What is review bombing, and how is it recognized?

[Tyndmyr]

To me, legitimate reviews are one per person, that they are not bribed to make.

Someone may like or dislike a show for any number of reasons that you may or may not agree with. The two sources of illegitimacy are payments for a review, either good or bad and multiple reviews from one person, skewing the supposed average.

I cannot imagine that many people will, for no pay whatsoever, make many accounts and type out many reviews. It seems as if the first problem is much more likely to be an issue. After all, while I have no issue commentating on movies, and will cheerfully shred a film I find to be lame, I've never bothered to review any movie on RT. After all, there is no conversation to be had there. Doing so over and over again sounds like a tedious chore.

Bought reviews are almost invariably positive, and exist for many products. Amazon has quite a number of these for some products, and they look well, halfassed and samey, for a dodgy product. You can generally tell. Usually nobody pays to make their product appear worse, though.

Is it possible the reviews are unrepresentative of the casual viewer, who's willing to tune it because it's on their subscription but don't love it enough to write a review? And the disproportionately negative tone of the fan reviews might drive them away.

Possible. The easy way to test it would be to see how rapidly viewership falls off. Lots of people losing interest in the show within the first couple of episodes would be a pretty telling metric of broad popularity. Companies have access to this information, and if they are not utilizing it, it's probably because it is unfavorable to them.

Audience scores and reviewer scores diverging isn't intrinsically a sign of bad actors. Reviewers are not randomly selected, and they commonly look for different things than the general public does. I mostly don't care at all what the reviewer score is, anymore than I care what number a video game magazine assigns to the video game.

For Rings of Power itself, the rating doesn't strike me as wild. It is a mediocre work. Not terrible, not amazing. All that is required to explain the reaction is upset fans hoping for the quality of the original films.

[pendell]

You're right. That was an error on my part -- I'd quoted you when I meant to quote bloodsquirrel. I'll edit my post to make it more clear.
EDIT: And it was actually Gnoman I was responding to:

The biggest signal of review bombing is the distribution. Any normal review pattern is going to cluster around 2,3, or 4 (assuming a 5-star system) depending on how good the product being reviewed is. This means that if you see something with hugely disproportionate number of 1s, it is a massive red flag.

As I said, I'm not sure that's valid given how 1 and 5 star reviews are weighted. Sigh.

Respectfully,

Brian P.

[Batcathat]

To me, legitimate reviews are one per person, that they are not bribed to make.

While it shouldn't be necessary, I'd add "that has actually seen/read/played/used what they are reviewing". I think at least some sites have taken steps to prevent it, but it used to be very common to see a movie on IMDB with a thousand 1-star reviews that wasn't even out yet, because of some controversial casting choice or whatever.

Not liking a work for some sort of meta reason can still be valid, of course, but as a potential viewer those "reviews" don't really help me much.

[Tyndmyr]

While it shouldn't be necessary, I'd add "that has actually seen/read/played/used what they are reviewing". I think at least some sites have taken steps to prevent it, but it used to be very common to see a movie on IMDB with a thousand 1-star reviews that wasn't even out yet, because of some controversial casting choice or whatever.

Not liking a work for some sort of meta reason can still be valid, of course, but as a potential viewer those "reviews" don't really help me much.

I would not. Trailers are not the whole work, but they are a portion of the work. A person can view only a portion of the pilot and decide they hate the work, and that is quite legitimate.

Trailers are literally cherry picked to be the most attractive introduction to them. Is it not possible that someone can see that and have a genuine reaction of dislike?

Their likes and dislikes may not be the same as mine, but that does not make them false. Perhaps someone despises a specific actor. Oh well, that's part of life. The fame of actors and actresses is counted on by studios for making money and attracting viewers, surely dislike of them is equally as legitimate as like.

[Bohandas]

If you believe that reviews should act as a means for customers to exert power in solidarity against companies, sure. Though even then, campaigns to get people to leave reviews exert undue influence compared to spontaneous reviews, so then you get to deal with the fact that it's more like influencers and not net owners getting to use the power of customers to make companies act the way they want.

