Saw a news clip last night that Consumer Reports was checking up on EPA mileage ratings. They found that at least half the cars tested differed significantly in reality compared to their rated number. Particularly bad were some of those supposedly super-efficient hybrids, whose actual numbers didn't look much different than my gas hogging SUV.

Might be useful if reviewers considered keeping a mileage log for a while on their cars and posted some actual MPG numbers....I know I'm usually pretty lazy about it and just take the gov't numbers at face value.

Have fun!
Mark

I actually stumbled across fueleconomy.gov recently, where the federal government is asking people to report the fuel economy they have experienced. They're not doing a good job of getting the word out--each model I looked at had only a handful of reported numbers.

Now for my CR tirade...

The Consumer Reports article is a bit cracked. They figure that since Congress won't raise fuel economy standards, then they should get the EPA to develop a more realistic test so cars would have more trouble hitting the existing standards. In all fairness, the current standards were agreed to knowing that they were not representative of the real world.

Kind of like tax assessments on houses. These tend to run at about half the true value of a house in places where I've lived. Imagine if local governments decided to increase property tax revenues by reappraising houses at their true market values. They'd be able to nearly double their revenue overnight. Only there'd be a bloody rebellion first.

I've only recently taken a close look at CR's own practices, to better explain why my own reliability research is necessary. The more I look, the clearer it is that they should take a look in the mirror once in a while. They talk about EPA numbers as deceptive, but don't seem to realize how deceptive their highly influential ratings are. In most cases not intentionally, but because of how people predictably tend to read them.

For example, by constantly talking about percentage differences from the average and problems per 100 cars, they deflect attention from the fact that the absolute differences are rarely as large as a problem per car per year, and are often just a few tenths of a problem per car per year. Since people are only buying one car, and not a hundred, the actual impact on people's lives of buying one versus another are often insignificant.

I've written a slightly lengthier critique on my site at www.truedelta.com/pieces/shortcomings.php. But the basic point is that CR's practices are seriously flawed, and as a result purchase patterns have been distorted. This makes them quite hypocritical when criticizing the EPA numbers. The difference is, most people know that the EPA numbers are not accurate.

I agree with you regarding Consumer Reports reliability numbers. The magazine has been so high on Honda for so long that it has blinded people to problems their cars have. I once had a co-worker brag about how reliable her Honda was, when I pointed out that she recently had it towed off the parkway because the alternator died. Her response was, "Yes, but it's been great other than that." My wife's boss won't buy another Honda after repeated and very expensive brake repairs. My brother had a Honda that wouldn't start in cold, wet weather. (I have another brother with a Honda Pilot that's been reliable, although he complains about a few little things.)

Meanwhile, my 1996 Mazda pickup (actually a re-badged Ford Ranger) has had exactly one shop-worthy problem in 4.5 years, and that cost me $230. Other than oil changes and similar maintenance, I haven't done a thing to the truck. My 1993 Porsche 968 averages one shop repair a year, which I do myself. None has stranded the car and most could be considered normal wear-and-tear, such as brakes. My 1995 Subaru Legacy recently had a differential failure that caused me to replace the transmission. If I had taken this to a shop, it would have cost $2,000 or so. I did it myself and it cost $550, including the cost of a new clutch... which it turned out the car didn't need. The only other notable problem the Subaru has had in seven years of ownership was a torn CV joint boot. I have replaced the brakes all around, as the car now has 145k miles.

I think that the reliability ratings in Consumer Reports tend to perpetuate the ratings. Cars that rate well continue to rate well because the cars' owners are blind to the maintenance that they're doing. It's the old thing about hearing what you want to hear.

Just my little rant, I guess.

--Bob

"...Uneducated?" Diplomatically put, Rex.

I wish CR would simply list the raw problems-per-year figures for each car, along with a standard deviation stat to expose vehicles that are hit-and-miss propositions (such as the recent Jetta/Golf/New Beetle). IMHO, rating reliability through a simplified code of dots is an underestimation of readers' intelligence, especially when those dots are based on a largely irrelevant "percent off average" figure. Maybe if their "average" consisted of the last ten or twenty model years, it'd be more representative...

That said, I do admire CR's automotive testing. They may be Honda stalkers, but you know they're not handing out praise for cars because of their styling (or because they have deep-pocketed advertisers). I don't always agree with their evaluations, but CR and ConsumerGuide are the only automotive publications I've found that at least try for journalistic objectivity.

As for the EPA figures, chasing a more-accurate MPG calculation seems like a waste of resources to me. Whatever a car's EPA numbers are, they go out the window as soon as an individual gets behind the wheel. My 2001 Eclipse was rated at 20/29, but depending on the circumstances, I got anywhere from 16 MPG (in frenzied city commuting) to 35 MPG (on 55 mph highways). The real value of the EPA figures--again, IMHO--is that you have a consistent test that you can compare vehicles' scores in.

Incidentally, my fiancee's Prius averages about 45 MPG, well off the EPA's 60/51 figure. But living in an area with not-so-defensive drivers and lots of full-throttle freeway merging, we knew not to expect those numbers going into the purchase. Besides, even if the Prius' real-world gas mileage isn't all it's cracked up to be, there's no question that it's more efficient than most new cars--and certainly any SUV.

None of my Honda-driving friends have had car trouble to date (except for my college roommate's '89 Acura Legend, which had over 300,000 miles on it). But I gave up on anecdotal reliability info after owning my first car--an '84 Chevy J-body that never so much as hiccupped until the rust rendered it structurally unsafe.

- P.J.

PJ, the numbers you want are coming. My panel just passed the 1500 vehicle mark. The plan is to get a good sample on a few vehicles, get some initial results, then get some press coverage. From what I've seen so far, a single article that reaches a million people would push my sample to a very healthy size. Assuming the servers at my host hold up--I'll have to see what I can do to get the best they've got for a few days.

Any update? I track my MPG online, www.ecarlog.com, its easy all the calculations are done for you. Just enter the number of gallons and price paid at each fill up. Oh yea you have to fill it up and no topping off. You also get a decent maintenance log and email reminders if you choose.

Prius get about 42-45 mpg not 60\51mpg (that;s more in the city). Kia gets 18mpg not 22/30mpg (very upset)
Seems toyota's Corrolla is honest per its a fuel minded car at 36mpg allround so it probably really does get 41 mpg highway.