Toyota Corolla IM, Honda Fit, VW Golf and the Mazda 3. the Honda Fit was nice but my ankle hurt during the test drive. other than that it was nice. The Corolla IM was really nice and very roomy which wasnice. I did sit in the 2019 model but the back seat entry caused issues. Mazda 3 was very comfortable but am a little worried about quality same with VW

Well, you chose well. These are among the most reliable cars on the market - particularly the Toyota Corolla IM, Honda Fit, and the Mazda 3. VW has a six year warranty so you are good for that amount - but it is less relliable than those three cars.

The original Corolla is reliable according to Consumer Reports all the way back to 2010 (2009, a model introduction year, not as reliable - but 2014 another introduction year, perfectly reliable - but it was not as a major change as 2009). The downside to the IM is it is now in its first year of production, and therefore relablity is not known. Any first year is a risk - half the time Toyota gets it right (particularly when they reuse technology). The other half, they are still reliable but not as much as they improve after the first year and have a lot of recalls (perfect example - the new Camry this year has a couple of major recalls, saftey related as well). So there is no record right now on that model - that is the risk here.

The Honda Fit according to Consumer Reports since its redeisgn in 2015 is only average in reliablity, which is bad for Honda but not too bad overall. Climate control problems, paint and trim, body integrity (rattles and creaks), and power equipment problems are more likely. So nothing major, but for the first three years of production they are more likely to have issues and you will have to bring it in to be fixed under warranty.

The Mazda 3 according to Consumer Reports is more reliable at this time than both the Honda and Toyota. They give it a much above average record of reliablity since its launch in 2014. Only thing below average in reliablity and only in 2014 is In Car Electronics. For 2015, 2016, and 2017 the only average thing is In Car electronics. Honestly, that is about as good as it gets in this class. I would say if you like the Mazda 3, get it. You have to go back to the last generation and to the 2011 model year to find anything wrong with its repair record - and that model is way below average with suspension issues. Back to 2010 there are a total of two other fields that get average - everything else is above average. In this case, the Mazda 3 is the best bet because the Toyota is so new.

The VW unfortunately cannot keep up with any of this. In 2015 and 2016, with the redesign in 2015 major things such as the engine, transmission, fuel system, climate system, paint and trim, body integrity, body hardware, power equipment, and in car electronics all have problems to one degree or another (some are average to above average, but all those catagories are not perfect and the cars are only three years old!) You will be using the warranty a lot for that car unfortunately although 2017 seems to be shaping up as a better year although it is still early as those cars are one year old or less.

So you won't go wrong with the Mazda, the Honda is less reliable than a lot of its siblings but is not terrible, and the Toyota is just new so you are taking your chances with it.

Best of luck.

Basically agree with AcuraT.

My 2011 Fit was trouble free for 6 years. 3rd generation Fit is not living up to the reliablity reputation of the 1st two generations. Still not bad. But if your ankle hurt scratch it --- the car that is!

When we moved to the mean, potholed streets of Seattle we decided we needed something about the size of the Fit inside and out, but with smoother, quieter ride & better overall performance. Took a chance on a 2017 Golf Wolfsberg. Now, two years and 12,000 miles later the fuel door has had to be replaced but all else is well so far. The car drives great, has lots of nice features, is very comfortable, has room for 4 people or loads of cargo/gear, and gets a real life 24-28 city mpg and 35-38 on the highway. I really, really like ours. Reliablity -- so far so good, but got my fingers crossed and know a good VW techie addict.

The reliabiity of the 2015-2018s are an enigma. There were at least 5 car sites that did long term tests of one year & up to 40,000 miles with no problems. Google 2015 VW Golf long term test drive. One example: https://www.caranddriver.com/reviews/2015-volkswagen-golf-18t-tsi-automatic-long-term-test-wrap-up-review Their complaints about the display and AT in their 2015 model been resolved by VW.

On the VW forums there are people who driven their Mark VII generation Golf for 3 years and 70K with no problems. Others are snake bit & have had lots of problems as AcuraT described. I don't get it but it drove so nice that I got one any way! That said, I only put on about 6,000 miles a year and rarely keep my cars for more than 5 years.

You mentioned that you want a car that you can drive for " a few years". Hard to know if you mean 3 years or 8 years. FWIW, the warranty on the 2018s is 6 yrs & 72K on the drivetrain. Some of those who have minor lectronic and trim issues just live with them & still love the car for how it drives. Still even under warranty, it is a headache if you have to take it back to the dealer multipe times.

I agree with AcuraT, the Mazda3 has a pretty darn good reliablity track record. As does Toyota. The rear visiibiity of the IM is pretty bad, but the standard back up camera pretty much makes up for that.

Since you are looking at small hatchbacks / wagons, there are a couple more that are I would add to your list.

Kia Soul and Kia Niro. Bobth very good cars cars and likely to be as reliable as any of the ones above.

The Soul is a quite comfortable, extremely reliable. I have driven them several thousand miles as rentals and gotten 32-35 mpg highway.

The Niro is a new hybrid that has gotten quite good profession reviews & rave reviews from owners. The base FE model can get 45-50 mpg and is still very well equipped. As a base model, dealers typically don't stock many. MSRP is $23,000 but will save you money at the gas pump. If you can find an FEs you might be able to haggle down to your price.

Did you test drive the new 2019 Corolla Hatch or the Corolla iM? The new hatchback is a much better car compared to the iM. The journalists are really liking it. It is the first year of a redesign, but Toyota usually does a very good job with their first model years in terms of reliability. I wouldn't worry about that.

