“I Re-Read The Article, But Can’t Find Why A Kia Costs 60% More To Maintain Than A Toyota. What Are They Getting For Kia Oil Filters?”
The explanation might lie somewhere within the reality that the person who compiled these stats can’t seem to distinguish between maintenance and repairs, which are two totally different concepts. But…who knows?..when something is as fatally flawed as this “study” apparently is.
The Kia and Mini positions are a little surprising to me. I’d have guessed Kia would be a little further down the list, and Mini a little further up. Two of the positions, MB and Lexus, if my neighbor is any indication, those are correct. My neighbor traded in his 5 year old MB and bought a Lexus replacement expressly to improve reliability & reduce maintenance costs. He’s had the Lexus for a couple years now, and seems very happy w/it.
You can make very very accurate predictions by looking at the overall maintenance costs of a manufacturer. Patterns emerge and you can deduce a lot from the overall patterns. Manufacturers with overall higher maintenance costs can indicate an overall company culture pattern. True there are individual cars in a manufacturer that may have extremely low maintenance costs but overall the manufacturers costs are high. What does that say about the rest of those cars. Their maintenance costs must be EXTREMELY high to average out so high.
That’s the thing about science and statistics. Nothing is 100% certain. If you want certainty, there’s death and taxes.
I don’t get everyone’s aversion to processing large amounts of data and presenting the findings.
No data is perfect. But what went into that article goes way beyond any of our scope of data.
For those of us who currently work with car repairs, our dataset might be in the 100-200 vehicle range. For those of us who who work on friend and relative vehicles, the data set is in the 10-20 range.
The article doesn’t quantify what the numbers are for their "massive data set " description.
While any data should be looked at with a critical eye, shouldn’t we be looking at this to see what can be learned rather than outright rejecting it?
Say you tossed a coin 100 times, and each time it turned up “heads”. %% There’s nothing about observing that sequence of events that would allow you to deduce with 100% certainty what would happen on the next toss. That’s just one of the limitations we deal with all the time. The options are
%% I borrowed the coin toss analogy from Tom Stoppard’s play Arcadia Rosencrantz and Guildenstern Are Dead concerning the relationship between past and present, order and disorder, certainty and uncertainty.
I work with probability all the time, and I’m comfortable with it. I know the difference between anecdotal evidence and a rigorously derived probability that an event might take place with the associated variability of the prediction. I doubt that the authors would present any of their data in that way, and it would be lost on a lot of their readers.
I object to the mixing maintenance and repairs as if they are the same thing, and applying what appears to be the same maintenance schedule to all autos. If we guess that vehicles are driven between 12,000 to 15,000 miles per year based on changing oil twice per year, then the authors might be double counting oil changes for the European brands on the extended change schedule. At $150 per oil change at your local Bimmer dealer, that’s an overestimate of $1500 over 10 years. The authors certainly leave this open as a possibility. The way they present themselves doesn’t give me confidence that they know what they are talking about. If I were King of the Forest, I would obtain the real maintenance schedules and cost them based on prevailing materials/parts costs and a reasonable estimate of the time to accomplish those tasks. As King, I could hire a queen or duke or earl with a good knowledge of the auto repair business to derive the parts and labor costs. They may have done that, but it isn’t clear from their presentation.
You’re saying BMW recommends longer oil change intervals, and if the author’s didn’t take that into account, their estimate for the BMW services might be too high? If that indeed happened, that seems a fair criticism.
It looks like they add repair and maintenance costs to get their rankings - that’s good. I don’t care as much about regular maintenance costs as I do unexpected repair costs. The rankings are more useful if they include both.
Don’t some manufacturers – if I recall this included BMW – cover the cost of scheduled maintenance as part of the purchase price? For a certain number of years anyway. I wonder if the study took that into account.
I seriously doubt the percentage of vehicle buyers that ever even think about a chart such as this being over 10%
BMW provides free maintenance for the first 3 years or 36,000 miles. That’s one or two oil changes for a new Bimmer. Based on that very high cost, I doubt they included it.
I agree that providing both repairs and maintenance costs are useful. I prefer to see them separately as Edmunds does. It seems to me that maintenance costs are easier to predict than repair costs since they are already scheduled when the car is purchased.
