Hyundai and EPA may be inflating MPG figures

[MorphDaCivic]

Here's a good article on the discrepency between posted MPG figures and real life results from owners of the Elantra.

Consumer Watchdog said it has asked the EPA to investigate Hyundai’s high-mileage claims because, “The Elantra has attracted an unusual number of consumer complaints about real-world mpg averaging in the mid-20s, far from Hyundai’s stated average of 33,” it said.

UPDATE: Hyundai Responds to Consumer Watchdog's Elantra Fuel Economy Complaints - WOT on Motor Trend

[Old Tele man]

...GM needs to do the same thing (damage control) about the MPG differences for the automatic Cruzes between EPA and real-world customers!

...the "manual" tranny Cruze owners seem to beat their EPA numbers easily, while the "automatic" Cruze owners seldom even meet their EPA numbers.

[Gone]

Here is a recent analysis that I performed based on comparing EPA numbers to fuelly.com numbers for cars rated at about 40 mpg.



Hyundai way underperforms their EPA numbers; Honda seems to outperform.

[BHarvey]

I beat EPA (in 11' Sonata) before learning to drive more efficient.
EPA
City-24
Highway-35
Combined-28

First ten tanks averaged 31.2 or so, and the last ten is around 37.

Have not had a combined tank under 35 in a long time now, and I probably can do better.

[GRU]

Originally Posted by BHarvey

I beat EPA (in 11' Sonata) before learning to drive more efficient.
EPA
City-24
Highway-35
Combined-28

First ten tanks averaged 31.2 or so, and the last ten is around 37.

Have not had a combined tank under 35 in a long time now, and I probably can do better.

Do you do EOC or just coast with engine on and drive smarter? Beucase their test would have to be done as a average driver would drive.

[Rokeby]

I can't see the EPA being involved in some kind of conspiracy here.

However, I can see how a vehicle manufacturer could design a vehicle's
propulsion equipment/algorithms such that the vehicle achieves overly
optimistic, read not easily reproducible in the real world, MPGs on the EPA
test cycle.

I think that most folks either don't know or forget that the MPG numbers
coming out of the EPA testing are intended for direct comparison purposes.

The EPA numbers are an estimate based on an arbitrary test cycle. They are
not a guarantee of anything meaningful. Given the infinite different driving
conditions and driver capabilities/commitment, the EPA estimate does not
necessarily predict any particular driver's MPGs. With a large enough random
sample, drivers of vehicles with a 30 MPG EPA estimate will most likely see
greater MPGs than drivers of vehicles with a 20 MPG estimate.

If the EPA estimate is taken as an average estimate, half the drivers of a
particular car will exceed it, and the other half wont.

Which half would you expect to self-select to raise a ruckus?

[BHarvey]

E-on-C and drive smarter, but the car could do much better with some simple reprogramming based on what the UG tells me. 2 small changes would probably net 3-4 mpg better.

[Gone]

Originally Posted by Rokeby

... I think that most folks either don't know or forget that the MPG numbers coming out of the EPA testing are intended for direct comparison purposes.
...

But they are not comparable! If Honda, say, under reports MPGs and Hyundai, say, ultra-tunes their engine to achieve higher EPA MPGs then the comparison is between apples and oranges. People expect to use EPA numbers to compare cars but if automakers are gaming the EPA numbers, then there is no basis for comparison.

If you are interested in MPGs then the only relevant comparison is from real-world MPG numbers. Those numbers are hard to come by; but less so with websites 'crowd sourcing' MPG numbers.

[Ladogaboy]

Of the cars that are listed, the Hyundais do tend to be a bit "sportier," and I wouldn't be surprised if some of those low mileage numbers are caused by the owners' driving habits.

[tjts1]

Self reported BS. All this tells us is which driver's are the biggest liars. I'll take the EPA test numbers over a bunch of self reporting nitwits any day.