If you believe reviews are there to provide information about the product or service as the customer should expect to experience it, this kind of thing destroys that signal. Especially in cases where the aggregate review controls visibility of the product to others in the first place.

If there's a review campaign to bomb a game because e.g. they haven't done a translation to Chinese, or because they asked a famous Twitch streamer to take down a video and all that person's fans bombed in response, or because the company declared that the game's sequel won't have VR support (even though this game does), or because there was a scandal that one of the employees at the company posted a sexist joke on Twitter, or because the company released a DLC that cost more in Brazil than in France... Negative reviews inspired by those things do not inform people about the contents of the game. They just inform people about whether groups able to mobilize coordinated action feel like rewarding or punishing the company.

If that's really what you want to be participating in when looking at reviews, well I guess you get to have that. But the point of calling review bombing out and making analytical tools to identify it is that not everyone wants to get pulled into other people's crusades - lots of people just want to know whether they'll enjoy the game or not. And protecting review bombing from being identifiable as such is harmful to those people.

No, I'm saying that if the company is going to harass you for playing their game wrong, as in the modding example, that is, in my opinion, important information about the game reasons why you, personally, might not want to but the game because it represents a liability to the buyer. The only one of your examples that is really comparable to the suing a modder scenario is the twitch streamer one.

What if the game was infected with spyware or viruses? It doesn't affect the gameplay. Should it be left out of the review?

or because the company released a DLC that cost more in Brazil than in France

Price gouging is definitely a valid concern. A game that's great for five dollars is a ripoff for twenty

[pendell]

I would not. Trailers are not the whole work, but they are a portion of the work. A person can view only a portion of the pilot and decide they hate the work, and that is quite legitimate.

Trailers are literally cherry picked to be the most attractive introduction to them. Is it not possible that someone can see that and have a genuine reaction of dislike?

Their likes and dislikes may not be the same as mine, but that does not make them false. Perhaps someone despises a specific actor. Oh well, that's part of life. The fame of actors and actresses is counted on by studios for making money and attracting viewers, surely dislike of them is equally as legitimate as like.

That's a review of a trailer, though, not the work itself. I'd go with the phrase "Don't judge a book by its cover" and insist a reviewer have seen at least a significant portion of the film. If they fall asleep three hours into a four hour snoozefest, I'll accept that. But judging a work by trailer content, which is often misleading and may contain footage not even in the film, doesn't seem very sound.

Respectfully,

Brian P.

[warty goblin]

Eh. The decision that matters is 'will I regret spending money or time on this thing?', and people can certainly have repeat experiences with looking at review data, making that decision, and seeing how it turns out. You don't need a detailed mechanistic casual model in order to improve the accuracy of one's inference there.

Calibration for example. If you find yourself enjoying things with 20% positive reviews 60% of the time, things with 50% positive reviews 40% of the time, and things with 95% positive reviews only 80% of the time, that tells you something about how you should use reviews without having to know why.

Things don't have objective ground truth ratings after all.

Right, if my goal is to decide whether or not to buy something, reviews can be useful as a predictive score function. All that requires is a correlation between thr score and my opinion.

They're still useless as a means of understanding the viewpoints of a population, except narrowly defined as the group of people who left a review. It seemed to me that a fair number if people were trying to goose user reviews into being just that.

(It's also worth distinguishing between an average user score, and the actual contents of a review. If I know I have wide feet relative to the population, I care more about people saying this mske if shoes runs wide than I do how they rate it because of that.)

(I suspect it also matters a lot if the object being reviewed is basically a commodity which is exchangeable with any other similar object except for matters of price and broadly understood ideas of quality. Very few people are deeply passionate about, like, particular brands of sheet copper, they just want copper of uniform thickness at a price comparable to other piecesof sheet copper, so a five star review is pretty good evidence it is basically that. For TV though, every single person's rating is going to be extremely idiosyncratic and TV show A is not generally exchangeable with TV show B; I can't just watch the other Lord of the Rings show).

