Hey Scotty,
All my past, current, and future cars (imminent purchase coming) I have never once looked at the crash safety scores/ratings or even paid much attention to safety during the car selection process. I also figured most cars are safer these days than a long time ago, so you can’t go too wrong. Anyways, are car safety scores/ratings overrated and am I being stupid for ignoring safety?
With modern vehicles unless you get something that's really a tin can like the Mitsubishi Mirage you should be OK, at least in the U.S.
The test results that are worth paying attention to are the IIHS and EuroNCAP.
These ratings companies are a joke. Kind of reminds me J.D Power.
First of all Euro NCAP sucks . Data simply doesn't exist for most cars. They don't even have data for popular cars like Honda Accord and Toyota Camry. Zero pickup trucks. There are mostly Euro models so it's useless to N.Americans.
I picked a few annual reports at random, to see how the ratings translate to the real world...
- The deadliest car in 2019/2020 was Nissan Altima (113 driver deaths per million registered vehicle, 94 drivers in other vehicles). IIHS awarded this car "Top Safety Pick". It got an almost perfect score.
- The 3rd deadliest car was the 2019/2020 Kia Optima. Awarded a golden "Top Safety Pick+" badge. (the '+' is the highest award)
- #5 deadliest car was the 2018-2020 Toyota Camry. Another golden "Top Safety Pick+" winner
What about Euro NCAP .... I managed to find a few:
2011 deadliest cars ....
- #3 - Chevrolet Aveo ... 5 stars out of 5 for safety
- #7 - Chevrolet Malibu ... 5 stars out of 5 for safety
2017 deadliest car - Ford Fiesta .... 5/5
I think I've seen enough about "ratings"
one of my main considerations when choosing a vehicle is survivability in case I get into a crash
"Main" consideration ... really?
Isn't that like choosing a hammer "mainly" based on how it feels when it hits my thumb?
Yes, one of my main considerations is not dying when someone T-bones me or switches lanes into me.
I care more about my physical wellbeing more than paying a few extra pennies on gas or paying more to service it.
I owe my family to return home safely, how can it NOT be one of mine main considerations!?!?
Risk assessment.
I don't go through life basing my decisions on the assumption that the worst will happen. I don't get into accidents a lot, so it's not worth devoting a lot of resources to.
I illustrated that with the hammer example. The main considering when buying a hammer, is how it hammers. That's it's primary function ... the reason why I buy it. If it can't do it's job then it's useless. Safety is just so far down on "likely to happen" scale, that it hardly registers.
Similarly, a car needs fulfill it's car functions first and foremost. Safety comes after. I think you know this very well, and you're just exaggerating.
Not that it's completely absent. I would never drive a Smart car for example. But these ratings mean nothing to me.
depends where you live I guess.
Densely populated city with heavy traffic ...
Lots of idiot drivers, and a high frequency of collisions ...
yeah then I would probably think about it a little bit.
I live in the suburbs. Most of my driving is on highway to and from work and to very remote places.
.
I don’t have to go to the cities, thank goodness.
Up to the mid-2000s, there were a lot more steel-based cars on the road. My Catalina was only around 25 years old at the time. I saw Chrysler Town & Country minivans being sold via advertising at the time, boasting they were rated a top safety pick. Stow & Go seating was all the rave. Airbags were also a relatively new thing in the mid-90s. Now, they have curtain airbags, in addition to standard airbags in the steering wheel and dashboard. These days, most steel-based cars are either non-existent, or they are few and far between. The other thing is people have grown accustomed to safety, only the actual body of most cars are made of metal now, even trucks. You don't see chromed steel bumpers anymore. It's all plastic. My Ranger is a 1999, the bumpers are made of metal, they stopped making chromed steel bumpers sometime in the mid 2000s.
Although some are much better than others today, they're all better than they were in the old days. I'll bet size still matters though - you can't change the laws of physics.
https://www.youtube.com/watch?v=uyjdmxh1MKU
These ratings companies are a joke
They are not, neither is a "company".
One is an EU government agency, the other is a nonprofit organization.
The deadliest car in 2019/2020 was Nissan Altima (113 driver deaths per million registered vehicle,...)
Nope. This is wrong.
