Is Tesla Fabricating Data To Make Its Model 3 Appear Safest?


After Tesla reveals the ultimate safety of its Model 3, the NHTSA clarifies its system and intentions, but it may make no difference.

Tesla stated this week that its Model 3 earned the National Highway Traffic Safety Administration’s lowest probability of injury score ever recorded and went on to say that its Model S and Model X vehicles hold the second and third positions of all time. How come the NHTSA and many mainstream news agencies didn’t report such amazing achievements?

Here’s the rub. The NHTSA is not a ranking agency. Never has the administration provided a list of cars with a list of scores and said that one particular five-star candidate exceeds another. In a recent attempt to assure to people that the NHTSA had no part in Tesla’s claims that its vehicles outscore all others, the NHTSA released a clarifying statement (keep in mind that Tesla was not mentioned in the statement, although many publications called this a direct rejection of the automaker’s claims). According to the organization:

National Highway Traffic Safety Administration issues statement about New Car Assessment Program’s highest rating

The National Highway Traffic Safety Administration’s New Car Assessment Program (NCAP) – known as the 5-Star Safety Ratings Program – provides consumers with information about the crash protection and rollover safety of new vehicles beyond what is required by Federal law. NCAP conducts a total of three crash tests on new vehicles: one frontal and two side crash tests, as well as a rollover resistance assessment – a driving maneuver test that assesses a vehicle’s susceptibility to tipping up and a measurement of how top-heavy a vehicle is. Results from these three crash tests and the rollover resistance assessments are weighted and combined into an overall safety rating. A 5-star rating is the highest safety rating a vehicle can achieve. NHTSA does not distinguish safety performance beyond that rating, thus there is no “safest” vehicle among those vehicles achieving 5-star ratings.

Tesla CEO Elon Musk assumed for some time that the Model 3 would be one of the safest cars ever produced. Even before any outside testing, he said it would have the best safety of any midsize car. After the recent perfect five-star rating, he was even more confident and promised to share injury probability stats once they were released.

The thing is, as you can glean from the above statement, the NHTSA doesn’t provide specific numbers to the general public, but it also doesn’t hide them. In order to determine a vehicle’s star rating, meticulous calculations must be applied. The organization doesn’t simply test the cars and then do a show of hands for star ratings. Instead, there are figures that are calculated and specific cutoff points for each star rating in each individual category and overall.

Thanks to Twitter user @anonyx10, we were directed to the website containing NHTSA charts with crash test data. Below is a screenshot of a partial chart, which includes the Tesla Model 3. Below that is a series of tweets better explaining the situation, as well as another chart to supplement the one provided in our previous article.

Let us know your thoughts and assessment in the comment section below.

Categories: Crashed EVs, Tesla

Tags: , , , , ,

Leave a Reply

52 Comments on "Is Tesla Fabricating Data To Make Its Model 3 Appear Safest?"

newest oldest most voted

The NHTSA data does not provide the “precision” or error margins to decisively assert that a particular model is superior to the other though. Raw data,

We also have to remember why IIHS started in the first place: insurance companies were not satisfied that NHTSA tests adequately represented safety of vehicles (they are paying in these accidents so have vested interest) . They wanted to take it a step beyond what NHTSA was testing as most real world accidents are not head on, but offset.

That being said, these are incredibly safe vehicles, as are many others. I am sure Tesla Model 3 will do well in IIHS testing too, although the Model S didn’t perform great in small overlap.

“as are many others”
I don’t know, there’s a lot of models scoring 3 stars ( Jeep ).

Doesn’t the IIHS not have a comprehensive crash test rating?? They test specific things which are good to know individually but don’t provide a clear picture as to a car’s overall safety. No, the insurance companies are trying to get automakers to work on specific things that worry them, one such example being pedestrian deaths at night. Imagine how expensive that claim would be for an insurance company. That’s literally the entire point of the headlight test.

I think a different kind of bias would result from that, and provide the wrong kind of incentive. People being crippled rather than killed is generally much more expensive – whether they are the driver, a passenger, in another car, a pedestrian, or something else.

The NHTSA is testing the physics of crashes with a uniform test method.
Unless there’s some massive difference from one Tesla Model 3 rolling off the assembly line to another, and with Robotic Welding, that’s highly unlikely, these numbers are valid.

