Been cross-shopping with....

[Xavier6162]

....the BMW 335i and noticed that there are a lot of reliability issues. And I see a lot of owners claiming they would never keep a BMW past the warranty period.

So now I’m back with the GC and all is ok with the world…until someone mentions reliability problems with this car too.

Has anyone mention anything about reliability?

 

[Sal Collaziano]

There will ALWAYS be problems with cars.. Some are more expensive to deal with than others.. The Genesis Coupe will have a longer lasting warranty and will probably be less expensive to fix after that warranty expires...

 

[mkaresh]

We have quite a bit of reliability information on the BMW, starting with this page:

BMW 3-Series reliability comparisons

Reliability for recent model years is consistently average. Of course, if you do have a repair, it will be expensive.

Too soon to have results for the Genesis Coupe yet. Might have a partial result in November. 23 owners signed up to participate so far, so it's going to be tight.

 

[TJPark01]

We have quite a bit of reliability information on the BMW, starting with this page:

BMW 3-Series reliability comparisons

Reliability for recent model years is consistently average. Of course, if you do have a repair, it will be expensive.

Too soon to have results for the Genesis Coupe yet. Might have a partial result in November. 23 owners signed up to participate so far, so it's going to be tight.

I question that data. There are too many versions of the current 3 that aggregating them all and spitting out a number is only going to give you an average. I mean do you think the maintenance on a M3 is going to run parallel with a 328i? What about the 335d, is this data just for North America or is it factoring all the Diesel and petrol version in Europe as well.

 

[mkaresh]

Nearly all of the 3s in the sample are 328s and 335s, and the OP mentioned the 335. Most cars are in the U.S. and Canada, and there are few if any diesels (the 335d has sold poorly).

But among fairly new cars engine problems are rare, and the parts that tend to require attention are shared. Among older cars it becomes more important to provide separate results for each engine.

Even if there were a number of M3s in the sample, you're not bound to get an average result any time you combine multiple variants. You'll only get an average result if all variants are about average, or if some variants are actually better than average (to counterbalance any that are worse than average).

 

[Xavier6162]

I have not seen the results. If the information is collected via internet forums then it is corrupt. The average owner is not an enthusiast and will not post to forums. The majority have no problems. The problems that are there are always blow out of proportion…except maybe for the High Pressure Fuel Pump issues with the 2007 BMW 3 series.

I’m just trying to find what genesisowners problems have been exploited on Genesis Forums.

 

[Phil_Collins]

Are you cross shopping versus a used 335i? The price of the cars new is not even close. I love the looks of the 335i and the car is very tuner friendly but I have to concur that it is not a car I would like to own past its warranty period. I have a good friend with an 07 335i. The car has been to the dealer several times for a wastegate rattle that the dealer is unable to fix. The wastegates have been replaced under warranty already once and they still rattle. His car sounds like an 80s diesel VW. The dealer is telling him that there is nothing they can do as they already performed the newest fix. No fuel pump problems so far and the car has always been bone stock.
I think for the money the Genesis is the way to go. Based on the reliability for the latest Hyundais I would say 05 and up the car should be fairly bullet proof in the long run. Also keep in mind that the average person keeps their cars for something like 3 years max before they sell/trade for another vehicle.

 

[Xavier6162]

Yes I am cross shopping with a used 335i 2008 and up. If I go that way I’ll go with a 2009 Certified Per owned 335i next year. The 2007 model year has too many problems. Either way a used 335i coupe is very close or a couple $K more (depending on options) than a new 3.8 Genesis Track. I’ve seen them from $29K to 32K for a 2007 model and $31K-$36K for a 2008. The sedans are slightly cheaper.

A CPO 335i gives the owner up to 6 years and 100,000 of warranty through BMW. Only a new Hyundai has a better warranty. I have a few months and will look to see what happens to the Gen'Coupe resale and what plans aftermarket tuners have for the genesisowners.

The 335i puts out 300 ft pounds of torque at 1400 rpm. If the Gen-Coup did that then I would not cross shop nor care what tuners are up to. I've driven them back-to-back and the genesis performance is not comparable to the 335i.