The worst are:
Mitsubishi Mirage G4 - 205
Mitsubishi Mirage hatchback - 183
Dodge Challenger 2WD - 154
Source: https://www.iihs.org/ratings/driver-death-rates-by-make-and-model
See note further on expired test results.
3rd deadliest car was the 2019/2020 Kia Optima
Nope. This is wrong.
The Kia Optima has got a deathrate of 80,
which is the same as the exact same as Chevrolet Impala and like 2 more than a Honda Civic hatchback.
And this car has got 89% in the 2015 Euroncap test which does demonstrate that it's not horrible, but not good.
Source: https://www.iihs.org/ratings/driver-death-rates-by-make-and-model
#5 deadliest car was the 2018-2020 Toyota Camry
Nope. This is wrong too.
The Camry has got a deathrate of 48, it's factually on the safer side.
Source: https://www.iihs.org/ratings/driver-death-rates-by-make-and-model
2011 deadliest cars ....
- #3 - Chevrolet Aveo ... 5 stars out of 5 for safety
Nope. This is wrong too.
Looking at US statistics the worst car wasn't the Aveo it was the 3rd worst was the 2011 Hyundai Accent.
Source: https://www.iihs.org/api/datastoredocument/status-report/pdf/50/1
The Hyundai Accent had HORRIBLE performance on the IIHS.
https://www.iihs.org/ratings/vehicle/Hyundai/accent-4-door-sedan/2011
You see how these are linked? It's almost as if the link has been scientifically verified (cause it was).
Now, overall the Aveo is arguably a dangerous car with a score of 99 deaths,
YET, it is significantly lower than the 115 average for the 4 door mini car category in 2011!
The 2011 EuroNCAP test checks how well the car performs in SV scenarios,
and the Aveo performed well in that aspect!
So now that we are clear on the fact the Aveo performs decently,
We also should that the tests are constantly evolving, a 5 star car from 2013 (89%) was retested in 2023 according to the new 2023 requirements and it has scored 0 stars (43%) - So needless to say a 5 star rating in 2011 is nothing by modern car standards - this is why the safety ratings EXPIRE, and a 5 star score that's expired is worthless.
2011 deadliest cars ....
(...)
#7 - Chevrolet Malibu ... 5 stars out of 5 for safety
Nope. This is wrong too.
The 2011 Malibu had 41 deaths per (...), which makes it pretty safe.
Source: https://www.iihs.org/api/datastoredocument/status-report/pdf/50/1
I think I've seen enough
I think you need to go over your data and see again.
Your statements have been verified and DISPROVEN scientifically many times.
https://pubmed.ncbi.nlm.nih.gov/21128188/
"Good correlation was found between Euro NCAP test results and real-world injury outcomes. The largest difference in injury risk between 2- and 5-star rated cars in Euro NCAP was found for risk of fatality, confirming that car manufacturers have focused their safety performance on serious crash outcomes"
And as far as the IIHS test categories, the side category has been also was scientifically verified.
https://www.tandfonline.com/doi/abs/10.1080/15389588.2022.2050223
"Because performance in measures collected in the IIHS side crash test are strongly associated with real-world driver death risk, one of the ways the crash test program could continue to encourage crashworthiness improvements is by requiring stronger performance on these measures."
I'm glad we cleared this up and shown how test results do DIRECTLY correlate with safety.
They are not, neither is a "company".
organisation ... whatever. It doesn't change anything.
The data was analyzed by S&P Global, by the way, who is a for-profit company.
Source: https://www.iihs.org/ratings/driver-death-rates-by-make-and-model
That's exactly where I went for the data . And the only source of summary data I could find, unless you want to spend months compiling it from the NHTSA database.
The first think you see:
The deadliest car in 2019/2020 was Nissan Altima (113 driver deaths per million registered vehicle,...)
Nope. This is wrong.
It's the worst normal size car.
and on a scale from zero to Mitsubishi soda can, it's bad.
The worst are:
Mitsubishi Mirage G4 - 205
Mitsubishi Mirage hatchback - 183
It's a micro car. Of course it's unsafe. I don't consider it a normal car. You wouldn't buy one.
