This may well be true, even if you don't assume that Tesla owners are "smug" as some commenters are doing.
I've often thought that this is a major inherent problem with trying to estimate the "quality" of a product or service by taking the average of user reviews. If a product has high ratings, does that mean it's actually better than the alternatives, or does it mean it has flaws that cause the more discerning and critical consumers to gravitate to other options?
My understanding is recommendation systems can attempt to model and compensate for such biases. (At least, I recall seeing descriptions of how this was done way back in the Netflix Prize days.) But recommendation systems usually have their own issues with transparency and trustworthiness.
What I frequently want is the answer to a very simple question: of people who have tried both of two options I'm considering, what fraction prefer one over the other?
I've often thought that this is a major inherent problem with trying to estimate the "quality" of a product or service by taking the average of user reviews. If a product has high ratings, does that mean it's actually better than the alternatives, or does it mean it has flaws that cause the more discerning and critical consumers to gravitate to other options?
My understanding is recommendation systems can attempt to model and compensate for such biases. (At least, I recall seeing descriptions of how this was done way back in the Netflix Prize days.) But recommendation systems usually have their own issues with transparency and trustworthiness.
What I frequently want is the answer to a very simple question: of people who have tried both of two options I'm considering, what fraction prefer one over the other?