True/False? “NHTSA reports an average of one accident per 484,000 miles”

As soon as Tesla was on the road it had to start reporting deaths. Source:

I keep reading the following sentence in safety reports about Tesla, but only about Tesla:

NHTSA reports an average of one accident per 484,000 miles.

Do you see the NHTSA reporting that anywhere? I do not. And I do not see any other car manufacturer quoting this number either.

I see only a sentence Tesla put on their website to claim they aspire “to be” the safest car on the road. And then they wrote that sentence without any source or qualifications.

In other words the 484,000 miles reference is found nowhere but the Tesla site, which claims it found it somewhere else without telling us exactly where.

This December 2020 NHTSA report (DOT HS 813 060) is perhaps the closest thing: “Overview of Motor Vehicle Crashes in 2019”

Source: NHTSA’s National Center for Statistics and Analysis, Research Note: “Overview of Motor Vehicle Crashes in 2019”, DOT HS 813 060

Wow, as a percentage of total fatalities since 1995 more and more people outside cars are being killed!

Speaking of charts, here’s a real one based on the data that Tesla itself publishes.

Source: Tesla Vehicle Safety Report for the first quarter of 2021. It clearly shows the average distance per accident while driving on Autopilot, and driving off Autopilot, both declined year-over-year.

I am not kidding when I illustrate their own data showing the precipitous decline in safety over recent quarters, while their NHTSA number is showing almost no change. These are the real numbers they publish themselves. Bizarre.

So can someone find the magic 484,000 number anywhere in NHTSA reports? I have questions even if you can:

  1. Why didn’t Tesla put in a simple NHTSA reference to their claim? Don’t they want us to connect directly to the NHTSA and read that report if true?
  2. Why do people keep repeating this without any direct NHTSA reference? People say Tesla says that the NHTSA says a number. What? Nobody just says please show us this report? Can anybody find an actual NHTSA report that says this number?
  3. Does anyone understand what NHTSA might actually be talking about when they are cited improperly in this Tesla quote?

Until I see this report where NHTSA says the exact magic 484,000 number, I continue to believe something is very wrong with media channels repeating it as though it’s true.

Take this report that uses the number for example:

Stock in the electric-vehicle pioneer Tesla is wobbling after a Tesla vehicle crashed and police said no one appears to have been at the wheel.

Here’s another one that uses the number:

Tesla Q1 Safety Report Shows Rise In Autopilot Accidents

Why is that 484,000 data point being sourced from Tesla in these articles about Tesla safety failures, and NOT some statement or report directly from the actual NHTSA?

Perhaps Tesla is engaging in disinformation such that safety news is always controlled by them and them alone to poison a safety narrative?

Here are some guesses why Tesla doesn’t want someone to find or read a NHTSA report, even though Tesla wants us to believe they base their safety engineering on it:

  • NHTSA averages are for all vehicles in all conditions everywhere
  • Tesla averages are for a tiny subset of vehicles and conditions
  • Tesla doesn’t define methods or terms such as miles, crash, accident
  • Tesla crashes have been increasing, worsening not improving
  • Other car manufacturers are reporting their safest records in history during rise in Tesla fatalities and injuries

Saying Autopilot in a Tesla is safer than a 1995 rust-bucket on a dirt road where Autopilot can’t even function is a completely bogus comparison.

Tesla seems to be willfully misleading with its claims about crash data.

As an example of more meaningful comparison here is an actual NHTSA report on factors in crashes in the United States:

Source: NHTSA’s National Center for Statistics and Analysis, Research Note: “The Relationship Between Passenger Vehicle Occupant Injury Outcomes and Vehicle Age or Model Year in Police-Reported Crashes”, DOT HS 812 937

From this table we see 1995-2011 cars are clocking 1,030,624 severe injuries.

Meanwhile, 2012-2018 cars have only 199,480. So is the 2021 Tesla safety marketing campaign comparing itself to a 1995 car on purpose?

Also the NHTSA issues a warning about their own numbers:

…while the present analysis shows that the newer vehicle model year groups were inversely associated with occupant injury severity outcome, this study does not identify which aspects of the model year group with particular vehicular designs are responsible for the reduction in the risk of severe injury to vehicle occupants.

That’s literally the opposite of Tesla marketing, which repeatedly says their particular vehicle is responsible for reduction of crashes… despite no actual evidence to support such claims.

Tesla put its first cars on the road in 2013, right? So you can see it’s patently unfair to compare a 2013 or later model with anything prior unless making a completely different point about car safety (e.g. buy any new car, not an old car, because data shows generic new cars safer than all old ones).

Do you see a problem with Tesla comparing its particular cars to all crashes ever for all cars on the road instead of doing a true comparison with proper analysis?

What if we just run the numbers of Teslas crashing versus Teslas delivered. What percentage of Teslas crash, and how soon after being delivered?

Remember that table at the start of this post?

After putting only a few thousand cars on the road, and a CEO publicly stating his cars are the safest of all cars on the road, Tesla had to report two deaths from a car that “veers into opposite lane”.

Is there another car manufacturer that has as many deaths per cars delivered?

If you went out to buy a car today, Tesla continues to claim misleadingly you should see them as safer than ALL cars ever made, even when you are only in the market for NEW cars.

And when you’re in the market for new cars, Tesla may in fact be significantly less safe than other options (Volvo, Honda, etc). Here’s some proper analysis:

The fundamental problem here is that Tesla does a poor job of driver monitoring. Unlike several other automakers, Tesla only uses a torque sensor in the steering wheel to try to detect when the driver is moving the wheel. This is a cheap but very imprecise method.

A brand new Tesla uses “cheap but very imprecise” engineering for its safety.

Why would Tesla hide the reference to the NHTSA and make it hard to see the actual math? Seems cheap and imprecise of them.

Again, here’s some proper analysis.

General Motors’ similar Super Cruise feature, which is advertised as hands-free, uses facial recognition technology to ensure that a driver is watching the road while it is in operation and recently ranked higher than Autopilot in a Consumer Reports test

I don’t like “hands-free” marketing either, but you have to recognize that Tesla was ranked lower than other brands in safety using independent analysis

If nothing else, you should know Tesla clearly doesn’t want the NHTSA to speak for itself because it never seems to say to anyone “here’s the NHTSA” or “go read the NHTSA”.

Until I see people start to use original source NHTSA documentation when talking about NHTSA reports, I am extremely skeptical of the NHTSA being fairly referenced by Tesla.

If Tesla builds cars like they build their arguments to drive their cars, you shouldn’t buy their cars.

Here’s some poetry that might help explain:

Electric cars were the future in 1981.
Reagan shut it all down.
Electric cars were the future in 2001.
Bush shut it all down.
Electric cars were the future in 2021.
Tesla is a dumpster fire.

If you want to know why people are sticking with fossil fuels, it’s pretty clear who is keeping them alive. Yes, that’s silly. Let’s get rid of the combustion engine and get in our electric cars.

Just don’t get into a Tesla unless you’re prepared to be misled by funny numbers straight into a tree and die in a fire.

Why put people into an electric clown car? That does not help bring electric cars to market faster, as it destroys trust in new cars and their manufacturers.

Perhaps the best take in the news so far has been the Chaser:

“When we say we want a fully driverless future, we mean it” said Tesla CEO, Elon Musk at a press conference on Monday. Musk harked back to his childhood days as the heir of a Zambian blood-diamond empire “this tactical disdain for human life is crucial for any entrepreneur looking to really embrace change”.

For my take on broader disinformation issues in Tesla marketing, see my earlier post on their CEO tactics.

One thought on “True/False? “NHTSA reports an average of one accident per 484,000 miles””

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.