Featured Mazda e-TPV to have Wankel dual fuel range extender

You had good luck. According to statistics on power-train problems in over 2M US trade-ins, Mazda does not fare well.
Mazda - Dashboard Light

 

The site is oriented to measure long-term reliability 5-15 years old. It can help you buy a good used car, even if that car is no longer produced. Everyone always says "oh, the new models are better". But often those new models start having problems as they age. Some brands have a reputation for problems, maybe not in all models, but in multiple models on an on-going basis. Other brands, Toyota and Lexus in particular, produce few lemons and continually produce vehicles with top reliability, even as they age.

When you buy a new car, you are taking a chance. No one knows how reliable it will be. You are trusting a manufacturer and giving it a large amount of money. Will your new Mazda turn out like a 2008 Mazda6 or a 2009 CX-7?

TrueDelta has repair statistics based on user reported repairs (not just power train). It also has a brand reliability chart where you can specify a range of model years. Here is their chart for the 2010 model year:
Car Reliability by Brand, 2010 to 2010 | TrueDelta
Mazda is half way down the list of brands.

I would also like more information about dashboard-light statistics. I wish he could report on problems vs mileage in each traded-in model's stats. How is his quality score weighted for older vs newer models? But even as it is, there is a wealth of data there and that data is based on a very large number of data points.

 

Another problem with dashboard-light.com ratings is that some cars rated with high reliability have much lower mileage at trade-in time. For example, the Porsche 911 is rated "Well Above Average" based on 2167 trade-ins, but the average mileage is only 65K -- half the average mileage of other cars. The Chevrolet Corvette is rated "Exceptional" in reliability but its average mileage is 66K. Cars like these tend to be second cars that are driven on sunny summer weekends. A "faults vs mileage" report would show which cars truly last.

A brand's reputation for long-term reliability is something that inherently needs to be built over decades. A few "clunkers" will hurt that reputation -- that's the way it should be. A new car buyer rightfully worries whether his or her new car will be another one of the clunkers.

As you say (Trollbait), it is a good resource for buying a used car. Every used car was once a new car. If the car had a fault when traded in, it is the original new car buyer who suffered from that fault. That fault was likely the reason they decided to replace the car -- possibly years before they had planned. This site is not useful for those who replace their new cars after 5 years. It IS useful for new car buyers hoping to get 10-15 years out of their purchases because it clearly shows which brands design their products to last.

When I see someone state that Mazda's or VWs have good long-term reliability, I immediately check dashboard-light and truedelta (or, for European cars, anusedcar.com) to see if the available statistics back that up. No set of statistics is perfect, but I would rather go by the data than by anecdotes.

 

No set of statistics is perfect. Unfortunately, since a year ago, TrueDelta data is no longer being updated via member surveys. There is still a ton of useful information there, though. You can drill down and see the repair histories of individual cars, including repair costs. I really doubt that members are falsifying their repair data, but I cannot prove that.

Your experience is anecdotal. Most cars these days are pretty reliable. Even if a particularly bad car has a 20% failure rate, 80% of the cars were fine. Eight out of ten owners would say they are good. Only proper, unbiased statistics based on a fairly large sample of vehicles can show significant differences in reliability.

CR, along with Edmunds and others, measures shorter-term reliability. You cannot say that brand X has improved long-term reliability until you have ample data from 10 year old vehicles. You CAN say they have fewer manufacturing defects after a year or two. Those defects were probably covered under warranty anyway. It's the out-of-warranty failures that cost you money.

Good long-term reliability does not just happen. It does not come from high quality assembly in the factory. It comes from a culture of "design to last". It comes when a company diligently analyses failures and incrementally improves the design of every part. It comes when a company tries to minimize needed maintenance even at the expense of performance or efficiency. Why do Toyotas not use turbos? Because they add a point of failure. Why was Toyota slow to adopt direct injection despite its efficiency benefits? Because it did not want intake manifold carbon build-up like that which plagued VWs. Toyota has a good reputation for long-term reliability and they are careful in the design of their products because they know that one or two clunkers can quickly spoil their reputation. Other brands have less to lose.