The Mazda3's quality is up there with the Corolla, Civic, and Fit. It is a very reliable car. My son has a 2015 and hasn't had any problems since we bought it last year. TrueDelta also shows the repair history is very good. The car is very fun to drive. It really has excellent steering feel like the Golf, but you do reach the limits a little quicker than you would with a Golf or maybe a Civic. The interior is laid out perfectly and feels high quality. As you will find with most journalists, the ride quality and sound insulation are lacking. 2017 and newer ones have improved ride quality and insulation. If you buy used, you can see that some Mazda3s were built in Mexico and some were built in Japan. My son got one made in Japan because we just feel they usually build more reliable cars to a higher standard. The Fit is made in Mexico (and I believe the Golf is, too?). The new Mazda3s are all made in Mexico I think.

The VW Golf should be the one you should worry about with reliability. It isn't that unreliable, but you should be aware that repair costs and maintenance are more expensive. Repair frequency is also higher than the others. The car is a blast to drive, however if you are looking for fun AND reliability, the Mazda is a better choice. Danlisahall's statement about the warranty means you'll be able to keep it under warranty for what I assume is "a few years." If I kept my car for that short, I would take the Golf.

The Honda Fit wouldn't be that comfortable for an hour-long trip. Otherwise, it is a good little car.

The Kia Soul has a more bumpy ride than the other cars, except the Fit and maybe the 2016 and older Mazda3s. It provides a lot of space however and Kia gives you a lot of features for the money.

The Prius should be considered, too. It is about as slow as the Fit. Look at the 2016 and newer ones. The driving experience is much improved over the previous generations, which means they drive well. Our 2005 keeps chugging at 257k miles and my 2016 is at 50k, both with minimal problems. My 2016 is an example of a first model year Toyota with flawless reliability. I could easily take my Prius on a 10 hour trip. I don't usually mind long drives so it may be different for you. It has more space in the trunk than the Kia Niro and gets 55 MPG. The 2017 and newer Prius has all the safety features standard.

So, the Toyota Corolla Hatch, Mazda3 hatchback, Kia Soul, and Prius are the ones you should look at. The Golf included if the time you keep it falls within the warranty. The Fit is just not that comfortable as Danlisahall said. I'd skip the average performing Corolla iM unless you only care about reliability. The new Corolla Hatch is just much better.

The VW Jetta is not a very reliable car. Every year it is built according to Consumer Reports it is average, below average, or much below average. Going back to 2000 the best it can do is average - and it only did that well in 2004, 2005, 2012, 2016, and 2017. Otherwise it was below average every year since 2000 or worse.

While the powertrain is usually fine for about ten years (engine, transmission to a lesser extent) the fuel system is a chronic trouble spot much below average - the worst rating - through 2015. Climate system is the same - trouble almost every year through 2015 (below average). Five years out, that is the first time major transmision issues shows up in the survey as it is average reliablity.

So for five years, you are highly likely to have trouble with those two components. Once you go older than five years, it starts having a lot more problems (like the transmission). So if you are okay having those two issuses and getting lucky on the transmission, those are the most likely things to have issues with. If that is okay for you, you are pretty much good to go as everything else is pretty much above average for five years on the car. It just depends if you want to deal with it or not.

The 2019 Jetta is a major redesign so previous history is of dubious value other than to say previous VWs do have a rep for requiring above average repairs. 2019 has an new 8 spd AT. Hopefully is much more reliable than past models but the jury is still out on that.

The earlier Korean twins, Kia & Hyundai were not the highest rated cars, especially the early Rios. The models from 2015 & beyond are light years better. Current models are both reliable, well screwed together, have great content, highly functional, and giving the Japenese automakers a run for their money. It would be a mistake to base your current decision on your previous experience based on owning a 1st gen Rio.

Enjoy the shopping experience and please keep us posted.

Danlisahall - Very true - I should have added this only applies to the cars in the older design, not the new design - that newer design is probably even less reliable as the last redesign in 2010 shows. Of course, this does apply to the 2018 he is considering since that is based on the old design.

I'm with everyone so far that has said the Jetta has never really been reliable. The previous generation didn't even drive that well so that car is a hard pass. The 2015 and newer Golfs are much better both in reliability (mostly) and overall enjoyment. I wouldn't really recommend either if you plan on keeping the car until it is 8 years old. I question Consumer Reports' reliability ratings with the Jetta. The "reliable" years are such random years. Almost every year had some type of electrical problem whether it was the horn or the door wiring. They take a deep plunge with depreciation, too.

Like Danlisahall said, you are making a big mistake not looking at the current Kias. Kia and Hyundai are some of the best mainstream brands right now. The new ones are about as reliable as current Mazdas and Hondas. Maybe test drive an Elantra GT, the new design. A buddy of mine test drove one, and he said he liked the interior, powertrain, and driving dynamics, but the ride quality was really bumpy. The Golf is more engaging and a better car overall, but the Elantra is cheaper and gives you a huge powertrain warranty. The first gen Rio is not something you want to remember Kia by.

I looked up the Golf on both TrueDelta and on Consumer Reports. Neither find the latest generation VW Golf all that reliable.

TrueDelta is nicer - it says it is average with a repair rate of 128% (100% being average). On the color scale it gives it a yellow.

