I'd say this is exactly where some kind of web of trust could work. Meaning that everyone is shown the reviews of their added family members, their trusted people (with less weight) and so on and on. If they betray the trust the system could even have a "trust less" button. It'd actually give people the ability to choose who they trust.
You don't normally need to trust their judgement as long as they aren't straight up lying. As long as they are making an attempt at being truthful, you just have to try to evaluate their review from their perspective.
E.g, Janet says product X works really well and she loves it, but you know she likes them with a certain feature that you don't like. therefore you should stay away because the product probably has that feature you don't like
It all depends on how you use reviews, I guess. I look for obvious problems and cost that in. Anecdotally it’s fairly difficult to hide bad reviews. It’s your expectations that make you unhappy; I don’t expect much buying a product sight unseen.
You see the same issues on Yelp. Reviews are always a secondary indicator of quality compared to, say, experiencing it yourself or getting a recommendation from a friend. I’d guess that the review style that is best depends highly on the good or service being sold, much like establishing consensus, and there is no general purpose review pattern that works well.
I imagine automatic reputation building for socket puppets will be right next on the market after such change.
People used to pay a premium for people that got out of their way and found the actually good stuff. Nowadays it's just more lucrative to build some trust and use it to defraud people as soon as possible.
I don't know where this path leads, but it's not looking nice.
I am really curious if this would actually work. I also think if you trust a reviewer and it turns out they are a bad reviewer than your trust rating should lower. (you turned out to be a bad reviewer reviewer)
That said I have no idea if this would. Seems those 87k people mentioned above would just trust each other to raise their trust ratings.
Its not about total trust "rating". That not how web of trust works. You, the informed and responsible user, have to look at who their connections are.
It's more like Linkedin: if someone adds you and they are a mutual connection with a bunch of coworkers you like, that's good. But if their mutual connections are all social climbing spammers, you know they either are one themselves or don't know enough not to fall for that stuff.
So now I have to spend time reviewing reviewers? This smells like the next bullshit job (or, since I'm not even paid for it, just a huge waste of time).
I'd say this is exactly where some kind of web of trust could work.
There is a temporal dimension to this too. Everyone loves their new toy just as they unbox it. What interests me is reviews after having used it for a few weeks or months.
Amazon has all the data it needs to filter reviews by “people who have actually bought this, review left after n weeks”.