 

Mazda 626 1998-2002 Chronic Reliability Issues
Mazda CX-7 2007-2012 Chronic Reliability Issues bad in years 2007-2010,2012
Mazda CX-9 2007-2015 Below Average Reliability bad in years 2007,2009-2011, 2013,2015
Mazda5 2006-2010 Below Average Reliability bad in years 2008,2009,2011
Mazda6 2003-2008 Chronic Reliability Issues bad in years 2003-2008
Mazda Millenia 1995-2002 Chronic Reliability Issues
Mazda MPV 2000-2006 Chronic Reliability Issues bad in years 2000-2006
Mazda RX-8 2004-2011 Chronic Reliability Issues bad in years 2004,2005,2008
Mazda Tribute 2001-2011 Well Below Average Reliability bad in years 2002-2009,2011

A new car bought in 2009 would be 10 years old now. If you had bought a CX-7, CX-9, Mazda5 or Tribute at that time you would higher than average chance of having power-train problems. Only 9 of the Mazda models on dashboard-light were available in 2009. Four out of those nine 2009 Mazda models had below average power-train reliability.

See a pattern? Does that pattern say anything about the company's dedication to producing reliable, long-lasting cars?

Only 2 of the Toyota models on dashboard-light had below average reliability:
4Runner 1990-1995
Tercel 1995-1998

A new Toyota bought in 2009 would have reliability at least "above average". 15 out of 15 models available then.
4Runner: exceptional
Avalon: exceptional
Camry: Well Above Average
Corolla: Well Above Average
FJ: Exceptional
Highlander: Exceptional
Matrix: Above Average
Prius: Exceptional
RAV4: Above Average
Sequoia: Exceptional
Sienna: Well Above Average
Tacoma: Exceptional
Tundra: Exceptional
Venza: Well Above Average
Yaris: Exceptional

See a pattern? Does that pattern say anything about the company's dedication to producing reliable, long-lasting cars? If you want to buy a car that will last 10 years or longer, don't you think buying a Toyota would be a better risk than buying a Mazda? Oh, but CR thinks Mazda reliability is good.

 

Are you aware of any other sites with reliability info on European models? That's the only one I know of. It may not be perfect but it still provides some idea of which cars have more problems, I think.

 

I'm not talking about people buying 10 year old cars or 20 year old cars or 5 year old cars. I'm talking about people buying new cars in 2019 who want their cars to last at least 10 years. Since no one knows how long a new 2019 Mazda will run reliably, the only thing to go on is past performance. How are 10 year old Mazdas doing now? Some OK, some not so good, some downright bad. Toyotas? All good.

I agree that assigning a single rating for a brand is an over simplification. Buyers should look at the model-specific data too. Of course, many will like the cars so much and believe in the brand so strongly that they buy anyway. I took a chance in 2004 buying a newly redesigned Prius. It worked out pretty well for me. I would not have bought a brand-new Chevrolet hybrid design had one been available in 2004, but I trusted Toyota. I think it is a testament to Toyota's dedication to quality that they were able to bring out a new power train design with such good long-term reliability. The old adage that you should never buy a newly designed model in its first year does not seem to apply to Toyota. As other brands will soon be transitioning their product lines to new power trains, will they have reliability problems at first? Will their new cars last? I hope their designs are as good as Toyota's was with the Prius. No one knows. Buyers will have to trust the manufacturer's experience and expertise as they fork out $25K or more. They should base that trust on the available statistical data, in my humble opinion.

 

No web site can predict the future. Dashboard-light presents information about the past reliability of cars. That information can be useful when trying to choose a reliable new car. That information can be useful when judging the commitment of a manufacturer to producing long-lived cars.

When I see a company bring out a model, have serious reliability problems, then drop the model completely, I see red flags. Maybe they don't quite know what they are doing.

Yes things can change. It's often surprising, though, how little they do change and how long change takes.