So you're saying the user reviews aren't valuable? Why does the company bother to collect them? Purely for their anecdotal value?

Companies collect and show customer reviews because good customer reviews drive business. You can probably extract some useful features (e.g. 23% of positive reviews mentioned the comfortable grips, 52% of negative reviews complained about the manual) using natural language processing or other data analysis techniques, but this is, again, limited by the flawed sampling mechanism. Maybe your product has a flaw rate of like .01%, but like half your negative reviews are about quality issues. This maybe indicates that you need to work on customer service to resolve those complaints, but it tells you far, far less about your flaw rate than sampling a thousand items off the assembly line.

I would hesitate to call the data useless; the fact that more than half of the 21957 people who bothered to write a user review downchecked it must tell us something. But I agree that it's not a "something" which would be valid statistically; the sample is self-selected and biased.

It tells you that more than half of the unknown group who left a review left a negative review. If you think that you are similar to that group, that may be informative as to whether you should buy/watch the product. As I said above, all this requires is a correlation between score and your taste; predictors don't have to be inferentially or causally valid. If you have specific interests or concerns, you can troll through the reviews to see what people are saying about them, which is really just trying to build a better predictive score.

If you actually make the show or product though, drawing inference from this sort of data is really dangerous, precisely because it is biased in unknown ways. For example there's a long history of hardcore full loot PVP MMOs, often made by and for dedicated fans. Most of these do badly, because the hardcore people who post on forums want these features, but the vast majority of non-hardcore players who aren't on the forums don't. This is why companies do market research with actual sample frames and stratification and other sampling methods.

As to what people or viewers of the show as a whole think? It tells you pretty much nothing. I understand the urge to want it to, not knowing is frustrating, but if the data is awful the principled and defensible answer is "I don't know" not "let's apply my existing biases to this data until I think it's useful."

Look at it this way, you wouldn't want to conclude anything about rates of alcohol use, and how people feel about drinking in the general population based on a dataset taken from an unknown mix of people outside a liquor store, an AA meeting, a frat party and a Mormon church, right? This is exactly that data, except even worse because you don't even know which groups it's being oversampled from.

[Batcathat]

I would not. Trailers are not the whole work, but they are a portion of the work. A person can view only a portion of the pilot and decide they hate the work, and that is quite legitimate.

Trailers are literally cherry picked to be the most attractive introduction to them. Is it not possible that someone can see that and have a genuine reaction of dislike?

Their likes and dislikes may not be the same as mine, but that does not make them false. Perhaps someone despises a specific actor. Oh well, that's part of life. The fame of actors and actresses is counted on by studios for making money and attracting viewers, surely dislike of them is equally as legitimate as like.

Yes, it's entirely legitimate to have an opinion about a trailer or an actor but it's less meaningful to me as someone wanting to know people's opinions about the work as a whole, rather than than a tiny piece of it. If "block reviews based on how much someone loves/hates one actor in the movie" was an option I'd absolutely take it.

Personally, I really like salty food and the presence or absence of salt can give a decent hint about whether or not I'll like a dish. But should I really start reviewing restaurants I've never visited based on the saltiness of their menu?

[tomandtish]

I do think it is interesting to note that as of this post apparently NO ONE has left a review of it on Amazon.

I did take a quick look at Wheel of Time reviews. Of the 30,166 posted 47% are 5 Star and 36% are 1 star. Those two therefore cover 83% of the reviews. And variations on "woke", "political agenda", and similar language dominates the 1 star reviews. There's one 1 star review in the "Top 10" 1 star reviews that is by a person who flat out states they haven't watched it.

