Feds told Tesla to stop making “misleading statements” on Model 3 safety

"Your company has issued a number of misleading statements," NHTSA wrote to Musk.

Read the whole story

 

[GoodOnYaMate]

This is a case where both sides are right.

(and where I'll be downvoted by readers who do fall on a particular side )

 

[Kruskal-Szekeres]

This is something that has been raised a lot in discussions, when The Faithful inevitably parrot The Leader. Its good to see it finally being addressed at a higher level, although I don’t look forward to the dumpster fire this comment section will inevitably become.

 

[r3ersx]

Tesla playing fast and loose with the things it claims in its marketing? Shocked /s

 

[Amos]

So Tesla was technically correct, which is the best kind of correct.

 

[ryanja]

It's better to be upfront than be like VW with all the "JD Power" awards

Ah, JD Power awards. The participation trophy of the automotive industry, purchased by the companies to make sure all their little cars and trucks can feel special about not quite being a good product.

 

[phat_tony]

If Tesla has said "lowest probability of injury in the NHTSA tests of any vehicle ever tested by NHTSA" then I think they'd be on pretty firm ground. The NHTSA probably still wouldn't like it because it's not how they want their data to be used, but it seems like a pretty factually solid claim.

However, Tesla's claim as stated really does sound like the NHTSA data shows that the Model 3 has the lowest probability of injury in real world accidents of any vehicle ever tested. They don't state that, but it's the kind of subtext that will probably be assumed by most readers. Tesla may be able to make a decent argument that their statement is not factually incorrect, but if you told the average person on the street what Tesla said and asked what it means, I bet the results would be pretty clear that this is a misleading statement.

Based on their statement, most people would probably think they'd be better off in a two car collision being in a 1,645 kg Model 3 than a 2,648 kg Nissan Patrol. This is almost certainly not the case - probably by a long shot.

 

[bigcheese]

“Safest” and “safest in this particular test” are not equivalent. If anything, I would consider that to be the most misleading thing about Teslas statement.

 

[jbauer5]

Ah physics, objects in motion tend to stay in motion. And the most massive land yacht carries more momentum on impact. So bring back the Hummers! Feel safe knowing you've used the other vehicle as a crumple zone.

 

[Jim Z]

Tesla should absolutely respond with hard data & close this one out

they don't have any. This is merely NHTSA reminding them "hey, our data doesn't say the things you're claiming it says, so stop claiming it does."

 

[r3ersx]

The NHTSA would have liked Tesla to stop there. Instead, Tesla dug into the NHTSA's data and spotted an opportunity to further toot its own horn. As part of its evaluation process, the NHTSA calculates a number called a vehicle safety score, which the agency has characterized as “relative risk of injury." The agency then awards each vehicle a star rating based on VSS ranges.

Tesla noticed that the Model 3 had a better VSS score than any other vehicle on the market. That, in Tesla's view, means that a Model 3 driver is less likely to be injured in a crash than a driver of any other vehicle.

What exactly did NHTSA think would happen, if they butt heads with a team of nitpicky software/hardware engineers, hell bent on proving a point

NHTSA cannot hand wave this away, just because they are a gov't agency.
Either clean up their own ratings or face the result

If Tesla is correct on what they did in court, it will be egg on the agency's face
& they will have opened up the possibility of digging thru stuff like this for other car makers, who will also be "technically correct"

Is there anything Tesla related you won't aggressively defend on Ars? Like Tesla is being dishonest here, there is no 'gotchya!' with government agencies. The cult of personality around Elon and luxury goods people buy is truly strange.

 

[Fatesrider]

So Tesla was technically correct, which is the best kind of correct.

Not so.

Musk left out the context of his claim. The relative safety level also doesn't mean "safe" in all conditions. That's another contextual omission. Omitting context when making gross generalization claims is a form of deception.

The fact is, the heavier a vehicle is, the more inertia it can absorb, and the less energy it will transmit to the occupants. Another factor is how much damage the batteries can take without burning. Front impact is usually the direction that safety engineering focuses. Side, semi-side and rear collisions aren't looked at as often, but any collision with a BEV has the potential to rupture the batteries, and cause a fire.

So with respect to safety, if the Model 3 runs into something, it's probably relatively safe. But if something hits the Model 3, there's no data to support a safety claim. That's another contextual omission. The claim wasn't quantified at all.

The NTSHA is correct to bring this to Tesla's attention, and to push for a retraction. Without the proper context, the understanding of relative safety is absent and therefore fails to support the claim. It is misleading, and could well be actionable by the FTC as false advertising.

So, no, not technically correct, either. And the article said why it wasn't.

 

[Jim Z]

I'm guessing NHTSA just doesn't want Tesla drawing from from their own data or being associated with them. They want to remain independent. If Tesla were to acquire the same data from an independent source, I'm sure NHTSA would have no qualms.

No, they're saying "stop telling people our data says this. it doesn't."

because it doesn't.

 

[AnchorClanker]

It's as if the libertarians all bought Teslas.