However, don't take my word for it - here is the page where TrueDelta participants tell of their issues on the Golf. A lot of electrical problems on new cars (cruise control failure, window wiper motor failure, you name it):

https://www.truedelta.com/Volkswagen-Golf/problem-histories-279,2015-2017

Worse is Consumer Reports. Much below average (the worst rating possible) for 2015 (the redesign year) and 2016. However, in 2017 Golfs improve to above average, which is better than most VWs.

Golfs in 2015 and 2016 have some incidence of minor engine problems, major transmission problems, minor transmission problems, serious fuel system issues (below average in 2015, average in 2016), climate system concerns (average), power equipment (average), and in car electronics (average and below average).

In 2017, they are basically delivered with squeaks and rattles (body hardware) but after that, no issues. So maybe they fixed the problems after two years. Again, you take your chances with VW as it seems no matter what, they have problems. I used to own one back in the mid 1990s, a VW Fox with power nothing - it was a great car. Today, loaded with electronics they are something I would never consider again - no matter how great driving. I would buy a Mazda, Kia, or Hyundai in this class.

I don't think we should be so dependent on CR's reliability. I would instead look at reliability trends. We can conclude that VWs have their problems and they are expensive to fix. If you are on a tight budget and don't want to pay for repairs, simply don't buy a VW. Some people are fine with that so the car is great for them. My first car was a 1982 gasoline VW Rabbit and it had a few electrical problems until a trash truck ran over it.

The 2016 Civic had the infotainment system problem. The bluetooth and USB randomly disconnect from the owners' phones. The system (running an old Android OS) also has reports of crashing. Pretty much everything else with the car is reliable. The 2017s and 2018s don't have this issue nearly as often, if ever. The 2016 is an example of an unreliable first model year. That is why you see the reliability ratings slipping. The problem with this is that sometimes the head unit replacement counts as an interior warranty problem. I don't know about the other brands, but my Prius had a 1 year/12k mile warranty for anything that breaks or rattles in the interior. Report the problem early during the warranty period and fix later if need be so you can get a free fix.

There is a reason why I rely on Consumer Reports so much - they have never lead me wrong. I buy a reliable car - I don't have problems. I buy one that is not so reliable, and I tend to have problems.

For example, back in 2001 I bought a Acura TL. Consumer Reports at that time showed some problems with the transmission but I gambled and bought it. Five years later, 60,000 miles - needed a transmission rebuild (Honda did pay for it, I paid for the rental waiting five weeks for a transmission).

In 2003 bought a Honda Accord V6 with the same warning from Consumer Reports about the transmission. This time it failed at 30,000 and again at 60,000 before I ran the car into the ground at 116,000 miles (it had a lot of problems then after 10 years).

In 2006 I cheated and bought the Saab 9-3. I say I "cheated" because at that point I was working as a consultant to GM fixing their quality teaching them six sigma and how to apply it to data they colllected from their warranty repairs. I knew that counterparts of mine had worked with Saab from 2004-2006 on at that point, the brand new Saab 9-3 (launched in 2003). By 2006 the quality had been fixed (I trusted my counterparts) and I bought that model year. That car was a miracle - troublefree for the first 100,000 except for an alterntor. That was it. As you know, lasted until this year when a tree fell on it. Consumer Reports in 2006 gave it a perfect reliablity score (after a way below average score in 2004 and average in 2005). This also showed GM in their own warranty data on the Saab brand six sigma really does work - and it is why Toyota has been onboard with that process (Kaizen in particular) since the 1980s when it was developed by GE, Toyota, and 3M (they shared ideas in putting the structure together). Now almost everyone in the car industry has adopted it (only American maufacturer not to adopt it but that is changing now was Fiat/Chrysler).

In 2013 I bought the Subaru Legacy H6 and again Consumer Reports warned me of one problem - the oil consumption issue with the reports that many Subarus have this issue (also, my parents owned two of them based off a recommendation I made for them and they love the cars - but they did have one that burned oil of the two). Mine of course started burning oil at 44,000 but that is easy to monitor and compenstate for. More importantly it works well in the winter and in the snow climbing my 3400 feet mountain. It means my wife gets home with no problem no matter what in the winter.

I definately use Consumer Reports for trends. Is it always going to be right? No. But it will give you a very good idea who builds a car (or model) that lasts and which do not. Also, it clearly shows no matter what your company is, the first year of production is the pits for reliablity. You can not rely on it on TrueDelta or Consumer Reports (many don't) and then get stuck with something you are always taking to the shop (say a Chrysler 300, or Fiat 500).

The Civic you are correct had a lot of infotainment issues. The electronic dashboard did not help matters either (faded in bright Florida sunlight for example). If you are in bright sunlight a lot, that would be a reason to avoid it.

Consumer Reports's methodology has some serious flaws, which is why I started TrueDelta's survey. Due to these flaws some of their ratings can be way off. For one thing, models that were introduced shortly before their April survey (such as the Kia Niro) can get a big bump. For another, if owners really like a car for reasons other than reliability its reliability score can get a big bump. Plus their data are usually old. TrueDelta's stats are currently based on data 14 months newer than CR's. Then there's just plain noise, due to which the ratings for nearly identical cars can end up very different.

Read their results with the above in mind, and you'll be okay. Just don't try to split hairs with them or consider them the gospel.

In response to the OP, my main concern with a Mazda in Canada used to be rust, but they seem to have improved the corrosion resistance of their cars back around the 2009 or 2010 model year. A close look at some 2010 Mazdas (especially the rear wheel openings) should confirm or refute this.

