Live data from Hacker News

Amazon's Fake Review Problem

brianbien.com

271–280 of 336 posts

Re: Amazon's Fake Review Problem

#271
post #249
post #188

Earlier quoted context omitted.

right? I have a HD280 from ~12 years ago? the headband started to disintegrate maybe 5 years ago but it's still quite usable at home and the sound and comfort is still as wonderful. bought a 2nd pair maybe 5+ years ago when it was discounted at literally half-price, so that's the one that gets taken outside on field-trips.

Joining this bandwagon. The headband on mine just sort of started to disintegrate into a weird sticky mess after 2-3 years. Then someone pulled them open too hard and cracked the plastic. Mine are from 2004, still working perfectly, but the whole headband area is now pretty much duct tape and glue.

Duct tape and glue?

I got angry, took my drill, a piece of metal, couple of screws and connected the whole thing Mad Max style. Yes, they were quite uncomfortable.

And you couldn't really wear them in public too.

---

And yes hd380s seem to have longer lasting headband than the hd280s.

Re: Amazon's Fake Review Problem

#272
post #206

Earlier quoted context omitted.

Listening is extremely subjective though. Headphone types can be better / worse for certain listening habits (there are people out there who don't think great bass makes great music, folks), and for different ears. Music taste itself is extremely subjective. Best thing is still to get to a trusted hifi store nearby with your own, known music in best quality and your current best headphone model and test through candi…

> Headphone types can be better / worse for certain listening habits No they can't. Loudspeakers (of all kinds) can either faithfully convert electrical signals into the vibrations they represent, or they can't. Equalizer presets can be better / worse for certain listening habits. There is no reason, when even a $5 MP3 player has EQ presets, that headphones shouldn't be aiming for a flat frequency-response curve. Now…

There is no reason, when even a $5 MP3 player has EQ presets, that headphones shouldn't be aiming for a flat frequency-response curve.

Sure there is. The market might prefer a baked-in EQ preset. In fact, different subsets of the market may prefer different presets. And that brings it back to the grandparent's point. The customer is always right; even when they're technically wrong, their preferences are what count.

Re: Amazon's Fake Review Problem

#273
post #248

Earlier quoted context omitted.

For loudspeakers it's more complicated than a simple flat on-axis response; the dispersion into the room is critically important. Which in turn means that a flat frequency response curve is the wrong target for headphones designed to mimic good speakers, because a flat loudspeaker is not flat when measured at the ear, which is where headphone drivers are located. Harman has done some really interesting research on th…

Good point, but I do wonder—is "full fidelity" really best defined as matching the sound profile of good speakers? Or is it better defined by e.g. whatever the sound engineer was hearing on their studio monitors when they finished the mastering process? That is: in most digital artistic media, the goal of the reproduction process is to faithfully carry across the art (and engineering) that "went in" on the other side…

That is: in most digital artistic media, the goal of the reproduction process is to faithfully carry across the art (and engineering) that "went in" on the other side

I can't agree with this. I don't think it's right at all. Art is targeted at a broad band of senses, from aesthetic through sensory to emotional. You're hung up on the narrow band of sensory reproduction, but that isn't what's most important to actual humans.

Take an example: some electro house track that person A might have heard on a night out, where they had a really good time. They would have heard it in the middle of a crowd (dampening a significant chunk of the sound), probably overly loud, probably with the bass pumped up.

If person A later listens to that track on studio headphones focused on faithfully carrying across engineering and "art", they'll be hearing the mechanics of the music and how it was constructed, but it will be subjectively a poorer reproduction of the remembered moment and experience. They'd actually be better off listening on much lower quality headphones that had boosted bass and muddy reproduction that let different frequencies intermingle, to make sure it doesn't sound spare and separated out like high quality headphones tend to.

Another example: person B might have been to a symphony or opera. They'll have a much happier time with high quality headphones separating out each instrument, appreciating the individuals even as they form the symphony.

Context.

An analogy: the "default", pre-choice-of-audience way to reproduce a painting shouldn't involve a frame and gallery wall, but rather should involve the painting standing alone as a canvas on an easel in a room in the same lighting conditions it was painted in

I think you're nuts! There's multiple frames to view a piece of visual art; from the artist's perspective, they may have designed the piece for a specific location, so you're better off if it's shown that way. It may have been designed to be contextualized (e.g. in a triptych, or as ironic contrast with its original commission). But the passage of time also contextualizes art; art can become meaningful not for its intrinsic artistry or aesthetic, but because of the narrative around the piece, or the artist, or the city / country at the time. A piece of art might be interesting only because e.g. it was once stolen by a very interesting person and got into the news that way; it might not have anything to do with the artifact as a thing-in-itself.

And I hasten to add that that doesn't mean it's bad art. Art is about provoking broad band reaction in the viewer. In order to do that, it needs to trigger the viewer's pattern matching neurobiology relating to different aspects of their lived experience. That's inherently contextual, both on the subject side and on the object side. There's no escaping context.

