Hacker News Reader

Comment by mdorazio | Hacker News Reader

These kinds of titles bother me because it's never clear if that's actually a lot or not. I did some basic research and it looks like there are about 180,000 car fires per year in the US on average, out of about 275 million registered vehicles. Or about 0.6 fires per 1000 cars.

According to Motortrend, there are around 250,000 Teslas registered in the US. The table on this site lists 96 Tesla fires. Or about 0.4 fires per 1000 cars.

So, your Tesla is actually less likely to catch on fire than a regular ICE car. It's just more likely to do so while parked in comparison to an ICE car that's more likely to do so due to bad/no maintenance.

View on HN

Many of those ICE registered vehicles are 10 or more years old. A better comparison may be to limit the comparison to vehicles of the same age as the oldest Tesla.

I thought that too until I (as a former/current BMW driver) saw how many new BMWs burned down. What I found most interesting is how e.g the Korea recall did not even register on any of our German newspapers or news agency in any way comparable to Tesla (there where some really small footnotes to be fair)

https://money.cnn.com/2018/08/14/news/companies/bmw-korea-fi...

https://abcnews.go.com/Business/bmw-mystery-fires-abc-news-i...

Wait how does this anecdote change the need to compare apples to apples? Also, if anything this anecdote highlights why this comparison is so poor. What if BMWs are responsible for all the ICE car fires? Why wouldn't Tesla compare against the safest or biggest automaker's stats then, to know if they're doing well? E.g. Toyota?

>Wait how does this anecdote change the need to compare apples to apples?

Because when you get into a car there are a lot more risks than it spontaneously combusting; and for those risks (like getting into an accident with a drunk driver), Tesla is already top of the class; almost as safe as a Volvo.

So is 7 in the last four days a lot; and am I in more danger in a Tesla, when accounting for other risks? If I am just as likely to; or even twice as likely to combust in a Tesla chances are the only car I'd be safer in is a Volvo.

I'm still confused, I thought we were responding to this comment: `So, your Tesla is actually less likely to catch on fire than a regular ICE car.` << this assertion is not supported by the named statistic, that's all we are saying here.

> for those risks (like getting into an accident with a drunk driver), Tesla is already top of the class

I guess while we're here, I don't know if that's easy to back up either. Overall fatality rates vs other models show a complicated story.

It is ain’t so safe if it swerves into oncoming traffic due to its shitty “fsd feature”.

No one was talking about comparing general car safety, the discussion was specifically around car fires on ICE vs Tesla EVs for relatively newer vehicles.

I don't understand why you went on this diatribe about general safety...

The one person I know whose car caught fire, it was a fairly new Toyota RAV4 - probably three or four years old at the time. They were driving on a highway, saw smoke from under the bonnet (hood), pulled over and within a few minutes the whole thing was on fire!

Are older vehicles likely to get fire?

poor maintenance would be the most common reason, e.g a lose hose.

Then you have to compare the class of the vehicles, Tesla ones are likely to be the more expensive ones.

Yes, especially if they have old rubber fuel hoses (Volkswagen).

Most of the car fires I have heard of are caused by incorrectly installed auxiliary components due to incorrect fuses and unsuitable cable diameters. This is a phenomenon that probably does not affect Tesla at all.

An even better comparison would also distinguish between differently fuelled ICEs. A diesel powered car is much less likely to catch fire than a petrol one.

And, as previous commenter points out, age comes into the mix too. I used to be involved in the old VW camper van scene. Due to the age of those vehicles, it was generally advised to replace all the rubber fuel lines on those when buying a petrol engined one, as so many people had had engine bay fires because of perished rubber piping.

That's just being bad at statistics. How old are the cars that catch on fire on average, and why do they catch fire? You can't just compare all cars to some subset of cars and derive a basic conclusion from there. For example; the average car on the US roads is 12.2 years. Tesla only started mass production 10 years ago.

By that logic, I think you'd first need to show that the age of the car impacts the likelihood it catches fire. Product failures tend to follow a bell curve over the period of their lifetime.

To compare in good faith you still have to cut at the same side of this bell curve though.

Tesla's imaginary PR department engages in this sort of nonsense themselves. They compare their fleet's crash rate, for example, to all cars on the road in the US, at a time when the age of the average car on the road is the oldest it has ever been.

The average car on the road also isn't a well-optioned mid-luxury car with driver assistance aides.

It would be like BMW saying "our new BMW M-whatever is faster than 90% of cars on the road!" Uh, yes, but how fast is it compared to other new, similarly priced sports sedans?

When you compare Teslas against cars made in a similar time period, in a similar market space (mid luxury) which are much more likely to have driver assistance aides like emergency braking, lane departure warning / intervention, cross traffic warning, etc - Teslas have a much worse crash rate.

If you have data that granular, please share. If not, a top-level analysis seems better than none at all.

It's worse than none, it's basically useless. Take into account safety ratings of all Tesla's on the road (one of the highest) vs safety rating of all cars on the road (average at best), and then think again whether 0.6 vs 0.4 says anything at all. The actual difference could be many factors higher.

Another reason this might be news-worth is if the rate of Tesla fires suddenly rose. I didn't look at any numbers and I suspect the number didn't suddenly change. Just trying to point out that ICE fires aren't the only baseline to compare to.

"I did not actually look to see if what I'm saying is true. I'm going to just declare it is and rely on someone caring enough to do the legwork to disprove me, assuming that's even possible."

Reply on news.ycombinator.com