The real world will add, of course, a different subset of drivers for each car. But if you’re a Volvo buyer, you can buy a Tesla with confidence your accident experience in a Volvo will be the same in a Tesla, and an actual crash will be better.
Along with Tesla having advanced accident avoidance systems.

Sure, if your real world crash exactly replicates the NHTSA test: driving into a wall head on at 35 mph. Or striking a pole which conveniently lines up exactly with the Tesla’s roof cross bar. Again, why IIHS started doing offset collisions.

To be fair, NHTSA has improved their tests. And I guess Tesla owners might be concerned with full width wall tests with how many I see accidentally pressing the accelerator and driving into walls of buildings 😉 (they have a lot of power so accidentally pressing the accelerator ends that way).

Don’t get me wrong, the Tesla is safe, but so are many others (I am not talking Jeep, look at the other 5 star vehicles, there are a lot). If it comes out on top of Euro NCAP and IIHS, then sure.

I think Tesla recently introduced provisions to mitigate the effect of pedal misapplication when obstacles are detected in the way?…

You are going to need a big error margin to account for 43% improvement over other cars. If their error rate is that high then the NHTSA scores and data isn’t worth the paper it written or the hard drives.

Tesla distorting facts? This is my shocked face: 8-|

Point of the article says otherwise, shocked one. Tesla got an A+, and they had this moment before. When the Model S was tested, they got “5.4 stars” and NHTSA said to stop saying that. They also ran their tests at higher speeds (I think 55mph versus 35 mph) and still passed. They *also* had to use a rope to pull the vehicle for the roll-over test. They *also* broke the crush-testing machine.

What is going on here is the human forces of hubris pushing back against the triumphant. Given how much glass is in the roof of the Model III, I think this is a testament to the best automotive glass ever! It is also really pretty with raindrops on it. There is some sort of diffraction going on that takes on colors of the rainbow, and it has been an obsession lately with my local Tesla club.


Model S only got an Acceptable rating in the IIHS small overlap test, has Poor rated headlights, and has poorer roof strength and more side intrusion than cars like E-Class and 5-series if you dig into IIHS test data.

It’s safe, but hardly the safest.

Didn’t read the article. Shocked Face!

Well, at least they have facts to distort, you, on the other hand…

Here is the proof that Bro1999 actually can’t read.

The regret at your Bolt purchase is still alive and well.

GM Stealership employee Mental MadBro using distorting, whining FUD against Tesla? I’m shocked!

“Mental MadBro”??? Really? We’re going to stoop to “mental” insults on this site now?

How about this: We take a few moments to pause before replying and try to act like responsible adults. I know, I know — Internet, Wild West, yadda yadda. But give it a shot.

While yours is a good goal to strive for, it is not possible to have a reasoned, adult discussion with trolls. Their purpose for posting comments is precisely to disrupt reasoned, adult discussion.

It seems you haven’t been around here long enough to notice that bro1999 is not interested in acting like a responsible adult — just bashing Tesla (or anything non-GM; but especially Tesla) in the most childish ways possible.

I’m not personally fond of the childish responses either — but they are no worse than the post provoking them.

No distortion. Did you actually read the article? NHTSA doesn’t *want* to rank cars to each other. That doesn’t mean NHTSA data (which is much more granular than 5 stars), can’t be used to rank cars. It most certainly can be used.

You can argue why a government regulatory agency shouldn’t rank products, and instead wants to set basic thresholds, but you can’t argue Tesla is distorting the facts by merely echoing the NHTSA data. They are just shining light on the facts.

Unfortunately, bro1999 is too deep in Tesla hate to even care. He just see’s a trolling opportunity. Too bad, he’d be a greeat ambassador for the Bolt if he was more reasonable on other topics. I love my Bolt. Granted, it’s being traded in for a model 3 next week.

bro1999 claiming Tesla distorted facts when the evidence shows it absolutely did not?

This is my shocked face: 😯

I also recall certain people trying to dismiss the Model S breaking the crush test machine as if “that doesn’t matter”. Well, in the real world, extraordinary results like that actually do matter in real world crashes.