VWs can be okay for four, maybe six years. After that they're much iffier. As with others, a model in its first year is much more likely to have problems.

Hyundai and Kia are a bit hit and miss. Most owners have good luck. Some do not, and Hyundai and Kia are very strict when determining warranty coverage. Their cars are riskiest during the first model year. Both companies are good at identifying and resolving common problems, so by their third model year the cars tend to be reliable.

Hondas are similarly riskiest during their first model year. Honda tends to be far more generous than Hyundai or Kia with out-of-warranty assistance, especially if engineering flaws produce a common problem.

If you must buy a car in its first model year, Toyotas are your best bet. But even these are better after the initial model year.

Of the cars you mention, I wouldn't have reliability concerns with the iM, Fit, or Mazda3. The Fit is a smaller, weaker, less substantial car than the other two.

AcuraT,

If you were teaching Six Sigma, you clearly know your stuff. But beware reconstructive memory and confirmation bias.

Looking at CR's April 2004 auto issue (which would have used data from their April 2003 survey) there are no warnings about transmission problems in the Accord, Odyssey, or TL. For the TL, the worst transmission dots are averages for the 2001 and 2002. The 1999's transmission is "much better than average" and the 2000 is "better than average." For the Accord the worst transmission rating is "better than average." For the Odyssey, which later emerged as the worst of the bunch (probably because its higher curb weight put more strain on the transmission) the 1999 trans got an average, but the 2000+ transmissions were rated better or even much better than average.

At the time you bought the TL and the Accord, the transmission wasn't yet a trouble spot in CR. The black dots for the transmission only showed up with their April 2005 survey, first reported in October 2005, far too late to have informed (or not) your purchases.

Now the Saab.

The 2006 Saab 9-3's reliability was first reported in the October 2006 New Car Preview--and it has a big black dot, the opposite of a perfect score. The 2005 is rated average. The 2003 and 2004 have big black dots like the 2006. A year later, the only change is that the 2006 now had a half-black dot. Two years later, the 2005 has "worsened" to half black, while the 2006 remains half-black. The 2003 and 2004 are full-black throughout.

Maybe you're remembering a "perfect score" from a road test evaluation, not from their reliability survey?

And oil consumption in Subarus? I don't think CR reported this until 2015, with a big story in June of that year. There's no indication of it in the October 2013 new car preview, based on the April 2013 survey. In that issue, the Impreza, Legacy, and Outback engines still have full-red dots--much better than average. So I don't think you were forewarned by CR in this case, either.

In the cases of the Honda transmissions and Subaru oil consumption, the evidence was eventually so overwhelming that even the worst survey would have caught them. Then again, Teslas require repairs so much more often than the average car that any survey should reflect this as well, but CR's doesn't. They're still predicting average reliability for the Model 3 based on the most recent records for other Teslas, which have benefited from years of quality improvements. And which also benefit from the flawed wording of CR's survey. The Model 3 requires repairs about fives times as often as the average car in our survey, the other Teslas 2-3 times as often (but as bad as the Model 3 earlier).

We all have memories that reconstruct our memories based on subsequent evidence, and a tendency to read from conclusions to the evidence rather than the other way around. The only safeguard I know of is to start from the bare facts every single time. Memories suck.

How do you have the information from all of these old surveys? That would be interesting to see. I throw away my CR magazines after they become irrelevant.

For Hyundai/Kia, I would just avoid the 2.4L engine from 2011-2014 because those were prone to failing. Most of their other products tend to hold up well.

CR definitely got it right with the Honda transmissions failing. Ours failed at 40k miles because we didn't have the recall performed in time. The second one has been good for the next 145k miles. Other than the transmission problem, the rest of the reliability ratings for the 2002 Odyssey have not reflected our actual problems at all. The early 2000s Honda transmissions are some of the most notoriously failing parts of any car in the last two decades. Simply, people now know to not buy them.

Quoting Michael" <<If you were teaching Six Sigma, you clearly know your stuff. But beware reconstructive memory and confirmation bias.

We all have memories that reconstruct our memories based on subsequent evidence, and a tendency to read from conclusions to the evidence rather than the other way around. The only safeguard I know of is to start from the bare facts every single time. Memories suck.>>

Wow Michael, that sucks! But it is good reminding us "... recontructional memories and confimation bias" great gut check for each of us. We can tend to think we know it all when in the end we only know what we know, and we don't know for sure if that is accuate. Thanks for the humbling reminder!

Did you major in psych or statistical analysis?

For the curious: https://www.linkedin.com/in/michael-karesh-b0a8763

Ph.D. Sociology, which includes training in survey research and statistics. And also some social psych. I have CR auto issues going back to 2004, with some even older than that. Also Consumer Guide for pricing and features from 1983 until they ceased publication with the 2009 edition.

Yup, thought so! That is quite a collection of CRs you have. I really liked the Consumer Guide and miss its very useful reviews. You definitely have the qualifications for a trustworthy car guru and True Delta bears witness to that!

I truly appreciate all the years, energy, & talent you have invested in TD so very much! Not looking forward to the loss of TD either - or at least your expertise and forthright opinions.

I'm not quite sure -- will this "My Next Car" part of TD continue an will you continue to contribute to it?

What is your next venture?