 

[Jim Z]

So Tesla was technically correct, which is the best kind of correct.

Not so.

Musk left out the context of his claim. The relative safety level also doesn't mean "safe" in all conditions. That's another contextual omission. Omitting context when making gross generalization claims is a form of deception.

The fact is, the heavier a vehicle is, the more inertia it can absorb, and the less energy it will transmit to the occupants. Another factor is how much damage the batteries can take without burning. Front impact is usually the direction that safety engineering focuses. Side, semi-side and rear collisions aren't looked at as often, but any collision with a BEV has the potential to rupture the batteries, and cause a fire.

So with respect to safety, if the Model 3 runs into something, it's probably relatively safe. But if something hits the Model 3, there's no data to support a safety claim. That's another contextual omission. The claim wasn't quantified at all.

The NTSHA is correct to bring this to Tesla's attention, and to push for a retraction. Without the proper context, the understanding of relative safety is absent and therefore fails to support the claim. It is misleading, and could well be actionable by the FTC as false advertising.

So, no, not technically correct, either. And the article said why it wasn't.

IIHS includes side impact in their battery of tests (pun intended.) IIHS's crash tests aren't regulatory requirements, but doing poorly on them can affect public opinion.

they haven't tested the Model 3 yet, but the Model S was rated "good" in side impact.

(as an aside, it's a testament to modern automotive design that a collision like that now has a "low risk of serious injury" when not 30 years ago it would likely have been a death sentence.)

 

[megatrex]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

 

[Person_Man]

This is a case where both sides are right.

(and where I'll be downvoted by readers who do fall on a particular side )

Funny, as I read this, you have 14 upvotes and 14 downvotes. I abstained from voting to keep it even.

-d

 

[Sarty]

NHTSA seems to be taking the position of "if you're going to play marketing-speak with the facts, don't attach our name to it".

It would certainly be worse if Tesla's claims were factually incorrect, but their being factually correct does not absolve them of using sneaky language to play on the public's good faith in a third party.

 

[Jim Z]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

What's missleading there?

they can't currently self drive, and they've not shown that the hardware they're selling is indeed capable of full self driving.

 

[Urist]

Weren't they already banned from making misleading statements? Or was that just Elon himself.

 

[r3ersx]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

What's missleading there?

Tesla has claimed they have been close to FSD for years, Musk had even promised they'd complete a coast to coast drive 100% autonomously by the end of 2017, and took large payments for FSD on the back of this promise. That drive has still not happened yet, and no indication they're close to it or anything resembling 'full self driving'.

 

[megatrex]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

What's missleading there?

they can't currently self drive, and they've not shown that the hardware they're selling is indeed capable of full self driving.

They updated their "full self driving" package after they failed to deliver FSD for several years after the promised date. So now you actually get something for what you paid, but none of it is actual full self driving.

 

[Dr Gitlin]

So Tesla was technically correct, which is the best kind of correct.

Not so.

Musk left out the context of his claim. The relative safety level also doesn't mean "safe" in all conditions. That's another contextual omission. Omitting context when making gross generalization claims is a form of deception.

The fact is, the heavier a vehicle is, the more inertia it can absorb, and the less energy it will transmit to the occupants. Another factor is how much damage the batteries can take without burning. Front impact is usually the direction that safety engineering focuses. Side, semi-side and rear collisions aren't looked at as often, but any collision with a BEV has the potential to rupture the batteries, and cause a fire.

So with respect to safety, if the Model 3 runs into something, it's probably relatively safe. But if something hits the Model 3, there's no data to support a safety claim. That's another contextual omission. The claim wasn't quantified at all.

The NTSHA is correct to bring this to Tesla's attention, and to push for a retraction. Without the proper context, the understanding of relative safety is absent and therefore fails to support the claim. It is misleading, and could well be actionable by the FTC as false advertising.

So, no, not technically correct, either. And the article said why it wasn't.

IIHS includes side impact in their battery of tests (pun intended.) IIHS's crash tests aren't regulatory requirements, but doing poorly on them can affect public opinion.

they haven't tested the Model 3 yet, but the Model S was rated "good" in side impact.

(as an aside, it's a testament to modern automotive design that a collision like that now has a "low risk of serious injury" when not 30 years ago it would likely have been a death sentence.)

The Model S is actually a good example of automakers responding to IIHS testing by making their cars better. When IIHS put one made in 2016 through the small frontal overlap test the dummy's head made contact with the steering wheel rim, so from 2017 the seatbelts were redesigned and Model Ss made after that date score better.

 

[Crackhead Johny]

Would it be OK if he makes the statements while obviously smoking a huge joint, so that people know he is high?

 

[lewax00]

But the NHTSA argues that this is statistical malpractice because it doesn't take into account vehicle weight. In a vehicle-to-vehicle crash, the occupant of the heavier vehicle is less likely to be injured. The NHTSA's tests, which involve crashing a car into fixed objects, don't necessarily account for this difference.

But wouldn't a fixed object be roughly equivalent to a very heavy object, in terms of how the collision works? Or even potentially worse? I'm not sure I understand this argument.