IiHS is more of a mixed bag for Tesla which shows areas where other cars exceed them. About 12-15 years ago there was a side impact IIHS test that plenty of 5 star NHSTA cars did not receive a good rarting for. Today, the front overlap is a new challenge. Botom line – compare IIHS and NCAP.

SurprisIng Mustangs did so well. But where are the other notable vehicles such as: S,X, Volvo, Audi,Mercedes??

Chart is incomplete. Just a sampling.

The Mustang is not a safe car according to EuroNCAP. only two stars in original test in 2017.

“recent seemingly passive-aggressive attempt”

I disagree with that statement. Not sure that the author understands the risk that a government agency would incur by ‘ranking’ any business, or individual. While the executive branch of government may not care, at the bureaucracy level, a department cannot overtly show bias, and they go out of their way to remove a perception of bias as well. Otherwise the department / agency exposes itself to increased risk of litigation. Say from a company that finishes LAST on the list.

NHTSA is simply making it public that Tesla cannot claim it is ‘Safest’ based on NHTSA data, they can advertise that they are the ‘Safest’, but NHTSA can’t be responsible for validating that claim. Important distinction.

Correct. I removed the wording. One of the original sources framed it that way. I agree that it shouldn’t be stated as such. Thank you for catching it!

I agree they shouldn’t provide a ranking of “safest cars”, because their testing isn’t providing a complete picture of the safety of a car. BUT… they already de facto have a ranking, because the numerical data collected in testing is publicly available. Here, for instance, is the test report for the frontal crash test for Model 3: which I found on the ratings page You can download the same report for another car and compare things like the maximum acceleration the neck of the crash test dummy underwent (in g), and duration (in ms). But they don’t do anything to make it easy to compare many data points for many vehicles quickly. A good visualization of the data would have made it enormously more informative. Imagine, for example, if each numerical value was accompanied by a histogram showing the distribution of measured values, the different star rating boundaries, and where the number you’re looking at is on the spectrum. The reader would immediately be able to see if there’s huge variation in that figure, and whether the result is better or worse than most other vehicles, and whether it’s barely five stars or way better. Sometime really should… Read more »

Seems that both NHTSA and Tesla are correct:

NHTSA is correct in that it does publish the NHTSA raw score data used by NHTSA to determine which cars meet the minimum threshold scores to receive NHTSA’s highest safety “5-Star” classification and that NHTSA does not have an *official* safety classification beyond the “5-Star” designation.

Tesla is also correct that the raw NHTSA published data shows that Tesla Model S scored the highest in the raw data metrics.

I can imagine the traditional car makers are giving NHTSA flack over Tesla’s self promotion about Tesla’s ranking #1 (“safest”) in safety… thus the NHTSA clarification news release which basically says “safest” is not an *official* NHTSA designation.


Car makers specifically do not want there to be stack ranked safety numbers. They don’t want to get in a spending war to see who can stay on the top of the list.

Meanwhile, Tesla is trying to debunk the age old myths about all EVs being unsafe by showing they are safer than ICE cars. Just like them debunking EVs being slow by having EVs faster than ICE cars.

It’s a question that we answer. Other publications have said that it is doing so. We aimed to look into it and prove whether it was right or wrong. We asked ourselves that exact question before writing the article. If the entire story is in the headline, people don’t tend to read the article. If we said, InsideEVs believes that Tesla is not fabricating data and all information is verifiable, there isn’t as much need to elaborate in the article. This is 100 percent no attempt to spread FUD, but actually to counteract it. Instead, we hope that the naysayers and negative publications would be drawn to it by seeing the headline and then read it to learn the truth. If every InsideEVs headline is 100-percent Tesla positive, we fall into the opposite issue. Of course, most people would immediately assume a pro-EV website would side with the EV maker, but the question posed sheds light on the fact that other agencies are saying Tesla Fabricated Data Again, and we’re asking … did they? Let’s take a closer looks folks. Sometimes, in order to impact the negative mainstream media and those pushing FUD, the headline is of the utmost importance.… Read more »

It makes sense to me. When I saw the headline on your site, which is familiar to me, I assumed I would be finding a considered answer to the question – a question that has been raised throughout the media by those determined to spread FUD. So I read the article and was not surprised by the answer, thanks.