Size is the biggest safety factor (and IIHS report this)
Despite manufacturers’ efforts to make them safer, the smallest late-model cars remain the most dangerous, according to the most recent IIHS study of driver death rates. Small cars and minicars accounted for 15 of the 20 models with the highest death rates for model year 2017 ... “Smaller vehicles offer less protection for the driver in crashes, and their lighter mass means that they take the brunt of collisions with larger vehicles,” says Joe Nolan, IIHS senior vice president of vehicle research.
And yet, IIHS still rated those death traps good overall ...
Nope. This is wrong.
The Kia Optima has got a deathrate of 80,
3rd worst car in its class ... yes it is.
Overall , still in the worst 20 And it received IIH's top award, which is the point you're avoiding.
(also it's the 5th worst car for killing other drivers)
which is the same as the exact same as Chevrolet Impala and like 2 more than a Honda Civic hatchback.
OK. And... ? The Civic is a small car by the way.
And this car has got 89% in the 2015 Euroncap test which does demonstrate that it's not horrible, but not good.
which further illustrates my point that these ratings are useless.
Nope. This is wrong too.
The Camry has got a deathrate of 48, it's factually on the safer side.
You are wrong. 5th worst in it's class. AND, you should add it together with the hybrid because it's the same body. (bringing it to 67, tied 4th with Fusion)
2011 deadliest cars ....
- #3 - Chevrolet Aveo ... 5 stars out of 5 for safety
Nope. This is wrong too.
I stand corrected. I misread the data. It came in 3rd worst in IIHS safety testing.
It's #4 in death rates. But wait ... does that make me wrong ... no.
In fact if you keep reading there's ANOTHER Aveo in the list! So with another 58, that brings the rate up to 157, making it the WORST car.
Another mini car ... shocking
And all of this is in stark contrast to the Euro testing, which was my point, that you again ignored.
Whether they "expire" it or not after the fact doesn't matter. Ratings are ratings.
#7 - Chevrolet Malibu ... 5 stars out of 5 for safety
Nope. This is wrong too.
Again, 7th worst in IIHS safety testing.
3rd worst death rate in it's class
and unsurprisingly, 2nd worst in class in 2020 (15th overall)
When you consider how many models there are, this car does not do well.
5/5 stars 👍
I think you need to go over your data and see again.
I think you need to see the bigger picture.
I really though you're gonna answer with solid info but this all seems wrong.
Not wrong. You just focus on insignificant details and completely miss the point.
"Good correlation was found between Euro NCAP test results and real-world injury outcomes
again ... what results? They're missing for most cars.
And from what I've seen, they're terribly unreliable.
"Because performance in measures collected in the IIHS side crash test are strongly associated with real-world driver death risk,
METHOD: " Logistic regression was used to estimate the association between crash test measurements and death risk, controlling for driver and vehicle information."
regression = guessing
so overall, they kind of get it right, maybe.
Can you trust individual ratings for vehicles ... it doesn't look that way to me.
Oh and look ... the study was funded by the IIHS, so they are congratulating themselves. Nice.
At least show me independent research.
Your statements have been verified and DISPROVEN scientifically many times.
A paper is a just an argument, like you're making now. Not proof.
There's no shortage of bad papers.
The chances of a catastrophic accident are low, yet I'd rather have a no wireless CarPlay than risk get stuck in a wheelchair or not come back home to my family. as simple as that really.
I drive a 2000 Acura Integra coupe every day. Chuck Tobias doesn’t even own a vehicle from this century. Look at Scotty’s Toyota Celica. All those cars would do terrible in modern crash safety tests. Yet, I am not concerned. What would you say about that as well as folks like us that keep our vehicles for a long time?
.
All our days are numbered. When it’s time to go, that’s it. I don’t worry about it.
Chuck Tobias doesn’t even own a vehicle from this century.
My primary ride is a 1997 Saab 9000, and that company was always ahead of the curve when it came to safety. No, it's not going to be as good as a new model with the latest technology but I'm satisfied it is safe enough.
https://www.theautochannel.com/news/date/19961008/news02157.html
In fact @dan asked about what if you have a catastrophic accident if something runs a red light and hits you doing about 60 miles per hour. About 20 years ago an accident like that happened to me, the other vehicle being a large SUV and I was driving a 1986 Saab 900 at the time - a car whose basic chassis design dated to the 1960s. My car was twisted into a pretzel, except the passenger compartment remained basically intact. The force of the impact was diverted around the passenger space as it was designed to do. (The SUV only had a dented grille and fender - size matters!) I just opened the door and walked away with no injuries. The car of course was a total loss.
