Comment on The Cybertruck Appears to Be More Deadly Than the Infamous Ford Pinto, According to a New Analysis
DrBob@lemmy.ca 4 weeks agoYou can draw conclusions because there’s only 35,000 on the road. That is a terrible rate.
Comment on The Cybertruck Appears to Be More Deadly Than the Infamous Ford Pinto, According to a New Analysis
DrBob@lemmy.ca 4 weeks agoYou can draw conclusions because there’s only 35,000 on the road. That is a terrible rate.
notsoshaihulud@lemmy.world 4 weeks ago
that’s how confirmation bias works, not statistical probabilities.
EM’s still a nazi and the CT is a horrible joke, but this is still insufficient data.
DrBob@lemmy.ca 3 weeks ago
Are you telling me that 35,000 vehicles is not a sufficient sample size to assess safety? Are you for real?
notsoshaihulud@lemmy.world 3 weeks ago
No. Incidence is a measure of probability of events over time (or with cars alternatively over miles). If the number of events is low (and 4 is low), your confidence intervals are extremely wide (which is the statistical way to say, we have no idea what the real number may be). The comparison is striking, the pinto had 27 fires over 9 years in >3M vehicles. fuelarc.com/…/its-official-the-cybertruck-is-more…
Let’s add that idiots buy cybertrucks who disproportionately think it’s bulletproof…
Again, “analyses” like this make great clickbait but contribute very little to our understanding, and that will remain the case even regardless of you getting angry at me about it or not.
DrBob@lemmy.ca 3 weeks ago
And the answer is"What is the Poisson Distribution" Alex.
There is literally a distribution that describes the occurences of low probability events in large populations. It was developed to study deaths by horse kick in the Prussian army. So confidence intervals never come into it. You’re applying Stats for Communications Majors reasoning to an adult problem.
Yprum@lemmy.world 3 weeks ago
It’s so great to be able to find comments such as yours, unfortunately it feels uncommon in Lemmy specially when certain names are mentioned, the bias and willfulness to shit on those are making people a bit blindsided and easy to guide through bad data usage. My first thought reading the title was about the statistical value of the numbers given, which doesn’t detract from the actual quality or lack thereof of the vehicle. At the moment using elon musk or tesla in a title of an article will increase the traffic automatically. Which is why we constantly get every single shitty comment made by him reported with useless data.