Prev: write your congress-critter - increased ethanol in gasoline
Next: Door guard to prevent dings/dents on Honda?
From: dr_jeff on 20 Jun 2010 12:54
Scott Dorsey wrote:
> Ashton Crusher <demi(a)moore.net> wrote:
>> It's really incredibly stupid. Toyota went from 1.08 last year to
>> 1.17 this year in terms of problems PER CAR. No retail car customer
>> in their right mind owning one or two new Toyotas (or any other brand)
>> would consider a change of 0.09 problems in the first 90 days
>> something meaningful in terms of making a buying decision. There is
>> so little difference between most brands that the whole thing has
>> become ludicrous.
>
> Well, the thing is, NONE of the cars today have enough initial defects to
> be considered alarming... all of them are far better than anything made
> a couple decades ago.
>
> So, really, using initial defects to compare vehicles is useless because
> the defect levels are all in the noise floor.
>
> The problem is, though, that what I want to know is how reliable a car
> will be after I've been driving it for twenty years, and measuring that
> without first driving the car for twenty years isn't possible. So there
> really isn't a good metric for what I most want to know, and the metric
> that _is_ available is effectively useless. It's a sad state of affairs.
> --scott

One problem with the JD Powers and Assoc. survey is that it combines all
problems. So, if there a problem with a loose screw in the dashboard or
the engine caught fire and the car is trashed, the problems are weighted
equally. Another problem with the survey is that these are owner
reported problems. So the owners of a particular brand or model of car
might be less likely to report a problem than owners of another brand or
model. Finally, this is only during the first 90 days of ownership. So
if a car's brakes completely fail, resulting in a wreck that destroys
the car is not reported at all.

Jeff
From: dr_jeff on 20 Jun 2010 13:34
Scott Dorsey wrote:
> Ashton Crusher <demi(a)moore.net> wrote:
>> It's really incredibly stupid. Toyota went from 1.08 last year to
>> 1.17 this year in terms of problems PER CAR. No retail car customer
>> in their right mind owning one or two new Toyotas (or any other brand)
>> would consider a change of 0.09 problems in the first 90 days
>> something meaningful in terms of making a buying decision. There is
>> so little difference between most brands that the whole thing has
>> become ludicrous.
>
> Well, the thing is, NONE of the cars today have enough initial defects to
> be considered alarming... all of them are far better than anything made
> a couple decades ago.

The natural assumption is that a car with fewer initial defects will
also have fewer problems down the road (both figuratively and literally)
and is of higher quality.

> So, really, using initial defects to compare vehicles is useless because
> the defect levels are all in the noise floor.

Nine problems per 100 vehicles (the difference between 2009 and 2010
scores for Toyota) might not be below the noise floor. There are
statistical measurements of that, like standard deviation. However, one
other thing that needs to be remembered is that over the last couple of
years, the costumer base for all automakers changed faster than normal.
Toyota's changes were brought by both a changing mix of vehicles
(particularly the Prius) and reports of people hitting the gas pedal
instead of the brake pedal or the gas pedal getting stuck under carpets.
GM's and Ford's sales are changing because they are closing brands. In
addition, the overall mix of vehicles sold has changed due to changes in
the gas prices (particularly two years ago), the clunker program (what a
stupid idea that was), the economy in general, and GM's and Chrysler's
financial problems.


> The problem is, though, that what I want to know is how reliable a car
> will be after I've been driving it for twenty years, and measuring that
> without first driving the car for twenty years isn't possible. So there
> really isn't a good metric for what I most want to know, and the metric
> that _is_ available is effectively useless. It's a sad state of affairs.
> --scott

I don't think the metric are totally useless. Ford is building a pretty
good track record of reliable cars, just like Honda and Toyota have
(and, IMHO, earned). I have had two excellent Ford cars and am thinking
about trading in my 2008 Ford Focus for a 2010 model, if I can do it
cheaply enough (otherwise, I will wait for the European model in '11 or
'12).