The next day I bought another Saab to replace it. I don't think the outcome would have been the same in a similar year Corolla or Civic, though of course those cars require much less maintenance to keep running. (Today's models are a different matter I'm sure, much safer than they were decades ago.) So at least to a certain extent it depends on what old car you're driving.
I personally wait a couple seconds before proceeding through green lights, now. 3 years ago, somebody blew a red light going 45-50 mph and crashed into the SUV right in front of me. It rolled right over and onto its top. Now, if I'm the first one at a red light, if there's more than one lane in each direction, I pause to look before I proceed, when it turns green. You never know anymore, way too many drivers are distracted by useless things while in a moving car.
Same here. Though the way that particular intersection was laid out you couldn't really see a speeding vehicle that was going to run the red light until you were moving through.
The cop who came on the scene was pretty amazed. He kept looking at the car and at me and said incredulously "You walked out of THAT?"
I do the same at red lights. I guess great minds think alike!
In July I was driving through a green light, when suddenly a Subaru XV run a red and crashed into the Mitsubishi Lancer driving behind me - the person in the Lancer got crushed by the dashboard coming into the passenger compartment (I think he survived).
While the Subaru XV received moderate damage, to my eye, it wasn't even totaled.
-
There's only a 150 pond difference in kerb weight between the two, so it's not weight or size.
The difference is the crap safety structures put in place by Mitsubishi and DaimlerChrysler, compared to what Subaru has designed.
And this is reflected in safety scores, the Lancer got the low score of 81% by the expired 2009 scale, (equivalent to ~50% today) while that Subaru got a very safe 94%. (by the 2017 scale)
-
Y'know it's not always size or weight.
Here's a head-on crash experiment between a Renault budget micro car and a heavy Volvo 940 station wagon.
https://youtu.be/emtLLvXrrFs?t=310
That Renault devastated the old Volvo.
Sometimes it's use of quality metals and design, and not just size and weight.
Cars today have gotten much safer than what they were years ago. You don't see people fly through the windshield from a front-end collision (my father has told me stories of what he's seen before). Also, when I got rear ended in my Town and Country, I walked away with zero injuries (and I didn't even have a seat belt on).
I have never once looked at the crash safety scores/ratings or even paid much attention to safety during the car selection process
Why?
This feels odd to me, one of my main considerations when choosing a vehicle is survivability in case I get into a crash - driving is a dangerous activity, so why not minimize risks.
I also figured most cars are safer these days than a long time ago, so you can’t go too wrong.
There's a huge range between really unsafe and safe cars.
Anyways, are car safety scores/ratings overrated
Absolutely not, crash test results from a reputable source a very important indicator of car safety.
NHTSA's crash tests are absolute crap, outdated, results are not useable - it's shameful how bad they are.
The test results that are worth paying attention to are the IIHS and EuroNCAP.
EuroNCAP simulates the outcomes of the most common fatal crashes.
- Head-on collision at a combined speed of 62mph with a 3,000 lbs car (deformable trolly)
- Collision with a wall / not a deformable object at a speed of 31 mph.
- T-Bone crash simulation at 37.5 mph with a 3,000 lbs car (deformable trolly).
- ( + The test includes assessment of kinematics for the other passenger too)
- Side impact with a tree/pole at 20 mph.
- The test also includes checking if the seats offer adequate protection against whiplash.
Additionally, the test includes results for:
- Child occupant safety in a side impact
- Pedestrian head, torso, pelvis impact
- Pedestrian autonomous safety systems (AEB, "Dooring prevention")
- Autonomous safety assistance systems (AEB Car-to-Car, Lane support, speed sign detection)
- Motorcyclists safety
This test program offers a progressive score which is a fair assessment of survivability in case of an impact.
All of these results are compiled into scores and categories,
together these scores are a comprehensive and conclusive assessment of crash safety.
IIHS adds valuable information on additional test categories, such as clipping a rigid object, side impact against a very heavier vehicle, Roof strength, and even headlight performance.
Interestingly, the most dangerous cars (Driver death rates by make and model, IIHS) usually do not enter the voluntary EuroNCAP test (or have an extremely outdated raiting from a decade ago) - so not having an IIHS score but still being sold in that region is a big red flag.
