Tesla Model 3 rollover crash shows its real-world safety

>I too wonder how this car is on the market without having passed federal safety tests.

You don't have to crash test a car to sell it. You just get in trouble for selling it if the results (either real world or lab) turn out to be unacceptable after the fact. This system of requiring compliance and then doing spot checks and enforcing after the fact seems to work fine (across many industries, not just cars).

From the linked Reddit thread.

> We'll be placing an order for a new Model 3 as soon as the insurance is sorted out.

Satisfied customer and Tesla will get another sale.

Glad to see other people prefer old.reddit too.

The pictures are indeed stunning but with no other _verifiable_ information (like the exact circumstances such as speed) these should be taken with a grain of salt especially on Reddit, which is considered basically Tesla's astroturf.

This is an important idea in performance vehicle design - in a crash, the 'capsule' around the occupant must not be compromised. You see this pretty vividly in bad F1 crashes. Granted, those vehicles are so weight optimized that they'll lose a wheel if they tap a kerb, but in particularly bad wrecks you'll see that the entire car, aside from the capsule has pretty much disintegrated.

example: Fendando Alonso's 2016 crash. Looking at photos, the car had been reduced to a capsule with a bit of engine and wheel sticking off of it. Alonso climbed out of the cockpit himself.

We don't seem to know how hard the roll-over was but we can see at least three of the doors still open. Part of that will be because they are pillar-less so less likely to jam at the top but it's still pretty impressive.

Sample size of one in uncontrolled circumstances (i.e. one crash in the real world as opposed to semi-scientifically conducted crash tests designed to give information from which we can infer real world performance) tells us very little.

Depending on the circumstances a barrel roll down the highway and/or median could be a batting practice pitch over home plate because there's practically an eternity over which the vehicle decelerates (or it could be a really tough crash, it depends on the details, we don't know).

Edit: and I'm wrong because? I'm not even bashing Tesla here.

Pretty sure the sales guy would mention, you know... actual crash test results. I agree this might be PR spin but the Tesla (all of them) tests quite well, well enough that they don't need to rely on "some guy on Reddit".

Edit: I realized this is specifically about the Model 3, which doesn't have official data yet. What I said is probably still accurate - what do BMW salespeople say about a new model that doesn't have data in yet? Probably: "Look how safe our other cars are!"

oh come on, surely you are not suggesting the crash was deliberate?

If you look up crash ratings, you will find it's not uncommon for new car models to be missing them. It just isn't something that is done before the first unit is sold - presumably because as other people mentioned, it has to be done independently to be useful. Of course there must be internal tests, but there's no use in publicizing them because of the conflict of interest, and I'm sure if Tesla did that, people would point out the problem.

People who are concerned about Tesla unwittingly give away the fact they don't browse the IIHS web site for fun. It is educational.

If you search for the Honda Insight on said IIHS site, you will find the old one that ceased production in 2014, not the new one that just arrived at my local Honda dealer. This is not shocking if you had any interest in the subject independent of Tesla.

The NHTSA site lists the 2019 Insight, but surprise, surprise, it says it is not rated.

Because the company that is responsible for crash tests hasn't bought the car yet?

They have to buy cars in the market to prevent manufacturers from adding reinforcement

That was my takeaway from the article. Why on earth are cars being sold which haven't gone through proper safety testing?

maybe for parts only, nobody will drive that thing as-is anyway

I don't think they can maintain 5,000 Model 3's a week pace without improving their process. However, ramping from 5,000 to 10,000 cars per week is simpler than getting to 5,000 per week.

Also, 5,000 is in reference to just Model 3's. "The delivery breakdown for the quarter was 18,440 Model 3’s, 10,930 Model S vehicles, and 11,370 Model X SUVs." https://electrek.co/2018/07/02/tesla-official-model-3-produc... 5,000 Model 3's per week ~= 65,000 per quarter.

How big is your short position?

While meeting those goals is important to Tesla's success, from the article: "... Tesla didn't sacrifice safety in its scramble to meet production goals"