Toyota why is the mpg readout not accurate?????

I'm sure that IS what he is saying but there is at least one flaw in that theory:

It is a known fact that, on most cars, the front tires wear differently than the rear because the "setup" is different; the alignment.
That difference is often exaggerated on a FWD car. It is natural and normal.

So.....if one follows the non-rotation method of tire "maintenance" you should expect to replace the fronts more often than the rears; maybe as much as twice as often. That leaves you open to the problem of mis-matched sets.

I expect my shop to inspect the tires when they are rotated AND to be able to tell the difference between normal and abnormal wear.
To that end, I inspect them myself quickly before and after a rotation.

 

He acknowledged as much in a previous reply, when he placed himself in the 'european' camp on this issue. While that is not my camp -- I prefer closely matched tread on all four corners -- it is a very big camp.

 

I would contend that alignment issues are evident in the 5,000 miles between rotations.

 

BobWilson is not your average car owner. He does his own 4 wheel alignment.

See the difference ? We obtain even wear through rotation so that uneven wear on different parts of the tyre catch up to each other; Bob insures even wear for each wheel. Nothing wrong with his approach in terms of wear, and I think he prefers 2x frequent replacement of his front tyres (and 1/2 frequent replacement of the back tyres) so that he can play around with different tyres.

The downside is the fiddling with alignment. You better be a BobWilson, or have a good mechanic friend happy to do the work for free.

Another disadvantage that is probably climate specific is sidewall cracking. My tyres end up needing replacement as much for cracking as wear, so I suspect if I tried the 'European' method I would end up buying more tyres over time.

 

Yep, especially is (a) hot desert areas and/or (b) high-ozone areas. In only 4-5 years, sidewall cracking becomes a major problem.

 

And there is another thread suggesting money can be saved by not rotating tires at all and changing only the two that wear out. The example given was 30k on OEM fronts, 54k on NEW fronts and 84k on OEM rears. I think it was just a coincidence that the rears lasted as long as the 2 sets of fronts and I suspect most would have to replace fronts, then replace a set of rears, then fronts again, and so on. I don't know the safety aspect of having a front set almost worn out while the rears are still good and then vice versa as one goes through the replacement cycle. I think Bob does the same thing, but I think he does it for different reasons.

And that is a very good point. Last year I replaced a set of 80,000 tires at the 86,000 mile mark and the tech at Discount said I was pushing the age (8 years) in spite of the tread lasting. However, he did acknowledge that it was okay because the car was garaged and not used for work where it would have been parked out in the heat. I'll probably still buy high mile tires in the future, but I don't know if they visually check for age related wear. I did have a set on my truck that gave out before the tread because it's parked outside in the heat. The sidewall developed hundreds of tiny holes and luckily gave out in the driveway.

 

I have been a fan of high mileage tyres for years but recently changed my mind because of the cracking problem in New Mexico. Welcome to the dry high desert!

I should mention that Costco/Michelin honored the tyre warranty by declaring the cracking a 'defect' and using residual tread-wear to decide how much money to return in credit. They also did not charge for mounting. I would not presume that to be true for other retailers or manufacturers.

 

To answer the OP's first question of why it is not accurate, the following issues affect the calculated MPG:

1.) Tire circumference changes with tire pressure or brand of tire or altitude,e.g. under inflated tires rotate more for the same given distance.

2.) fuel expands and contracts with temperature or altitude, this effects the measured fuel usage.

3.) type of fuel, alcohol mixes or different octane ratings or seasonal fuel blends all have different energy per quart/liter. This effects the measured fuel.

4.) accuracy of electronics, even advanced electronics have an error. This even changes from batch to batch of electronic components.

All these thing stack up to make it nigh impossible to design an accurate MPG display across the whole Prius manufacturing range, so like all things in life it is a compromise and as an design engineer, you chose your margin of error. In this case Toyota chose to err on a 5-8% optimistic reading. This is to their advantage as it gives most drivers a warm fussy Prius feeling I guess.

To get the most accurate(but this still will have a margin of error):
1.) Use a GPS(has an error too) to measure the margin of error of your car's odometer over a large distance like 300miles/500kms.

2.) fill up your tank at pump (has an margin of error too, you are not getting exactly what the pump reads)

3.) calculate the MPG manually and compare the calculated MPG against your calculations.