[Tyndmyr]

Price gouging is definitely a valid concern. A game that's great for five dollars is a ripoff for twenty

That seems reasonable. Mulan had some controversy over the price point when it was released, because it was originally set at $30. Fans thought that was simply too high for a rental, and certainly it was a price point that was not terribly normal for streaming.

I can see someone reacting poorly to that announcement...and that reaction would be genuine. Someone for whom cost is no object might not share that reaction, but if many people feel that way, surely it is a legitimate complaint.

That's a review of a trailer, though, not the work itself. I'd go with the phrase "Don't judge a book by its cover" and insist a reviewer have seen at least a significant portion of the film. If they fall asleep three hours into a four hour snoozefest, I'll accept that. But judging a work by trailer content, which is often misleading and may contain footage not even in the film, doesn't seem very sound.

Respectfully,

Brian P.

Books are judged by their cover all the time, and rightly so. If the publisher believes this book is of such low caliber as to not warrant time or attention on the art, well, that seems like a perfectly reasonable cause to place it back on the shelf and look for something else. If the author, the genre, or the description fail to interest you, put the book back, and find another.

You should absolutely judge a book by its cover, and a movie by its trailer. That is the purpose of covers and trailers.

I will agree that relying on only the *number* of the review, and not the actual contents is certainly inaccurate. If many people hate it, you should probably find out why. Perhaps the reason matters to you, perhaps it does not, but a few seconds of reading will inform you either way with more detail than any number can provide.

[pendell]

I do think it is interesting to note that as of this post apparently NO ONE has left a review of it on Amazon.

That's because Amazon disabled the reviews . It's their company; apparently they don't want to host a lot of negativity regarding their own product. Still seems an admission of weakness, though. Either that they mistrust the reaction, or that they don't believe they can filter out a reviewbombing campaign.

Respectfully,

Brian P.

[tomandtish]

That's because Amazon disabled the reviews . It's their company; apparently they don't want to host a lot of negativity regarding their own product. Still seems an admission of weakness, though. Either that they mistrust the reaction, or that they don't believe they can filter out a reviewbombing campaign.

Respectfully,

Brian P.

I was just getting ready to edit my reply when I found that out. Yeah, that's not a good sign regardless of the reason.

[NichG]

No, I'm saying that if the company is going to harass you for playing their game wrong, as in the modding example, that is, in my opinion, important information about the game reasons why you, personally, might not want to but the game because it represents a liability to the buyer. The only one of your examples that is really comparable to the suing a modder scenario is the twitch streamer one.

What if the game was infected with spyware or viruses? It doesn't affect the gameplay. Should it be left out of the review?

Price gouging is definitely a valid concern. A game that's great for five dollars is a ripoff for twenty

Utility of a given review for a certain propose and legitimacy of an opinion are totally different things. You can have completely valid concerns about something, and yet prioritizing those concerns in a certain context over other information can be harmful to others' ability to use that context for their needs.

So whether it's appropriate to, say, give people a tool to filter out certain reviews from their own decision-making is based on the overlapping interests of the review site, the users of that site, and for things like storefronts the content producers as well.

The store wants people to be receptive to it's advertising, so it doesn't want to have people buying things they will end up regretting or returning. The content producers want people to feel good enough about the things they make to make the decision to buy them. The users of the site ostensibly want to not waste money on things they'll regret buying.

There are also other interests that may be less cooperative. Users may want to force a company to change things about what they do. Producers or the storefront itself might want to make a sale even if the buyer will regret it. And the storefront might want to encourage producers to give them a bigger cut or things like that by e.g. letting them buy review moderation or something. However, if the arrangement is stable, everyone in the interaction has some ability to make it so that in the long run it's worse to prioritize the non-shared interests than to prioritize the shared interests.

So any question of whether a review is 'appropriate' has to be in context. Does it serve those shared interests?

From the user point of view, reviews give them the opportunity to find out details about the product that you would only find out at some cost to yourself otherwise. Anything public like 'the price of the game' is not new information. Having your computer slow to a halt because of spyware could easily be new information.