PS:I agree with your acessment that the P5 was / still is a wonderful little, timeless car. I had its predecessor, an 85 Mazda GLC. When the P5s came out I drooled over them and was always was tempted to buy one. But before they came out, I got fixated on Civic wagons startin in 88 AWD then their wonderful 5th generation 92-95. Between myself and my son, we had a VX & Si, plus a couple Del Sols - like the P5, they were light weight, had reasonable power, and pretty basic except for the engines & suspension -- sweet cars to drive. And you could coax 40 mpg out of them. Then I needed more room & went to Subies, so I never got a P5.

Weighing a few different things for the next venture.

Now that I've freed up a bunch of time I hope to participate more than I have been in My Next Car? There might be ways we can tweak it to improve it. I do like the instances where a real discussion with the car buyer happens, and there might be a way to make this happen more often.

Thank you Michael. Obviously my memory of the results is wrong, and my CR results only go back to 2007 as I threw out the older stuff about six months ago. Last time I go by memory.

Looking at the 2007 ratings it clearly show a black circle on transmissions for the 2003 Odyssey, V6 Accord, and shockingly - the 2003 Honda Civic Hybrid. So it must be the survey showed it after I bought the car (within four years). For Acura by 2007 there were a lot of black circles on the CL, MDX, RSX, and TL for 2003. Showing the problems after I bought both Hondas.

Looking at April 2007, the Subarus have black ciricles on their engines for 2001, 2002, and 2003 for most models including the Legacy, Outback, and Forrester. They did not call out oil consumption until later, but were reporting engine problems long before that. I only can expect that those early engine issues are oil consumption as friends of mine in the northeast who bought those very models many had oil consumption issues out of college. That does not mean there is a link, just a strong guess that it may have something to do with it. Also, looking at April 2012 the engines decline over six years to average or below average on those three cars and including the Impreza now - so something is constantly bringing down the engine scores even five years later in CR. The trend is there - going back in time.

Also, you can cheat on Subaru using the web - here is mention in 2012 - about the chronic oil consumption issues on these cars:

https://allwheeldriveauto.com/why-does-my-subaru-use-oil/

I did use that site so I knew what I was getting into with my Subaru.

As I mentioned - I did not rely on Consumer Reports in buying the Saab. For that I relied on my fellow consultants. In 2007 after one year it had only two black circles and more red partial and full circles (and some average fields), better than the previous year (and much better than in 2003 the remodel year and 2004). Problems were more likely at that point with electrical and power equipment. As I mentioned, I lost the alternator in the first 100,000 miles but that was it - which matches the black circles on that model year.

Going to April 2012 the ratings for the Saab 9-3 rose to average, not above average. My car (2006) had issues with engine cooling, fuel system, electrical, paint and trim, body integrity, and power equipment. Besides the alternator at 70,000 and some squeaks and rattles and a number of rubber parts failing in the engine compartment (which is normal), I guess I was lucky - I did not experience any of it.

I note while you publish all the data regarding the electrical issuses on the Tesla 3, you compare it to the general "average" score they threw it on Consumer Reports with no data behind it. With this I complete agree with you - when that happens I don't trust Consumer Reports at all. Trying to create a reliablity rating when you have no data on a model is at best misleading and at worst can be competely wrong (as it is in this case). The only time you can leveage their data for trends (such as engine issues with Subaru) is looking at the historical record of real data ignoring the guessing they throw in there. On the Mazda 3 however, TrueDelta and Consumer Reports come close to complete agreement, they both say the car is reliable over time. This is when both have data on a model.

I don't disagree with you - I use it as a guide which is why I refer to it in many answers (I write they don't have any results when they are guessing in all my responses, like on a BMW question on the boards recently). When I bought the Envision this year, TrueDelta had nothing on it - I only had Consumer Reports to go by. It was above average for two years on everything except Climate System and in car electronics (so overall fairly reliable) and it was not the launch of the car since it launched in China the previous year (fourth year of production in 2018). I know wny the Climate System is below average - the controls are ones that were eliminated by GM on all Buick models but this one - and it is the "touch" screen that does not react well. It works, but it is slow so people don't like it. I know about that issue and it does not bother me, but it does work. That is why the electronics score suffers as well. Assuming that is it, it will serve my purposes and last a long time.

Thank you for correcting me and my faulty memory.

The older Subarus most likely have black dots for their engines due to head gasket issues.

You don't have to be lucky to avoid a problem in a "black dot" area. One of my major complaints has long been that how CR presents its results leads people--even those as knowledgeable as you are--to think problems are far more common than they actually are. A system could earn a black dot with a problem rate as low as three percent--meaning 29 out of 30 cars might not have experienced the problem.

I wrote about the current system when it was introduced in 2006:

https://www.truedelta.com/pieces/newdots.php

And my basic critique:

https://www.truedelta.com/pieces/shortcomings.php

Basically, you've probably been reading far more and far more precise information into those dots than is actually present in the data behind them.

Unfortunately, hardly anyone has really cared about these issues, such that CR has felt no need to rectify them, and I stopped harping on them. The 13 people who were receptive got the message. To the extent other people criticize CR, it's for simplistic BS like "bias."

Forgot to comment on the original Civic Hybrid. We haven't had much data on those recently, but back when we did (for the 2003) they were quite bad. One of the goofs Honda makes from time to time, and best avoided.

