The current "industry standard" is 1 accident per 10 million flights. I think the MAX 8 has racked up something around 150K flights with 2 accidents. So yes, the statistics represent an anomaly.
How did you get that number? Assuming a binomial distribution with p=1/10^7 and n=150*10^3 samples as the null model, the probability of 2 or more accidents happening by chance is 0.011%.
I assume you meant: If an airplane is as safe as average then it has PUT_NUMBER chance of having 2 incidents after 150k flights. 0.01% is actually the number I'm getting, assuming parent estimates are correct and making naive assumptions. In other words only 1 every 10 000 airplane models will have 2 incidents that early on if they are of average safety.
That is different then stating the probability of it being as safe as the average airplane, which you can't do as easily without additional modelling/priors and bayesian statistics.
Lies, damn lies, and statistics.
The NH is that a plane has 1/10 [M * flights] failure rate. The odds of 2 failures in 20M flights falling in same (random) stride of 150K flights are 150K/20M = 0.75%.
[E: fixed numbers]
[EE: yes, I admit this calculation is incorrect]
If we assume that accidents are rare and independent per mile, then the correct distribution is a Poisson distribution with a parameter of 150,000/10,000,000 = 0.015. In which case the odds of no accident should be 98.5111939603...% and the odds of one accident should be 1.47766790940...%. The odds of 2 or more accidents is 100% minus those, which is 0.0111381302891...%.
It gets even worse when you consider the fact that both crashes happened while the plane was in the air. Nearly half of all accidents happen at takeoff and landing. This makes the per mile, just flying along failures even less likely.
Is that figure per model or per actual airplane? If the latter, 0.01% seems close to something I'd expect to actually see just by chance.
If I understand correctly, that's the probability for 2 crashes within any span of 150 kiloflights. I'm curious if it makes sense to ask about the first span of 150.
It's likelihood per 150k flights per model (although likelihood per 150k flights per individual plane would give the same answer, it's just a weirder question).
The assumption is that for the average plane all accidents happen independently from one another. Under such an assumption, the probability for any span of 150k flights is exactly the same as the probability for the first span of 150k. So it would not change the answer.
If you roll a die a 1000 times, the odds of landing on a 6 on roll 999 is the same as it landing on a 6 the very first roll.
reply