Why this 'new information' thing is relevant...

Let's say I have a biased coin and I want to estimate the bias. I get more information in N flips if they're actually independent flips than if, say, there's a 50% chance each flip to receive the same result as the last regardless of the bias. Furthermore, if I believe that the flips are independent but they're actually correlated this way, I will strongly overestimate my confidence in any guess I make about the bias.

That effect of correlated reviewing is what diminishes the ability for the reviews to be informative, regardless of the specific review contents.

For example, 100 people saying independently 'the plot was bad' is more informative than if someone writes a popular article about how bad the plot was, and 100 people read it and say 'yknow, this guy is right, the plot was bad' and leave corresponding reviews.

That doesn't mean that plot is an invalid thing to review. It just means that reviews which are induced by external drivers contain less useful information than spontaneous organic reviews. And that's a statistical property that is independent of the content of the opinion itself.

[crayzz]

Books are judged by their cover all the time, and rightly so. If the publisher believes this book is of such low caliber as to not warrant time or attention on the art, well, that seems like a perfectly reasonable cause to place it back on the shelf and look for something else. If the author, the genre, or the description fail to interest you, put the book back, and find another.

You should absolutely judge a book by its cover, and a movie by its trailer. That is the purpose of covers and trailers.

This is why Daikatana is my favourite game. I mean, I never played it, but the trailer and promotional material seemed so cool, it easily clinches number 1.

[Blackhawk748]

Audience scores and reviewer scores diverging isn't intrinsically a sign of bad actors. Reviewers are not randomly selected, and they commonly look for different things than the general public does. I mostly don't care at all what the reviewer score is, anymore than I care what number a video game magazine assigns to the video game.

Its actually incredibly common and has been happening since the literal dawn of cinema. Dracula (1931) got panned by critics but general audiences freaking loved it. You can find literally hundreds of movies that had this happen, a none small portion happen to be horror films.

This is why I prefer to get my reviews from Youtubers, simply because the ones I watch fully explain what and why they liked or didn't like in the film and I can find someone who matches up roughly with my wants in media.

[Rynjin]

I would not. Trailers are not the whole work, but they are a portion of the work. A person can view only a portion of the pilot and decide they hate the work, and that is quite legitimate.

This is a pretty unpopular opinion among fans of a particular work, unfortunately. I had someone tell me the other day I wasn't allowed to have an opinion on Bleach because I only read about 90% of it before dropping it in disgust.

[MetroAlien]

This is a pretty unpopular opinion among fans of a particular work, unfortunately.

To play devil's advocate, there's a difference between saying "I watched the whole thing. It's bad. Here's why...." and "I was turned off by XYZ, so I didn't watch all of it / at all"

For instance, I couldn't sit through whole episodes of Cowboy Bebop on my own, because I thought it was boring. But when I watched the whole thing with a group, I discovered some parts that I liked more than the beginning of the series led me to believe.

My opinion as a whole hasn't really changed, but no instead of telling people "It's boring" I tell them "I found it hard to sit through the slow start"

[Liquor Box]

I think someone needs to have seen it to be qualified to review it.

I don't think I would trust a review or recomendation/condemnation based on someone watching a trailer alone. If someone had watched half or more, but not continued from that point because they didn't like it, I think their views carry less weight than someone who watched the whole thing.

[Bohandas]

I think it depends whether a show is serialized or episodic. If it's serialized than you can certainly judge it based on just the start (or any other isolated part), because unlike an episodic series you can't easily just skip parts that you don't like

[Rynjin]

I think someone needs to have seen it to be qualified to review it.

I don't think I would trust a review or recomendation/condemnation based on someone watching a trailer alone. If someone had watched half or more, but not continued from that point because they didn't like it, I think their views carry less weight than someone who watched the whole thing.