If the car makers built the exact same cars for many years (the Model T,
VW Bug and Peugeot 504 were all built for over 20 years with basically
the same design and model), then you can get an good idea of how
reliable a new one would be. But, you would never get any new
technology, like improved crash-worthiness, better emissions controls, etc.

So there is always a trade-off between ignorance about projecting repair
rates and advancing technology. If I am not mistaken, small aircraft
engines still use magneto technology because the aircraft companies are
afraid of the liability if the engines stop working mid-air and because
the magneto engines, when maintained properly, are very reliable.

So the aircraft companies have chosen to maintain proven dependability.

Jeff
From: dr_jeff on 21 Jun 2010 06:18
Derek Gee wrote:
> "dr_jeff" <utz(a)msu.edu> wrote in message
> news:BqydnfsimcxP2oPRnZ2dnUVZ_tednZ2d(a)giganews.com...
>> One problem with the JD Powers and Assoc. survey is that it combines all
>> problems. So, if there a problem with a loose screw in the dashboard or
>> the engine caught fire and the car is trashed, the problems are weighted
>> equally. Another problem with the survey is that these are owner reported
>> problems. So the owners of a particular brand or model of car might be
>> less likely to report a problem than owners of another brand or model.
>> Finally, this is only during the first 90 days of ownership. So if a car's
>> brakes completely fail, resulting in a wreck that destroys the car is not
>> reported at all.
>
> What do you mean by the last sentence? If the brakes fail within the first
> 90 days, it's going to be reported. Why would think it wouldn't?

If the crash occurs on day 91, it is not reported.

> Problems outside of the IQS 90 days, would be covered by the JD Power
> Vehicle Dependability Study which would be three years from the current
> model year. (e.g. the 2010 survey covered problems with the 2007 model year
> vehicles)
>
> Derek
>
>
From: Scott Dorsey on 21 Jun 2010 09:28
dr_jeff <utz(a)msu.edu> wrote:
>
>The natural assumption is that a car with fewer initial defects will
>also have fewer problems down the road (both figuratively and literally)
>and is of higher quality.

The thing is, this has repeatedly turned out not to be the case, especially
when the initial defect levels are all so low to begin with. Perhaps it is
better than nothing, but not very much so.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
From: jim beam on 21 Jun 2010 10:10
On 06/20/2010 10:34 AM, dr_jeff wrote:
> Scott Dorsey wrote:
>> Ashton Crusher <demi(a)moore.net> wrote:
>>> It's really incredibly stupid. Toyota went from 1.08 last year to
>>> 1.17 this year in terms of problems PER CAR. No retail car customer
>>> in their right mind owning one or two new Toyotas (or any other brand)
>>> would consider a change of 0.09 problems in the first 90 days
>>> something meaningful in terms of making a buying decision. There is
>>> so little difference between most brands that the whole thing has
>>> become ludicrous.
>>
>> Well, the thing is, NONE of the cars today have enough initial defects to
>> be considered alarming... all of them are far better than anything
>> made a couple decades ago.
>
> The natural assumption is that a car with fewer initial defects will
> also have fewer problems down the road (both figuratively and literally)
> and is of higher quality.

<snip for brevity>

the above is a false assumption. a really dumbed down example is this:
have you ever encountered the exploding chinese capacitor problem in
electronics? to the consumer, there is no difference in initial
quality. but a few months down the road, when they literally go "BANG",
you'll discover why the initial quality, which was prima facie perfectly
adequate, was in fact no indicator of subsequent performance.

same for cars, "initial build quality" makes no differentiation between
one vehicle with cheapo chinese bearings [for example] and another with
quality american. but you sure are going to know it a few thousand
miles down the road when one needs replacing and the other will keep
going for another few hundred thousand miles.


--
nomina rutrum rutrum
 |  Next  |  Last
Pages: 1 2
Prev: write your congress-critter - increased ethanol in gasoline
Next: Door guard to prevent dings/dents on Honda?