Not that I disagree that this is probably misapplying the statistics, but that's a standard marketing tactic.

 

[pkirvan]

So Tesla was technically correct, which is the best kind of correct.

If that's what you got from this article you need to read it again. Tesla's are possibly the safest vehicles in collisions with fixed objects and other identical Teslas. Their claim, however, was to make you the safest, period. In fact larger vehicles, such as perhaps a bus or tank, could indeed keep you safer. So no, not technically correct at all.

 

[Sarty]

But the NHTSA argues that this is statistical malpractice because it doesn't take into account vehicle weight. In a vehicle-to-vehicle crash, the occupant of the heavier vehicle is less likely to be injured. The NHTSA's tests, which involve crashing a car into fixed objects, don't necessarily account for this difference.

But wouldn't a fixed object be roughly equivalent to a very heavy object, in terms of how the collision works? Or even potentially worse? I'm not sure I understand this argument.

Not that I disagree that this is probably misapplying the statistics, but that's a standard marketing tactic.

I think a fixed object is identical to a vehicle of equal weight moving at equal speed in the other direction. It seems like this must be so for reasons of symmetry. A heavier vehicle is worse than a fixed object -- rather than arresting you completely, you're going to be thrown backward.

Unless it's a Wednesday afternoon and I can't brain good.

 

[pkirvan]

But the NHTSA argues that this is statistical malpractice because it doesn't take into account vehicle weight. In a vehicle-to-vehicle crash, the occupant of the heavier vehicle is less likely to be injured. The NHTSA's tests, which involve crashing a car into fixed objects, don't necessarily account for this difference.

But wouldn't a fixed object be roughly equivalent to a very heavy object, in terms of how the collision works?

No, not at all. Collisions with a fixed object depend basically on how gradually your vehicle can slow you down. Crumple zone size and whatnot. Any car can perform well in that measure. Collisions with a much larger vehicle going the other way will essentially reverse your speed instantly leading to huge accelerations. There is very little that the maker of a small car can do to protect you from the accelerations that will be produced if you hit a semi trailer head on.

 

[Jim Z]

But the NHTSA argues that this is statistical malpractice because it doesn't take into account vehicle weight. In a vehicle-to-vehicle crash, the occupant of the heavier vehicle is less likely to be injured. The NHTSA's tests, which involve crashing a car into fixed objects, don't necessarily account for this difference.

But wouldn't a fixed object be roughly equivalent to a very heavy object, in terms of how the collision works? Or even potentially worse? I'm not sure I understand this argument.

Not that I disagree that this is probably misapplying the statistics, but that's a standard marketing tactic.

no, because you can't simulate all the possible things you might collide with. NHTSA's frontal crash test is full frontal into a rigid barrier. Which is actually kind of rare in the real world. IIHS has done moderate overlap into a deformable barrier (crumples like a car) to simulate offset collisions (fairly common) and small overlap against a rigid barrier for things like clipping a post or an unprotected concrete divider. But the specific amount of overlap between the colliding vehicles, or the angle at which they collide, or a bunch of other possible variables come into play which can't be simulated by a single crash test. What NHTSA did is use their model to say "based on the Model 3's results on this test, compared to other vehicles' results on this test we calculate a lower chance of injury." Tesla took that ball and ran it all the way over to "the Model 3 is safer than any other car period." And NHTSA's responding "that's not what we said."

(the key words above are emphasized.)

 

[Dilbert]

It's better to be upfront than be like VW with all the "JD Power" awards

Ah, JD Power awards. The participation trophy of the automotive industry, purchased by the companies to make sure all their little cars and trucks can feel special about not quite being a good product.

And now the award for the best car in 105 to 110 inch wheelbase, red body, hatcback, with alumimum wheels, and 1.8L turbo, goes to: <drumroll>

And now the award for the best car in 110 to 115 inch wheelbase, blue body, coupe, with steel wheels, and 1.9L turbo, goes to: <drumroll>

 

[Dilbert]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

What's missleading there?

they can't currently self drive, and they've not shown that the hardware they're selling is indeed capable of full self driving.

That's the thing. The hardware is supposed to be 100% complete, and customers are just waiting for this mythical OTA software upgrade so they can sleep, read a book, or eat, during their commute.

It just isn't going to happen people.

I think there will be a lawsuit, and then all these Teslas will be recalled and brought back to retrofit with sensors that would work. And of course that will be years from now.

 

[Jim Z]

Misleading statements is their entire marketing plan. See: Selling "full self driving".

What's missleading there?

they can't currently self drive, and they've not shown that the hardware they're selling is indeed capable of full self driving.

That's the thing. The hardware is supposed to be 100% complete, and customers are just waiting for this mythical OTA software upgrade so they can sleep, read a book, or eat, during their commute.

It just isn't going to happen people.

I think there will be a lawsuit, and then all these Teslas will be recalled and brought back to retrofit with sensors that would work. And of course that will be years from now.

probably cheaper just to offer a partial refund of the price paid for the package.