Honestly, I really don't see how. "This product was so bad I couldn't even finish it" is a perfectly valid opinion, and just as "weighty" as someone who says "I watched all of it and it was fine". Perhaps moreso, given that for longer form entertainment (a series of novels, a manga, a long-running TV series) there exists a sort of sunk cost fallacy where a lot of people might not actually like the series but feel they have to finish it because they already read/watched a ton of it.

This is exactly what I was talking about earlier, by the by.

[Batcathat]

I think there's quite a difference between, say, watching five hour long episodes out of ten and writing a review that it was bad and watching a minute long trailer for the same show and doing the same thing. While trailers are (or at least are supposed to be) representative of the content, they are a really tiny part of it. Basing an opinion on a trailer is fine and kind of the point of them ("This movie looks bad so I'm not going to watch it") but reviewing and rating the entire movie based on it seems dishonest to me.

[Mechalich]

I think it depends whether a show is serialized or episodic. If it's serialized than you can certainly judge it based on just the start (or any other isolated part), because unlike an episodic series you can't easily just skip parts that you don;t li,e

I think you can judge both based on the start, but a weak start matters more in the case of a fully serialized show because you are obligated to watch the early not-as-good material and can't just skip ahead to the good stuff. For example, in the case of Star Trek: TNG, you can watch the pilot and then skip straight ahead to season two and miss hardly anything. But in The Expanse that's not going to work. Both shows are noted for having weak opening seasons, but the prospective viewer only has to fully embrace the sunk cost in one.

Honestly, I really don't see how. "This product was so bad I couldn't even finish it" is a perfectly valid opinion, and just as "weighty" as someone who says "I watched all of it and it was fine". Perhaps moreso, given that for longer form entertainment (a series of novels, a manga, a long-running TV series) there exists a sort of sunk cost fallacy where a lot of people might not actually like the series but feel they have to finish it because they already read/watched a ton of it.

It is possible, though admittedly fairly rare, for some long-running piece to dramatically change for the better in the latter half or even to incorporate a final act plot twist that completely reframes the first half. This also matters more in certain genres than others. For example, in a whodunit, the ultimate resolution of the central mystery is paramount. Generally any story where the ultimate resolution is in tension until the absolute climax puts more weight on having reached the work's conclusion. This is not usually the case in the action-oriented melodramas with the fate of the world at stake because it is generally not actually in doubt that the world will survive.

[Liquor Box]

Honestly, I really don't see how. "This product was so bad I couldn't even finish it" is a perfectly valid opinion, and just as "weighty" as someone who says "I watched all of it and it was fine". Perhaps moreso, given that for longer form entertainment (a series of novels, a manga, a long-running TV series) there exists a sort of sunk cost fallacy where a lot of people might not actually like the series but feel they have to finish it because they already read/watched a ton of it.

This is exactly what I was talking about earlier, by the by.

Well, because the person who watches the whole thing and says they don't like it consumed the whole product before making that decision. You sampled only a portion.

Game of Thrones is a good example. It is widely regarded that the first 7 seasons were top tier television viewing, but that season 8 was poor lowering people's assessment of the series as a whole. A person who watched only 90% (or 87.5% to be more exact) would tend to have a different opinion from those who watched all of it - and a less well informed opinion because they did not view the whole product.

[NichG]

"I couldn't finish it" is relevant information to whether I'd want to start it, that I'd have to spend time myself to find out.

"I heard it was bad so I didn't start it myself, but I have opinions about stuff I've heard about it" lacks that property of having relevant information.

I'd happily filter out reviews from people with less than a half hour of play time of a game, but not for having less than 2.

[Rynjin]

Well, because the person who watches the whole thing and says they don't like it consumed the whole product before making that decision. You sampled only a portion.

Game of Thrones is a good example. It is widely regarded that the first 7 seasons were top tier television viewing, but that season 8 was poor lowering people's assessment of the series as a whole. A person who watched only 90% (or 87.5% to be more exact) would tend to have a different opinion from those who watched all of it - and a less well informed opinion because they did not view the whole product.