This years (so far) best performer is the Nio ET5:
The previous year's was a tie between the Model Y and iONIQ 6:
With a weaker but still decent result from Toyota
I don't get into accidents a lot, so it's not worth devoting a lot of resources to.
you're just exaggerating
I didn't say I do not care about other things: I just said that I would obviously prefer a higher rated car, and that I would not buy a car that performed poorly - I care about it about as much as I care about fuel economy (slightly less than reliability).
That's it's primary function
So the primary function of a phone is phone calls.
Yet I also need my phone to have a large display, have good app availability, a good camera.
There are lots of phones that can make great quality phone calls,
yet obviously buyers have other considerations and requirements that are critical.
The primary function is part of it,
And of course no one would buy a safe car that doesn't drive - yet, to of the cars that perform their primary function, I definitely wouldn't choose the one that failed it's safety tests.
As an example, a car I considered has poor frontal crash performance (week protection of the drivers chest)
Hence I don't want it.
Safety is just so far down on "likely to happen"
That's exactly where I went for the data . And the only source
If you press on the dropdown menu, more options will appear.
You are looking at midsize cars only.
The first think you see:
Exactly the first thing you see, is only midsize cars.
And yet, IIHS still rated those death traps good overall ..
It literally got the 2nd worst score in the main category,
and also refer to my post, I clearly said "IIHS adds valuable information on additional test categories".
Again, 7th worst
I have disproven this, but lets see again.
Nope, still not on the 2011 chart for the worst cars.
And lets see the EuroNCAP results - 94% - room for improvement but quite safe.
Also, just for general information note the margins of confidence around the observed value,
The statistic analysts has shown that there's a 95% chance it could be as low as 26, and as high as 56.
In other words, with such margins, it can be actually even better - we just lack data.
Also the actual 7th worst is the Silverado Crew Cab - and it performed poorly on the IIHS test.
It has received a not-good score in 3 out of 4 categories.
and unsurprisingly, 2nd worst in class in 2020
That's a different car on a different platform.
what does that have to do with anything?
I think you need to see the bigger picture.
That actual studies were conducted by statisticians and I am proven right?
Cars with poor results tend to be deathtraps and that they aren't worth buying.
So of course it's reasonable to verify that the car you're looking at ain't a deathtrap.
again ... what results? They're missing for most cars.
Yes, the North America region does not have a good new car assessment program.
Not great for buyers of cars offered only on the US market - for the rest that buy compact asian cars it's fine.
This is solid additional info from the EU gov, and it's free - what more can I ask for?
they're terribly unreliable
I'm yet to see a car with a non-expired raiting that gets a good score and yet fails irl - but this is minor and I'm sure there are edge cases.
What I'm concerned about is the opposite - cars that perform poorly always turn into deathtraps.
So far the examples you gave turned out to show a somewhat different picture...
And reading the studies - the results are that cars with a higher safer rating are safer beyond margins!
regression = guessing
When I studies my degree we did lots of regression and it is NOT guessing.
It's a statistics analysis method, aka linear regression. (usually, simple linear regression)
Basically it's a method for finding a function that expresses the relation between two variables.
(X - safety score, Y - deadly crashes - the regression method will give us a function that relation and we then can judge how strong the relation is)
so overall
Overall they demonstrated a mathematical relation using accepted statistical methods.
A paper is a just an argument, like you're making now. Not proof.
They have shown proof, you can disprove their work if you want.
There are open articles, like those who verified that the full frontal crash test by EuroNCAP does generally match the behaviors you see in actual crashes: https://www.tandfonline.com/doi/full/10.1080/15389588.2019.1597348 (free access)
Oh and look ... the study was funded by the IIHS, so they are congratulating themselves
Not really, they shown with statistics that although some dispute the accuracy of the side impact test it still is an indicator of safety.
Again, if you have an argument for why their analysis is wrong, or evidence they're wrong - show me.
There's no shortage of bad papers
There's a shortage of papers who claim that crash test results are crap.
(Note: There are papers that offer critique of the methodologies of some tests, particularly of the IIHS, but these do not claim that the scores and real world results don't correlate.
Also, to the IIHS' credit, after critique they updated the test.)