I would like to see how any organization, not just this one, predicts repair costs. The way Edmunds does it, the extended warranty is divided into profit, sales costs, and finally the actual cost of repairs. The repair cost estimate would be the average cost of repairs amortized over all vehicles of that model. Clearly, you can’t do partial repairs, and this method would cost it that way. I am not sure how else to do though. If I ran the repair insurance business, I would hire actuaries to estimate the average cost of repairs based on actual repairs for that model. I wonder how the company discussed here does it.
Surprised volkswagon came in so reputable.
“Surprised volkswagon came in so reputable.”
I’m not surprised by the Volkswagen ranking. Nothing about the findings drawn from the “massive dataset” would surprise me. :neutral:
CSA
Just my 2 cents, but I’m of the opinion that the only way a reliability study can be done is if a thorough survey of all owners and total research of all complaints made to the dealer and verified by repair orders is done.
That would be godawful time intensive and very, very expensive to accomplish.
What should be done is to compile complaints say over the first 30k miles of ownership and weigh those; excluding maintenance costs of course.
Complaints would have to be weighed individually as to whether it’s a legitimate gripe with the car or whether the problem is owner inflicted in some way. Again, near impossible to do.
Just my 2 cents, but I'm of the opinion that the only way a reliability study can be done is if a thorough survey of all owners and total research of all complaints made to the dealer and verified by repair orders is done.
No…you don’t need a 100% sample rate. Very valid stats can be done with under 25% sample rate.
One factor that is not taken into consideration is whether a complaint, or multiple complaints, ever makes it ino the files. This has been especially true with Subaru.
Warranty repair done and claim is denied or brushed off by dealer management; ergo, the problems never happened.
Also, the root cause of the complaint may never be known. If the car owner is the cause of a problem there is a tendency among many of them to blame the dealer/mechanic/car manufacturer instead of taking a long hard look in the mirror.
An example (out of countless examples) could be the guy who bought a new Subaru and installed some kind of oddball mobile communiations system. There was a stubby (6") antenna mounted on the top right corner of the windshield.
This caused a wind noise at highway speeds and not only did he refuse to remove the antenna to shut the noise up, he steadfastly refused to accept that the problem was anything other than bad design of the car by Subaru. Imagine his answer to a poll question…
To edit the last paragraph, he also continued to insist that we lowly mechanics “re-engineer the windshield so it would remain quiet”.
A study of ownership costs for 2005 and 2006 model years may benefit a new car buyer.
I am not interested in data from 1985 to 1995. Surely vehicles this old would not be included in the study.
Why are Plymouth and Oldsmobile included in the findings? The last year for a full line up with Plymouth was in 2000. Do the statistics represent the model years from 1990 to 2005? Does a Plymouth Horizon have similar ownership costs as a Chrysler 300?
The statistics are also contrary to most of the experts here with respect to the cost of ownership of a Volkswagen or Mini.
If It’s In Writing And On The Web, It Must Be True And Accurate, Right?
:neutral: CSA
Personally, I Love All These Published “Reliability” And “Cost Of Ownership” Reports…
…not to help me choose a vehicle, but for the influence they have on some people. It drives the cost of purchasing certain high rated vehicles up and brings down the cost of some fantastic used cars that are not reputed to rate highly, whether they’re done correctly, accurate, or have any real value whatsoever.
Most cars I’ve owned and operated (and still own and operate) require virtually nothing in terms of repairs and 99.9%+ reliable have not been rated highly in these studies, and have been bargains.
As I’ve said, rather than rely on blanket recommendations based on a mix of repairs, maintenance, miles, age, make, model, etcetera, lumped together…
I research specific makes, specific models, specific model-years, with specific engines / transmissions and equipment, and that’s what you ordinarily won’t glean from these brilliant reports.
Another tool in my arsenal is the one where owners are surveyed and asked if they are happy with the vehicles and asked if they’d buy another. That says a lot and takes into account an overall satisfaction of ownership, not just “reliability”. For me, there’s much more to a car than reliability. I have to enjoy owning it. I want comfort, quiet, power, and something that’s pleasant to drive.
If I bought many of these cars based solely on some goofy reliability report I wouldn’t be able to stand driving them. I read on this site many times that owners buy cars based on MPG or reliability and then complain about noise, ride, comfort, etcetera, and they want a different car.
Also, most of these published, are published by commercial entities, and part of the “service” they provide is to attract people to the publisher’s business for the purpose of attracting customers. Many of these take miniscule differences in “reliability” and make it show big differences (or at least it is perceived that way by many folks).
So, bring it! I love reports like the one being discussed, as do the folks who buy into the results as being meaningful. I just benefit in a different way. 
CSA