Tesla scrutinized by the NHTSA over Model 3 safety claims

[Twiglett]

I think the actual objection to the original "news" article is that its 8 month old new being resurrected for no reason.

 

[woodisgood]

I think the actual objection to the original "news" article is that its 8 month old new being resurrected for no reason.

Well, there’s a reason...

 

[omgwtfbyobbq]

More like Tesla is saying, "we got the best A in any test ever at the school", and the prof is saying, "WHOA there Sparky! All those grades from all those tests and all those years are not directly comparable."

I think it's more like the prof putting all of the grades together in a single spreadsheet, Tesla saying, "we got the best A in any test ever at the school", and the prof then saying "WHOA there Sparky! All those grades from all those tests and all those years are not directly comparable even though I dumped them into the same spreadsheet and didn't do anything in the spreadsheet to indicate that they are not directly comparable."

FWIW, here's the spreadsheet...

Regulations.gov

My opinion is that if NHTSA doesn't want Tesla using their data that compares the VSS across vehicle classes then the NHTSA needs to break up that spreadsheet by vehicle class.

 

[vickh]

I think it's more like the prof putting all of the grades together in a single spreadsheet, Tesla saying, "we got the best A in any test ever at the school", and the prof then saying "WHOA there Sparky! All those grades from all those tests and all those years are not directly comparable even though I dumped them into the same spreadsheet and didn't do anything in the spreadsheet to indicate that they are not directly comparable."

FWIW, here's the spreadsheet...

Regulations.gov

My opinion is that if NHTSA doesn't want Tesla using their data that compares the VSS across vehicle classes then the NHTSA needs to break up that spreadsheet by vehicle class.

Great link
3 RWD is the lowest in a sort, so I stand corrected in OP.

but why the AWD #NUM! error?

19-Sep Tesla Model 3 RWD 2018 5 5 5 5 5 5 5 5 5 5 0.38 5
19-Sep Tesla Model 3 AWD 2018 #NUM! #NUM! #NUM! 5 5 5 5 5 5 5 #NUM! #NUM!

 

[SageBrush]

The NHTSA makes a very straight-forward claim that it’s inappropriate for an automaker to make broad statements about a vehicle’s relative safety using the NHTSA crash test scores because those tests don’t appropriately factor in the effects of vehicle weight in a vehicle-to-vehicle crash. From a technical standpoint, I don’t see how anyone could disagree with this in good faith.

If that is the point then how do you explain 5 star ratings that ignore relative weight ?

 

[SageBrush]

 

[NoMoGas]

Truth is a valid defense. Tesla just stated the facts, nhtsa doesn't seem to think it's fair that Tesla tells people they literally scored the highest. The answer for the other automakers is to make a safer car which NHTSA should be supportive of.

 

[NoMoGas]

Everyone complaining about anti-Tesla bias should read the actual letter.

The NHTSA makes a very straight-forward claim that it’s inappropriate for an automaker to make broad statements about a vehicle’s relative safety using the NHTSA crash test scores because those tests don’t appropriately factor in the effects of vehicle weight in a vehicle-to-vehicle crash. From a technical standpoint, I don’t see how anyone could disagree with this in good faith.

No. It factors in the RESULTS of a crash test. I don't care WHY my car is safer, I care that it IS safer and by ANY metric, Tesla's are the safest cars on the road.... if you are IN them.

The rules about what automakers can/can’t say about NHTSA test results in public statements/advertisements are what they are, and they exist the make sure the public isn’t misled. But Tesla’s all like...

The "rules" are very deliberately NOT called rules... they are "guidelines".

 

[dskid]

Right, because per the test results, their A was better.

Incredibly, tests have scores measured primarily as a percentage which means that some A's are better than others! Some might be 86% and some might be 100%! They are all A's but not all A's are equal. Just like all 5-star safety ratings aren't equal; there's an underlying measurement system that grades where they are within that 5-star rating, called probability of injury (among other things).

 

[SageBrush]

If not for the NHTSA we would have manufacturers screaming for a NCLB (no car left behind) act.

 

[scottf200]

It's Tesla saying, "We got a 99.5% on the exam! Highest score in the class baybee!"

And the professor is saying, "It's an A, nothing more, nothing less. 10 other kids got A's too."

Tesla says, "But my A was better!"

If you car shopping for your spouse and kids would you like to know your car is
major injury free in accidents
99.5% of the time or 90% of the time.
Those are very very different " A's ".

 

[afadeev]

This is really old news (October 2018) dredged up again. What else would you expect from CNBC?

All referenced communications originate in March/April of 2019.

I do agree with Tesla's response below, but not sure if they can keep saying "safest"
https://www.plainsite.org/documents/fnrhg/tesla-nhtsa-foia-response/

Tesla's assertion is based on non-publicly released data.
Without seeing and analyzing that data first hand, none of us can draw ANY conclusions about the validity of Tesla's claims.

