I've had a defect rate of less than 1% percent with a sample size of enough items to reduce the margin of error on that statistic to something rather negligible.
To get your configuration you need to have two items that aren't defective and five items that are.
There are seven choose five ways you could have a particular configuration where you have five defects out of seven purchases.
7 Choose 5 = 21
So therefore the probability of what happened to you happening to me would seem to be:
9.801e-11 * 21 = 2.05821e-9
This written out in numbers is: 0.00000000205821 That number as a percentage is: 00.000000205821% For comparison the probability of being struck by lighting over the course of a person's life is 1-in-15,300 which is 0.00006535947. Many people struggle with reading really small numbers like that and so we often decide to round. If we do that the probability of what happened to you happening to me is roughly 0% if we choose to round at six decimal places. Seconding the OP claim that the general experience is shocking, because it is extremely divergent. It would be more shocking to be hearing everyone on Hacker News was regularly being hit by lightning bolts than that people are getting your defect rate.
When you see a result like that it’s a good idea to check your assumptions.
That calculation assumes these probabilities are independent when they aren’t. What was ordered, when it was ordered, and where it’s being shipped to are all likely to impact the odds. On top of that are ability to detect fakes and defective items are likely to be different.
Having said that, I have been hit by lighting. Or at least it stuck the car I was in and I felt some effects. I didn’t report it anywhere that records such strikes, which suggests non serious lightning strikes may be underestimated in those statistics.
> What was ordered, when it was ordered, and where it’s being shipped to are all likely to impact the odds.
given the wildly diverging experiences people report, this is my assumption also. at least to my knowledge, I've never received a counterfeit item from amazon, but there are also certain things I avoid buying. for example, I would probably not buy lipo cells, especially not from a third-party merchant. they are trivial to rewrap and poor QC can literally start fires.
would you mind sharing what kinds of things (or just categories of things) you are buying to see such a high rate of counterfeits?
I've had a defect rate of less than 1% percent with a sample size of enough items to reduce the margin of error on that statistic to something rather negligible.
To get your configuration you need to have two items that aren't defective and five items that are.
P(!defect)^2P(defect)^5 = 0.99^20.01^5 = 9.801e-11
There are seven choose five ways you could have a particular configuration where you have five defects out of seven purchases.
7 Choose 5 = 21
So therefore the probability of what happened to you happening to me would seem to be:
9.801e-11 * 21 = 2.05821e-9
This written out in numbers is: 0.00000000205821 That number as a percentage is: 00.000000205821% For comparison the probability of being struck by lighting over the course of a person's life is 1-in-15,300 which is 0.00006535947. Many people struggle with reading really small numbers like that and so we often decide to round. If we do that the probability of what happened to you happening to me is roughly 0% if we choose to round at six decimal places. Seconding the OP claim that the general experience is shocking, because it is extremely divergent. It would be more shocking to be hearing everyone on Hacker News was regularly being hit by lightning bolts than that people are getting your defect rate.