it in the Gen 3. I would constantly have to remind myself that the numbers I saw on the MFD were optimistic and adjust them via mental calculation, depressing! I don't mind the bone china but I DO mind that.

 

My assumption is that any odometer error or gas measurement error would weigh equally on each MPG calculation. And I think of the numbers as "entertainment".

 

You can change tires to adjust the calibration. Just choose tires who revs/mile compensate for the error. It is not that hard.

Bob Wilson

 

Yes, exciting day at the movies. I concur.

After my new car check on a very long trip with a GPS cross check, I am done. Dash info is close enough for me.

I think the quality of gas is doing stuff. I noticed with the same driving pattern different mpg results with different gas stations.
Also I want a longer life reality more than optimistic hopes.

 

OK, then why bother with the spreadsheet and calculations? Why not just use the MFD numbers and save yourself some effort?

I've mostly stopped reading posts in this section for exactly this reason: I never know how the numbers were arrived at so I have no idea how meaningful they are.

 

Bad idea. You'd need one size tire to correct for the odometer error, another to correct for the mpg error, and a third to correct for the speedometer error.

 

As Mr Rogers and Billy Joel said, "I like you just the way you are". This Prius is the cats meow.

The title of the thread sounds like a quote from the Hitch Hikers Guide to the Universe.

My improbability drive is working fine.

 

Perhaps but it worked with the Gen-3. If the BMW i3-REx deal falls through and another does not show up before the Level 2 ECO, I'll be able to run the experiment. But let's assume you are correct:

1. Use a GPS for speedometer and trip distances
   
2. Adjust the tire diameter for true MPG
3. The odometer reading is less useful

Bob Wilson