4.) now over a number (e.g. 10) of tanks calculate the average MPG ( Fuelly - Track and Compare your MPG does a good job) and you will get a good feel of how optimistic your car's MPG display is.

However this will change as your tires wear or when you change tire pressure or change tires or change fuel or seasonal fuel blends.

 

Three tanks is more than enough* if you use the same fueling pump and do not try to stuff fuel in past the first 'click' off.

Regarding Fuelly, does the software weight different fill-up distances ? I suggest instead simply dividing miles driven over the tanks being considered by the total volume of fuel bought (not including the amount to reach 'full' in the beginning.)

*I say this based on the following: Let's say the deviation from 'full' is up to 100 ml. Three tanks use ~ 100 liters, so the error is 1/1000

 

To address kiwiscoot's first list:

1. Doesn't matter. Any distance-travelled-error is the same for you and the car's computer.

2. My 2 brain cells are struggling with this, in exasperation I'm going to dub it irrelevant. There!

3. See 2

4. I like the bit about warm, fuzzy feeling. The rest not so much.

And to second list:

1. Irrelevant, see response to 1 above.

2. Yup, we're going to do that. Note, any variation in pumping practice is going to cancel out with the next fillup.

3. Uhuh, that's another given.

4. Fuelly, spreadsheet, back of a napkin. Anyway you want, you'll get a feel, not necessarily good, about the very consistant bias Toyota's foisted on us.

 

Fuelly calculates a number of things, e.g. like a running average. You can even compare it to other cars or the average of all the same cars on fuelly, which for my 2011 Prius is of 359 cars which have traveled more than 6.5million miles.

Info on fuelly About Fuelly

 

Been here for 20 years, but this was my first set of high-mileage Michelins on my wife's van. I forget what the truck tires were, but they weren't high mileage and they were 8 years old too, so I was probably pretty lucky.

 

My question was whether the average is harmonic.

 

Taking your first point 2.) in consideration, I accept your whole reply and view.

To the original poster:
The bottom line is as an design engineer you cannot design a system that is totally accurate without being calibrated periodically to account for changes in the system. No owner is going to buy into a system where you have to take your car in to be calibrated every service with the associated costs. Seems Toyota is damned if they do or if they don't. "Over read" offends some owners and "under read" would offend some others. IMHO Toyota covered their bases very well, a very well designed system to give such consistent performance. Simple matter to gather the info on your average parameters for your car and location and then calculating a good estimate of your situations margin of error.

What do you mean by harmonic? Do you mean seasonal averages?

If that is what you mean, no. It is not that advanced.

 

I agree with Mendel -- the system is quite precise with a 5% over-estimate. Toyota could have included a correction factor.
I personally do not care though, since it is easy enough to handle.

Harmonic Mean

I use the term 'unweighted average' to refer to the average of a series of simplified ratios (e.g. MPG) that originally do not have the same denominators.

Example:
50 mpg from point A -> B
100 mpg from point B -> A
What is the MPG for the trip ? The answer is 66.6 and not 75 MPG because the return trip used only half the fuel as the forward segment. I express that difference as the forward MPG having twice the weight as the return segment. So while the simple average is (50+100)/2, the weighted average is (50+50+100)/3

If Fuelly is calculating a simple average of the tank MPGs a similar issue arises because the tank refills are not equal.

 

Has he ever actually said that he purposely deviates from the standard alignment by much ?
If so, I think that COULD BE a foolish thing to do.......and certainly does not justify advocating that approach to your average Joe owner.

The tire alignment specs are there as much for handling as they are for wear.
Deviating from that by much might make the car handle unpredictably at times.

 

So far as I know, BW has not advocated 'European' alignment to anybody. He posts what he does on his car.

Would you like to censor his posts ? (Sorry Bob -- I couldn't resist.)

 

Nobody's asking for Toyota to tighten up their accuracy. As it is (see snapshot I posted from my spreadsheet) the error varies by at most +/- 1%. That's fine. No need to throw money at that.

But the numbers are pushed off to the plus side, by a head-shaking 7+% (in my experience). That's the subject of the thread title. It's an embarrassment that Toyota persists in this.

 

A consistent error is called bias...

A random error is called tolerance (wink,wink).

 

Yes, although precision and accuracy are helpful terms here. The MPG meter is precise but not accurate. This is why we can talk about a bias, and a fudge factor to correct it.