Michael,
Those black ciricles on Subarus are partially due to what you describe, but you are just wrong that the burning oil problem is not the reason for them and it is just the head gasket. The article I linked you to above is dated February 2012, when the first reports were logged on Subarus burning oil on cars a few years old - not just the head gasket issues you are claiming. Here is the first few paragraphs descirbing exactly what I have been stating - that the Subarus were burning oil well before Consumer Reports reported on it in 2015.

There is the Jetta Sportwagen, which is the old wagon model sold up to 2014. The Golf Sportwagen is the new wagon 2015 and newer. The Alltrack is the "off road" version of the Golf Sportwagen. The Sportwagens and Corolla iM will be the most comfortable out of the cars you are considering. The Mazda and Hyundai are more sporty and have a stiffer ride. It depends on what you prefer. The Mazda should be more reliable and efficient, and it would definitely be the one I would choose. The Toyota would be the most reliable though. You could probably get an outstanding deal on a 2018 Corolla iM because dealers are trying to dump them.

Hyundai and Kia are sister companies by the way. The current Forte5 and the 2017 and older Elantra GT are very similar cars mechanically, but I see you felt otherwise.

FWIW I reviewed the Elantra GT, comparing it to the Honda Civic Si,here.

Lectrofuel:

Thanks for clearing up the VW wagon situation.

If CGfighter drove the new Elantra GT, it is a newer and much better car than the 2018 Forte. I've personally never felt that the Forte hit the mark. Kia is introducing an all-new 2019 Forte this fall--but initial impressions aren't very promising. I'd personally prefer the Elantra GT, at least in Sport form, to either generation of Forte.

AcuraT:

No issue with using CR as a source--I check their info from time to time myself. I was just cautioning you against reading more precision and consistency into it than is actually there, and against thinking the differences are larger than they actually are.

One other point: your pesonal sample size is one, and you know as well as I do what that means. Odds are against your personal experience always matching what CR reports, even if CR was 100% correct about THE AVERAGE.

You might not believe how many people have told me that TrueDelta's sample sizes are too small, only to then make a point based on one or two cars.

I was comparing the previous gen Elantra GT and the 2014-2018 Forte. The new Forte will probably be better, but now it has a CVT so that is a deal-breaker. The 2019 has standard Apple CarPlay/Android Auto and a lot of other nice features on the base trim. From what I've read, it isn't engaging and is more comfort tuned.

Michael, Appreciate your comments, and you are right about CR being overly sensitive about 1% differences on late model cars (the last couple of years). Only as the years go out do the differences really mean more and it is all related to that average they calculate. Sample of one? The Saab? It is statistically insignificant, I am quite aware. Must have been much better than the average Saab. Made up for the worse than average Hondas I have owned.I guess. All averages out in the end. True Delta's database is too small? I thought you don't publish results unless you have a certain number of responses in any case? Thanks.

He publishes the results, but notes with an asterisk the models that have a small amount of responses. Those Hondas were still reliable. The transmission wasn't, but everything else on those cars were very well put together, albeit with barely any sound insulation, an old Honda tradition. If you got the 4 cylinder Accord it wouldn't have the transmission problem.

To not have an asterisk we require at least 25 responses that cover at least 75 months of ownership. Some people think we need sample sizes in the hundreds or even thousands, even though no one has a minimum that high, because they don't really understand statistics or are basing their opinion on a field where very small differences must be measurable with a high degree of certainty.

If you want to distinguish between a 2% problem rate and a 3% problem rate then, yeah, you'd better have a very large sample size. Even CR doesn't have sample sizes large enough, which is why whenever they try to split such fine hairs they end up with iffy results. But there's no practical reason to measure differences that small--they'll have no meaningful impact on anyone buying just one car because they'll be swamped by random variation.

Sounds like your getting close. There is no perfect car but there is one out that that will suit you well. Love to learn what you get in the end and why.

Michael, how reliable do you think RepairPal's reliability ratings are, in your opinion? They rate the Fiat 500 "Above Average," which is the same score as a Toyota 4Runner. Seems flawed. They say, "Last year, RepairPal created the first ever Reliability Rating based on actual vehicle repairs. For 2018, the updated RepairPal Reliability Rating has analyzed millions of repair orders from over 2,000 auto shops across the country, measuring dependability by evaluating the cost, frequency and severity of repairs. While other reliability scores are based on subjective results from consumer surveys, RepairPal's Reliability Rating uses real-world data to help consumers seek out the best car for them."

Thanks for that clarification Michael. I always thought the data you presented was good data, just was not aware of the asteriskmeaning.

Yes, Consumer Reports does not have a thousand cars in many surveys - except perhaps cars like the Toyota Corolla or Honda Civic. So in rare cases the averages will work - problem is, we don't know which cars have that many responses as they don't tell you which gets back to your point - what does as 1% difference really mean?

Lectrofuel, I posted all the problems I had with the Honda Accord dating back to 2007 (not listing the transmission replacement as it happened before I joined TrueDelta and I bought the car in 2003). It is quite extensive. Here is a summary of all the problems that car had from 2007-2009:

January 2007 - transmission overheating problem, corrected by adding a oil sump pump installed at no charge from honda

April 2007 - driver seatbelt latch failure. Dealer claims nothing is wrong.

June 2007 - driver seatbelt latch failure - dealer replicates (gets stuck in car, I have to help him out). Sheepishly agrees that the seatbelt latch needs replacement.

Sept. 2008 - rear defroster malfunction - it does not turn off automatically anymore and worse, does not turn on sometimes. Defroster timer needed to be replaced.

January 2009 - wiper motor shows corrusion and needs replacement.