 

[mkaresh]

I have not seen the results. If the information is collected via internet forums then it is corrupt. The average owner is not an enthusiast and will not post to forums. The majority have no problems. The problems that are there are always blow out of proportion…except maybe for the High Pressure Fuel Pump issues with the 2007 BMW 3 series.

I’m just trying to find what genesisowners problems have been exploited on Genesis Forums.

Actually, we're seeing far fewer HPFP failures in our survey than the forums and forum-based surveys would suggest. It helps to have a structured survey process that only collects data going forward.

 

[ImInPA]

I question that data. There are too many versions of the current 3 that aggregating them all and spitting out a number is only going to give you an average. I mean do you think the maintenance on a M3 is going to run parallel with a 328i? What about the 335d, is this data just for North America or is it factoring all the Diesel and petrol version in Europe as well.

I too have problems with the "statistics" in this site. Not very sophisticated, and, the data gathering mechanism does not even attempt to provide random observations (critical for this type of analyisis). Most of the users have been contacted via postings on other forums, and, it is my belief that most forum populations are somewhat biased to specific problems.

______________________________

Help support this site so it can continue supporting you!

 

[mkaresh]

Before I reply, might I suggest that anyone who wants to continue this discussion do so in the thread about my research, which can be found in this forum. We're veering OT here.

A quick response to the above post:

The methods we use are far more sophisticated than you seem to realize. You cannot tell what methods are involved in producing the results simply by looking at the results. So why assume you can?

One key part of our method: unlike other reliability surveys, we effectively randomize the data by covering the enrolled cars beginning with the time they sign up, not retrospectively. The past problems that might lead someone to sign up are generally excluded from the analysis.

"Common problems" are often far less common in our data than forum discussions suggest. Not only this, but there are quite a few models for which very few owners have reported problems.

Over 85% of site visitors come from search engines. Forums are responsible for less than 10% of our traffic. You can't tell where our traffic comes from by simply looking at the site. So why assume you can?

There is no perfect way to gather such information in the real world. But the methods we use are as good as any currently used, and better than most.

 

[ImInPA]

You've made a number of untenable assumptions.

To begin with, the methods we use are far more sophisticated than you seem to realize. Not surprising, since you cannot tell what methods are involved in producing the results simply by looking at the results. But why assume that you can tell?

One key part of our method: unlike other reliability surveys, we effectively randomize the data by covering the enrolled cars beginning with the time they sign up, not retrospectively. The past problems that might lead someone to sign up are generally excluded from the analysis.

Your beliefs are not backed up by our data. In fact, I've typically found that that problems commonly discussed on forums are far less common in our data. Not only this, but there are quite a few models for which very few owners have reported problems. Not surprisingly, these models also score well in other reliability surveys.

Also, over 85% of site visitors come from search engines. Forums are responsible for less than 10% of our traffic. You can't tell this by simply looking at the site. By why assume that you do know where participants come from?

There is no perfect way to gather such information in the real world. But the methods we use are as good as any currently used, and better than most.

I notice you on many different forums soliciting owners of various brands to come and join your surveys. If your sampling method is tainted, it does not matter if you randomize non-random data from a questionable sample size. I do know that among my co-workers and friends, they all seem to share the same concerns about your methodologies and survey collection practices. With regards to the black and red dots from a certain consumer organization, not only do they conduct surveys, but, they tend to actually purchase vehicles and test them themselves. I have issues with their methodologies too as they tend to only survey from within their own membership. A couple of the best vehicles I have ever owned were not recommended by them. As Mark Twain once said, "There are three types of lies: Lies, Damn Lies, and Statistics."

 

[mkaresh]

Where do you work? I find it intriguing that so many people there have discussed my survey.

CR gets all of its reliability data from surveys. The vehicles they buy are used for road test evaluations, and have zero impact on their reliability stats. So they're irrelevant.

No one (except the manufacturer, and they're not talking) has a truly random source of data. So if that's what you want, then you get no reliability information, from any source. Luckily, in practice a random sample is not necessary to measure repair frequencies if you:

1. Don't try to split hairs.

2. Minimize the impact of subjectivity.