They’re All Very Safe Cars…However, If you are taking a Hit it’s Safer To Be In the Heavier “S or X ” Car. …..If you are Hitting a Solid Or Heavier Object than yourself , it’s Safer to Hit it In Lighter Model 3 Car For Less Inertia/Impact & You’ll Likely Bounce 0ff…So How do you Plan an Accident ?, it’s 6 of one Half Dozen of the Other Assuming That The Cars Involved are All 5 Star Rated Tesla’s…

Tesla fans should stop being so excited with any small detail :P.
In Europe the first car that really took seriously crash tests (and crash safety) was the Renault Laguna II. It was a big deal as it was the only 5 star rated car when all others were 3 or less. After Renault “all” cars in Europe are “safe” with the exception of cars for other markets or very cheap ones.

I’ve heard more about how Tesla is a little better than others than Renault advertised the fact back then.

My grow with the way you reported it originally was with the headline and article both giving the impression that the NHTSA had concluded Model 3 is the safest car they’ve ever tested, and publishing a “probability of injury” percentage – when both were Tesla’s own claims. Moreover, the basis for these claims was unknown at the time. In fact, the calculation of probability is still obscure, as well as irrelevant. The performance of different vehicles in the tests relative to one another are meaningful, but they are not complete. I’m glad to see an explanation of what the basis for the claims was. But it really would be better if you reported facts (including the fact that TESLA claims this data means X, Y, and Z) and left it to the reader to interpret what they mean. It’s also fine, in fact often desirable, if you do a little critical research yourselves, form an opinion about interpretation, and state it clearly AS SUCH – not as undisputable fact, and certainly not as what the NHTSA said when it didn’t. By the way, the website where all the NHTSA test data is published is the NHTSA’s web site, The… Read more »

It’s a fair point that the NHTSA said it does not approve of the way Tesla is using NHTSA’s data to make a claim beyond the official NHTSA rating for the Model 3.

But I also note that the NHTSA has not actually stated that anything Tesla/Elon said on the subject is wrong. There’s a big difference between saying “That’s not how we do things around here” and saying “What you’re claiming isn’t true.”

Just because the NHTSA doesn’t like the way Tesla is using NHTSA’s test results to advertise the Model 3, doesn’t make any of Tesla’s claims untrue, and it certainly does not indicate Tesla is “fabricating evidence”!

Go Tesla!

Another Euro point of view

I am not saying anything….hope it will be appreciated.

Did anyone actually read the numbers?

Two of the 3 column are probably relatived to roll over. Model 3 is the least likely to roll over, thus the “safest” claim.

But apparently Elon came out and admit that Model X is safer because it is larger and heavier.. LOL. Just another day to spin the data for dumb public.

Did *you* read the data? The whole table in the screenshot is clearly all about the roll-over test data. That is just a sample (taken for this article) to demonstrate the sort of data published by NTHSA — it has nothing to do with how Tesla arrived at their claim!

Whatever. Good enough. From my un trained eye it looks safe. Less intrusion into the cabin than a lot of competitors. So good enough. I’m not going to sweat over it.

As long as it performs well when I need it to perform. No real crash will be exactly the same as a test.

In a classroom, there may be many A students but there’s always a smartest kid in the group. The Model 3 appears to be that kid.

I hope I never have to find out how safe my Model 3 is. Elon might play fast and loose with the facts but one thing is for sure: it is a JOY to drive. Haters either haven’t driven one or can’t afford to buy one.

Same crap happened when Model S was first tested and technically achieved 5.5 stars out of 5. NHTSA provides data to Tesla but tells them not to share it. Tesla, proud of its achievement, uses said data to promote its vehicles.

Um, no, this data is public.

The last graph shows in the lower right corner, “Source: ARK Investment Management LLC, NHTSA”. This may explain why the Model X and Model S are not in the X-axis.

In a real world accident (meaning not in a test facility), i’d most likely want to be in a Volvo, just my 2 cents…

Please provide column headings for the data chart. While some are obvious,(brand, model, year) we need to know the headings for the remaining 6 columns preceding the final, one would assume, “star” rating column.

Fixed. Thank you.