February 2009 -Clamp on left front brake froze prematurely. Replaced.

May 2009 -Bumper broke from ice buildup. The mudflap for the tire collected ice, and it broke off the bumper breaking it. The bumper needed to be replaced so it did not drag on the ground.

November 2009 -Second transmission problem, this time a seeping transmission seal identifed during oil change. Part not in stock so had to return a week later to have it installed - took 2.5 hours and would have cost $250 without extended warranty.

November 2009 - Wear of rubber "boots" on trasnmission seals. Normally would have cost about $250-$300 to repair but extended warranty covered it.

As I said, my experience with Honda may not be normal but all I had was problems with it. Bringing the car in multiple times a year to keep it running is not normal (especially early on in the life of the car as I list above from years four to six of ownership). I had the extended warranty on that car for six years, for $800, the dealer put over $3000 into that car with all the transmission and other problems it had before it ran out in 2009. I then put more into it to get it to 140,000 when I finally gave up and sold it when the engine mounts, transmission, and GPS were all failing at the same time.

I had a 2001 Acura TL before that which was before I joined TrueDelta (never put that car into the system). It did not have problems the first three years which is why I bought that Accprd. Then everything started to go on the Acura in year four so it only lasted five years and 120,000 when the second transmission was failing on it (high mileage car). As you can imagine, my two Hondas with terrible reliablity problems swore me off for life. Before these cars I had owned a Acura Integra but that car was troublefree and I sold it with 170,000 miles on it to a friend who drove it another 40,000 before the transmission went and he junked it). I don't know what happened but for me, in the early 2000s, their quality went downhill rapidly and all I had was headaches with the two cars I owned. Reports from TrueDelta and Consumer Reports that with the V6 engines they still have transmission issues from time to time (particularly around redesigns, as there have been extensive reports of rough shifting from the newly launched 8 speed transmissions the first year or two of production).

Since then I switched to GM and Subaru and done much better, with better built cars that last over 100,000 with few issues (still have to wait on the Subaru which only has around 75,000 but besides burning oil, runs fine - although my parents have made it over 100,000 with few problems except oil burning on some cars based on a recommednation I made for them). I am pretty tolerant of cars and will fix them (either myself or with the mechanic) and I do try to keep them a long time (150,000 to 200,000), so finding a reliable one is important to me. Toyota and Honda used to do it for me, now GM and Subaru do since I need AWD as snow tires won't get me up my mountain alone and only Lexus has AWD except for some Toyota SUVs.

Lectrofuel,

RepairPal's results are all over the map. As far as I can tell they're analyzing a bunch of unsystematically gathered data with no controls for its obvious weaknesses, and just posting whatever they come up with.

I asked them if they'd like to work together to provide truly useful reliability information. Some initial discussion, then crickets.

Bottom line is there's little interest on either the supply or demand side anywhere in providing high quality car reliability information. On the demand side, most people just want enough material to build a story in their head that they like. On the supply side, there's no business case for doing it the right way because, apparently, too few people can tell the difference between good and bad information and are willing to pay for the former.

AcuraT:

It is possible to post all of this through the reliability survey if you want to--it won't affect the stats, but will be posted.

Historically, Subarus have gotten expensive once over 100,000 miles. Aside from the head gasket issues, their suspensions started requiring frequent repairs (control arms, wheel bearings, many headlight bulbs). Granted, most of these repairs were't terribly expensive, so more of a nuisance than a big hit to the wallet. Those since the early 2010s might be different, we'll see.

AcuraT: Man, you have had your share of car woes. I assume that you don't drive any harder than the ave Joe and maintain them per manufacturer's recommendation.

The old saying that the way to keep car costs at a minimum is to "drive it till it's dust" may have been true in the old days, but given the complexity of the electronics, transmissions, emissions systems... since the start of the 21st century, I wonder if that is still true.

The cars of the 50 - 70s needed full tune up every 5-10K, oil changes every 2-3K, muffers every 3-4 years, annual wheel alignments, new belts and hoses every 4-5 years... Plus anyone with a screwdriver & a few wretches whowasn't afraid of dirty fingernails could do it themselves. Now all those maintence intervals have been greatly extended - as much as 10x for some of them. But for the DIYer to do anything besides change the oil you need a OBD tool, tech degree, and 6 different types of screwdrivers before you even open the hood!

So I can't help but wonder -- and now I enter into pure speculation based on my hazy recollections... Prior to 1990 the 1st 5 years, 60,000 of a cars life were probably more expensive than since the 90s. On the other hand, I suspect that the cost of ownership after 5 yrs / 60K has gone up in the past 20 years.Of course parts and labor were less expensive. Or were they really if you figure what percent of the average annual income that might have been? Are there any sound figures on the cost of car ownership as percent of average annual income over the decades?

As a young guy I wasn't very hesitant to buy a used car and wasn't concerned if my car had 150+K. I could replace and alternator or muffler, due a tune up... Not so much as as cars get increasingly complex and I get less agile! Again, this is speculation but now days I figure I'm even or ahead of the game to sell my car after 5-7 years as it approaches 70,000 miles (my annual mileage is fairly low). That plus the advances in overall performance, fuel efficiency, comfort, and safety happen so quickly that a 7 year old car is pretty dated.

Do the rest of you think there are still significant advantages to long term (> 7yrs) to long term ownership?