Re: Amazon's Fake Review Problem

#274

Amazon does do a lot to detect bribed reviews automatically. However, in the long run, this may be an arms race that Amazon will lose. One issue with detecting fakes with identical phrases is that many real reviewers will take phrases from the product description or from other peoples' reviews either consciously or unconsciously. This also matches up with a common bribed review technique of giving people canned phras…

> smart reviewers salt their reviews with the occasional negative one (generally for people who aren't bribing them) This is exactly the problem with the "consumer protection" racket--companies like Consumer Affairs will curate a negative presentation of your company or product, then call you up offering a $2,500/month subscription which basically allows you to remove old negative reviews and prevent new ones from ap…

Agreed, Yelp does the same thing.

Re: Amazon's Fake Review Problem

#275

Earlier quoted context omitted.

I wish that Amazon could split the market place off. I disliked it from the first day they introduced it. I want to buy from Amazon not some company I've never heard of. It's if course never gonna happen. It's too convenient for Amazon and most users probably don't even understand that they aren't buying from Amazon.

> It's if course never gonna happen. It's too convenient for Amazon and most users probably don't even understand that they aren't buying from Amazon. Out of curiosity, why do you assume people don't realize they aren't buying from Amazon?

Whenever I tell anyone "I sell on Amazon," they are really confused and say "what? How can you, Joe Random, sell on Amazon?" When I try to explain it's a marketplace "like eBay," where anyone can sell things on, they get confused and can't understand.

Even if some people understood "some inventory comes from different places," the vast majority of retail operations vet their sources and have quality control. So it's easy to understand why people would assume it's not different from Target selling inventory provided by multiple distributors.

Re: Amazon's Fake Review Problem

#276
post #11

Amazon is starting to feel more and more like eBay. The same product sold by multiple vendors under different names, questionable reviews, clearly fraudulent sellers and so on. I used to love Amazon but now when I look for something I dread to do so on Amazon because I don't feel I can trust anything there.

I had this realization the other day. I was buying something that's currently out of stock most places. I looked on Amazon, saw some "3rd party" sellers that said they would have it on stock various dates about 1 week in the future, and I didn't believe them. Some sellers had hundreds of positive reviews, others had only a few, most were an "appropriate" price (not to high or low), and all of them had the prime desig…

IIRC, when my listing says "back in stock on x day" it's caused by Amazon moving my inventory between wearhouses. I don't know if you can manually mark your listings like that though.

Re: Amazon's Fake Review Problem

#277

Earlier quoted context omitted.

If Fakespot is able to spot the fake reviews, Amazon could have also utilized the similar tech. What stops them?

Profit they make from extra sales due to fake reviews? Why do you assume fake reviews are bad for Amazon bottom line?

When a customer is burned by a bad product with excellent reviews, especially more than once, ordering from Amazon could quickly become more trouble than it's worth.

Re: Amazon's Fake Review Problem

#278
post #51

Earlier quoted context omitted.

How would you go about tracing the steps? Are there some large database of OEM factories where you can contact them in English (assuming China). I too would love to know exactly where and who made it, how many middle-men have stepped in with their own markup to finally Amazon. Because it's clear that Amazon doesn't give two shit about who and what gets sold, they just care about volume. Fixing reviews is not part of…

With great difficulty. Most major brands never even touch their products. They design them and have them built by contract manufacturers. They then get sent through distributors, shippers, brokers, etc. Each of which has terribly old systems which may or may not track each individual item. Most links in this chain are also disinterested in any third party knowing what and how many products they move, including and es…

I think your last paragraph is absolutely key - many players treat that information as strictly proprietary. A big reason is because oftentimes a supplier makes identical (other than the packaging) products for two different customers, one sold as generic, one name brand.

Re: Amazon's Fake Review Problem

#279
post #3

I don't care much about the reviews, but I care a lot about the fake / phony / not-as-described merchandise: http://seliger.com/2017/01/09/tools-continued-careful-buy-am... . I've actually changed behavior and tried harder to buy from the original maker, if possible.

Unfortunately all goods at Amazon are intermingled including those Amazon itself ships. The only exception is companies who specifically pay not to do so in which case you lose access to prime shipping.

Amazon seller here.

This is not strictly true, you can still have your orders shipped "prime" but not commingle.

Re: Amazon's Fake Review Problem

#280

I am much more wary about buying from Amazon than most of my peers, who easily click and buy on a whim or out of habit. I went to buy a toy at a brick and mortar, and compared the price to Amazon while there - $12.99 in the store and $27.xx at Amazon! I think they've marketed them to be the low price leaders but that's definitely not the case. I'm also concerned about counterfeit products - I don't care as much about…

Most toys sold by third parties are overpriced on Amazon in December.

December is also a great time for coupons and sales from traditional retailers (you can still order online) so you're doubly getting hosed.

Post reply on HN