My father in law has a GV70 with the 2.5T engine and I've been on trips a few times with him. Its certainly a nice car, but I didn't find it "better" than the NX, comparable would be my description. Actually didn't feel as smooth as my NX turbo (but maybe it was his driving, ha!) and I've driven and been in a colleague's NX350h a few times and I know those are even smoother. He had a Kia Sorento back a few years ago and liked that as well. In between he had a top trim Acura RDX but he wasn't very impressed with it (liked my NX more) and ditched it for the GV70 after just a couple years. It really didn't give him issues, its just that he found it noisy and a bit unrefined even with ALL the toys. Its nice to see Genesis establishing a better rep, certainly the Koreans have some very unique but still very pleasing while not too over the top styling, but the Korean's long term reliability reputation still needs some time for additional mending I think. Hopefully will continue to improve with better cars today. For my father in law, he gets rid of his cars usually after 4 or 5 yrs, so not something he has ever had to care about. He's definitely happy with the GV70. For me, 10-12 yrs has been my typical ownership duration with Lexus/Toyota (used to be 3 to 5 yrs with various GM and Chrysler because, well, you know...) I'm kinda stickin' with what I've experienced that works even with some interesting competition.

But that doesn't count my 71 Toyota Corona I had in high school. That was such an interesting and fun little car but it was already in its end days when I first got it. A friend I had going back to those times just earlier this week sent me an ad for a really good condition 72 Corolla he came across (very similar to the Corona) for sale near us. I was tempted... but not yet quite ready for a "pet" car. LOL.

 

How does your theory account for all the cars in CR that are rated as problematic, poor, very poor, etc. if, as you assert, "Most Consumer Reports survey respondents are happy customers"?

 

I’ve had my 2025 NX 350h for less than a month and it only has about 180 miles. Will have to see about mpg once it’s broken in and I don’t sit in the garage figuring out all the new-to-me. It’s my first Lexus. I’m going to have to look up the 12v issue mentioned above.

 

Agree 💯 %. Simple logic tells us more people will be willing to fill out a survey if they have problems than if they love the vehicle. Frustration and anger is much more a motivator than pleasure when wanting to share information. Ask any manager of a restaurant do they get more complaints or praises from customers.

 

Don't think that’s an issue anymore. Mine has sat for two weeks twice over the summer. No issues.

 

2025 350h made in Japan. 7 weeks and 700 miles. Zero issues. Love it.

 

They really, really like the NX so they respond but they got one or two burs in their saddle that they don't think are representative of Lexus ideals and they air their gripes.

YMMV,
MidCow3

 

MC--we need some data. Not just an anti-CR bias. You've peddled this perspective, citing survey research and polling research as things you reject repeatedly. That's fine as an opinion but you bring a lot of certainty to your posts. Evidence. Data. That's what is needed to refute the evidence and data provided by established sources like CR and JD Powers, etc. YMMV. And Peace brother.

 

Basically marketing, statistics, and surveys do not provide accurate data. I have nothing against Consumer Reports, JD Powers or any other polling and research data. However, you have to understand from a pure mathematical and statistical standpoint you are never going to get an accurate sampling of the desired information. While marketing and sampling companies may provide an indicator of what actual data consist of you can never depend on the outcome with any degree of certainty, especially not to the the degree the marketing and sampling companies purport. I guess my bias is from a strong statistical and engineering background working with detail computer data. Analysis of data are only as strong as the accuracy of the data and how representative the data set is of the whole. Peace to all.

YMMV,
MidCow3

 

Consumer reports record is unimpeachable. They don’t just use surveys, they actually buy the vehicles and other things they test, unlike most organizations. Good way to make sure the manufacturer hasn’t gone over the product beforehand to fix any flaws the average consumer would receive.

• The score in their chart is their road test score. They actually purchase a car and long-term test it (usually a year) amongst their pool of drivers, and then the review and score are published.
  
• The "Predicted Reliability" is based on surveys that CR sends out to subscribed owners, asking about all sorts of stuff with the car. If you subscribe, chances are you care about and pay attention to quality and value more than the average person, so they and their survey responders have always weighed stuff like build quality, NVH, comfort, etc more heavily than technology and performance in their testing. The respondents are aware of CR’s mission, and are far more likely to give relevant responses. They are engaged and are the preferred pool to sample as a consumer.
  

A car has to earn a certain road test score and have above-average predicted reliability or better to earn their "recommended" stamp. But there are plenty of instances of cars that aren't recommended that get the highest road test scores in their category. Corvette has always been one of those.

They’re also a nonprofit, their goal is accurate information, not profit, they have no ads, and zero manufacturer influence. The latter is the most important difference between them and other car reviewers.

 

The problem I see here is the inability to see research through more than one lens, i.e. "pure mathematical and statistical standpoint." Leaving aside how often statisticians, mathematicians, and computer scientists disagree about their own methodologies and conclusions, applying your need for absolute certainty in all fields outside those you value would paralyze our species' ability to reach conclusions and make decisions. Because in the real world conclusions and decisions can't wait for the absolutists to give us their version of certainty. By the logic you employ MC, we would need to discard almost all the research done in the social and human sciences. Not to mention all the contributions made to civilizations (and to the thinking of the sciences) by philosophy. The field of law could not function if the only standard applied was absolute certainty in all things. Same for military science. On and on. The need for certainty and absolutes often leads to absolutism and absolutists. Which often give us authoritarians. Some people cannot cope with the realities of life's ambiguities. But that doesn't make ambiguity the problem. Pretending we can only get valid answers and data from one mindset is. XLR8TOR does an excellent job of underscoring how CR does its work. Rocket scientists are pretty good at rocket science (though they can't give us rockets that will always perform as hoped). But we would not want those who only know their field of science to resolve political issues in the real world, for example. I believe it was Einstein (a pretty good scientist) who said "Not everything that can be counted, counts. And not everything that counts, can be counted."

 

My best confirmation for J D Power 24 out of 28 years #1 in dependability and Consumer Reports #1 in reliability is have had six Lexus's and zero problems. Friends and family with also Toyotas with dozens same story. The other side have friends owning the bottom of the list like Land Rover Volvo Mercedes etc. and almost everyone multiple problems.

 

Yes, Consumer Reports does specific testing on products which is very good. However they still use a lot of consumer surveys based on users who respond. That is not a representative set of all people.

@Romer an accurate set of results which is then definable as a representative answer, is based on a random set across all users which is representative of the entire data spectrum you are trying to analyse. This applies to just about everything from cars to rockets, to product evaluation, etc. Car owners replying to a CR survey or a JD Survey are NOT a random set of users. Most of these users are new car buyers and typically respond like their cars. To get a truly random set you need to send out surveys to a random population that includes car owners new and old, and non-car owners. Then you would get fairly accurate results. Looking at surveys from new car owners who subscribe to CR is not a random set and therefore is almost meaningless because of the bias.

Here is a great book layman on statistics if you haven't read it I would at least recommend skimming over: How to Lie With Statistics book by Darrell Huff

YMMV,
MidCow3

 

Alas we've arrived in the land of OZ. Yikes.

 

J D Power surveys owners of three year old vehicles with their dependability study hardly new owners.