Michael, yes, that was pulled from all the stuff I put into TrueDelta - it is in my record. I am someone who probably dragged down the Honda average with my relatively recent bad luck with them. Like I said, before the turn of the century, had few problems and they lasted a long time. Since 2000 - both were much worse than my preceeding Hondas, and one of the cars - I put all the repairs from the third year on into the system. I have been dilgent since then, and when I look at my 12 years of Saab ownership, it was much less problematic than the Honda was. I did not bother going back when I owned the Honda because I never made the time to do it - sorry. Don't own the car anymore so don't have the records. I have been diligent since I joined your board for over the last 13 years.

Danlisahall, you are right. Older cars were much easier and you needed much less computer equipment to fix it. Today, the computer tells you what is wrong 80% of the time and you just replace what it tells you to do. It is why I don't really do much on my cars anymore although I have all the tools still (wife wants me to get rid of the socket wrenches I no longer use as I have switched to working on the house - and you don't need socket wrenches for that).

I do keep cars a long time and I don't "beat" on them, and I am diligent about changing the oil every 3000 miles not using synthetic, and every 5000 to 6000 using synthetic. On the Honda Accord it was every 3000 miles, on the Saab every 5000 miles. On my Enclave it was every 6000 miles (I sold that car early at just over 70,000 since I did not need a car that big anymore after nearly 6 years although it ran well). My first new car in 12 years is this Envision (although my wife got a new car 5 years ago) - and I did the first oil change at 6000 miles. Up to nearly 9000 miles and I started driving it the last day of April, so it is high mileage car.

Yes, the cars of today are more expensive and have more expensive equipment. However due to the use of more advanced statistics in troubleshooting (as in six sigma for manufacturing processes), cars today built in many ways are far better than they used to be. It is why I am hoping the Subaru lasts me more than 100,000 and just burns oil for a long time. Most cars do last 100,000 now and you can have few issues, the trick is getting it to last another 100,000 without too many repairs.

I am very diligent in keeping track of my costs after the first three years of ownership. I have a spreadsheet where I track repairs and maintenance costs separately in two columns. When I see the costs rising that is when I get rid of the car. Happened to the Accord at 10 years, happened to the Saab at 12 years and a tree fell on it. Right now the Subaru needs maintenance (oil) beyond what is normal, but that is it - nothing is breaking. So we will keep it until it starts costing more than what the car is worth to repair. In other words, until something major breaks like the transmission after 12 years.

You are right - the newer cars have a lot more technology, so I was planning in any case if the tree did not fall on the Saab to replace it in the next couple of years with a modern car with the crash prevention technology. Now I have that car. For me, it is cheaper to keep my current car running than replacing it every four or five years because I hit 70,000. That just is very costly and I don't want to do that. However, when I do buy a new car it tends to be nicer as I want the latest safety features that make sense (why my Toyota in the 1990s has automatic seat belts back then as well as a driver's air bag).

I used to buy used which is why I have that checklist I posted on here one time that I put together from working on cars as well as reading a lot of books. I may still do it again when my kids need a car to drive (twins, they can share). It all depends if we can get the 2013 Subaru up to 200,000 or not without spending a fortune on it. If it runs well, they get the car in 2024 when it has about 150,000. Otherwise they get a nice used car I select using my skills and that checklist I developed (and worked for me and others very well so it has been tested with over 10 cars purchased using it).

Michael, I think the other reason that reliability data websites struggle is that most people think car shopping is a chore and default to Honda or Toyota if they want a reliable car. They aren't willing to pay for the good data or they just trust CR's data as if their data is a fact, which isn't the case for any website.

AcuraT, sorry to hear about those problems. Best of luck on your Subaru and Buick.

Problems with our 2002 Odyssey have been sporadic since it was new. It used to leak oil and the A/C Compressor (belt driven) would squeak whenever the AC was turned on. Those problems somehow fixed themselves without us having to do anything. Power steering pump broke, 4th gear pressure switch broke, transmission replacement right after the recall started at 25k miles, the rail for one of the power sliding doors broke, the rotors have had to be replaced or fixed probably 7 times, the radio turns up to full blast whenever you quickly turn the steering wheel all the way, and all the wear and tear my kids have put on it over the last 16 years like breaking the 2nd row seat recline and cupholders. The engine had a 5th cylinder misfire in February on a trip to Mammoth. That CEL cleared itself. Also, while parked with the engine on it sometimes smells like oil is burning, but the oil level is always perfect. In the end, the rotors were the most annoying problem because it made the car feel unsafe under hard braking. Literally, the dashboard would jiggle, your teeth would chatter, and the steering wheel would shake your hands off of it.

This OBDII adapter I got on Amazon is great for people who drive a lot or have an old or unreliable car. Every car enthusiast should have one as the possibilties are virtually endless with what you can do with it. Connect it to an IFTTT account, Amazon Alexa, Google Drive to make automatic spreadsheets, etc:https://www.amazon.com/Automatic-Connected-Realtime-Diagnostics-Detection/dp/B01JRBQ9PC/ref=sr_1_3?ie=UTF8&qid=1536264701&sr=8-3&keywords=automatic

Danlisahall, I think there are still advantages to long term ownership depending on what car you have. I don't see myself selling my 2016 Prius until it dies unless I get a reliable weekend car that I can daily drive like a Miata. My car has all the nice safety features and I'm expecting it to last to 250k miles with little to no problems like our 2005 Prius is doing. However, there are a lot of safety reasons to buy a modern car if you have one from the early 2000s or older.