NHTSA, on the other hand, does have access to the data, and states that Tesla's claims are inaccurate.
So, we have two opposing claims:

1. From NHTSA (no known agenda)
2. From Tesla (obvious self-promotion agenda, and history of inflammatory and misleading pronouncements)

Tough choice, I know.

FWIW, here's the spreadsheet...
Regulations.gov
My opinion is that if NHTSA doesn't want Tesla using their data that compares the VSS across vehicle classes then the NHTSA needs to break up that spreadsheet by vehicle class.

The above is certainly interesting, particularly since the data in that XLS does not support Tesla's claim that Model 3 is the "best ever".
Not for Side MDB or Side Pole tests!
Yes for Rollover and Front, but only based on the 2018 test data sample.

I don't know where to find the results of 2017 and earlier years that would include all the cars with which Model 3 actually competes in the marketplace (3-Series, C-class, A4, Arteon, vs. X1, GLC, Q5, Atlas).

No. It factors in the RESULTS of a crash test. I don't care WHY my car is safer, I care that it IS safer and by ANY metric, Tesla's are the safest cars on the road....

Except that the 2018 NHTSA data doesn't support that assertion.
And NHTSA just sued Tesla to stop it from claiming so.

Do you have any data, other than Tesla's word, to dispute NHTSA's argument?

a

 

[NoMoGas]

[QUOTE="afadeev, post: 3906271, member: 97163]

Except that the 2018 NHTSA data doesn't support that assertion.
And NHTSA just sued Tesla to stop it from claiming so.[/QUOTE]

Yes, the data absolutely does back Tesla up. All one has to do is look at it.

and...

Can we PLEASE stop misrepresenting what happened? NHTSA sent a c&d... a lawyer letter, there is NO lawsuit.

 

[Dr. J]

but why the AWD #NUM! error?

19-Sep Tesla Model 3 RWD 2018 5 5 5 5 5 5 5 5 5 5 0.38 5
19-Sep Tesla Model 3 AWD 2018 #NUM! #NUM! #NUM! 5 5 5 5 5 5 5 #NUM! #NUM!

 

[SageBrush]

Tesla's assertion is based on non-publicly released data.

You should also note that the NHTSA has no complaints that the data should not have been used or published, or that the data in incorrect. Their *only* objection is that all 5 stars ratings are not represented equally.

This is not a case of Tesla misrepresenting data. The NHTSA is flat out wrong in its demands and approach, as befits a politicized agency

 

[atomicpunk]

Ok, I wasted 20 mins of my life trying to navigate the NHTSA site to find some actual rating of vehicles in this class, and I'm at a loss as to how you're supposed to derive actual comparison data, but could be me. So, in it's class, which cars offer a superior crash test rating? Isn't this the entire point? Which car can I buy today (in my class) that is the safest choice????

 

[CarlK]

NHSTA published hard numbers. Tesla used those numbers to make the interpretation. If NHSTA does not agree it need to make it's only interpretation and inform Tesla and the public which other car is the safest one.

 

[SageBrush]

Ok, I wasted 20 mins of my life trying to navigate the NHTSA site to find some actual rating of vehicles in this class, and I'm at a loss as to how you're supposed to derive actual comparison data, but could be me. So, in it's class, which cars offer a superior crash test rating? Isn't this the entire point? Which car can I buy today (in my class) that is the safest choice????

That is not the point at all, unfortunately.

This is the NHTSA telling Tesla to not hurt other car's feelings.
It makes their parents cranky.

 

[qdeathstar]

gonna side with a Tesla, all the data supports in and 40 percent of cars have a 5 star crash rating so the test is essentially meaningless.

 

[omgwtfbyobbq]

The above is certainly interesting, particularly since the data in that XLS does not support Tesla's claim that Model 3 is the "best ever".
Not for Side MDB or Side Pole tests!
Yes for Rollover and Front, but only based on the 2018 test data sample.

I don't know where to find the results of 2017 and earlier years that would include all the cars with which Model 3 actually competes in the marketplace (3-Series, C-class, A4, Arteon, vs. X1, GLC, Q5, Atlas).

Tesla claimed that the 3 was the safest based on the 3 having the best VSS score of any car tested, not the 3 scoring better on every single part of every single test that the VSS is based off of.

The VSS scores are I believe available for other vehicles/years if you search the regulations.gov site for other VSS spreadsheets. You can also search for specific vehicles in the NHTSA's crash test database.

NHTSA VSR | NHTSA vehicle Crash Test Database

The 3 doesn't score better than every car on every part of every test, but overall it has scored better than any other car on the VSS, which is a score based on the test results in the aggregate, and each test result is based on the test data in the aggregate.