Game of Thrones is an extreme outlier when it comes to television, however*. Most series don't start off good to great and then immediately fall off a cliff 7 seasons in. And most series, likewise, don't start being good 7 seasons in either.

If someone watches 50% of the available material and says "it ain't good, don't waste your time" you better believe I'm more inclined to believe that opinion. Life's too short to consume garbage media.

*At least in the general public sentiment. Apropos of the conversation, I thought the series started to suck about the start of season 5 and stopped watching in the middle of it. There's a pretty steep drop off in quality from the exact instant they run out of book material to work with.

[veti]

Game of Thrones is an extreme outlier when it comes to television, however*. Most series don't start off good to great and then immediately fall off a cliff 7 seasons in.

Well, "7 seasons" is a bit - precise, but the general pattern isn't that unusual. I'm old enough to remember Frasier, Spin City, The X Files... To say nothing of Big Bang Theory.

"Series that started well then fell apart" is quite a popular theme for threads on places like Reddit. After a while, the producers' desire to milk a franchise outruns the writers' and actors' ability to sustain the characters and situations that make it a hit.

[Rynjin]

Well, "7 seasons" is a bit - precise, but the general pattern isn't that unusual. I'm old enough to remember Frasier, Spin City, The X Files... To say nothing of Big Bang Theory.

"Series that started well then fell apart" is quite a popular theme for threads on places like Reddit. After a while, the producers' desire to milk a franchise outruns the writers' and actors' ability to sustain the characters and situations that make it a hit.

Yeah, but usually it's a pretty slow decline, and is often preceded by a slow climb as well; eg. the first AND last seasons of TNG are the worst.

And I'd say if someone decided to drop off a series 4 seasons in, that is in itself telling.

[snowblizz]

That's because Amazon disabled the reviews . It's their company; apparently they don't want to host a lot of negativity regarding their own product. Still seems an admission of weakness, though. Either that they mistrust the reaction, or that they don't believe they can filter out a reviewbombing campaign.

Respectfully,

Brian P.

Amazon knows the results won't be good. They have already seen what happened on other platforms they can't control fully like Youtube, where their trailers and teaser have been I believe the kids say "ratioed" even though efforts are made to clean out the negative commentary.

[Bohandas]

Game of Thrones is an extreme outlier when it comes to television, however*. Most series don't start off good to great and then immediately fall off a cliff 7 seasons in.

30 Rock also did this. Six seasons great, and then boom, right off a cliff the seventh season.

[Willie the Duck]

Yeah, but usually it's a pretty slow decline, and is often preceded by a slow climb as well; eg. the first AND last seasons of TNG are the worst.

There have been some examples of Good... Good... Good... Ohmygodmyeyes!Whatwereyouthinking? Moonlighting -- they got together and then... had nothing left to have tension over. Mad About You -- five brilliant (if incredibly 90s) seasons of couplehood dynamics right up until they had a baby and... did not make the transition into a parenthood series. I don't necessarily think Dukes of Hazard was a masterpiece, but the series was very much itself (and wildly popular) right up until they switched out the leads. GoT is unusual, but certainly not unique.

Regarding the main topic -- I don't think there's one way to define or detect review bombing, nor a clear and universal right or wrongness to it. All it tells me is that an online public venue with low entry-bar to contribution of voices-heard is not going to be representative of the larger population (making it less-than-of-perfect use to the wider population, who nonetheless use it as a source), which I already knew. We have had plenty of examples of this already -- a small dedicated group of volunteers at Wikipedia have disproportionate control over the output; people who treat Star Wars, Ayn Rand, or the Greatful Dead (or Tolkien) fandom as a lifestyle are waiting in the wings for voting to open on something like PBS' list of most top 100 most influential movies/books/bands of all time; heck, just look at the roleplaying section of this and similar boards.

Regardless, it highlights that many of these places to publicly share ones' opinion on a product work best if left to just be that, and when used as some larger representation of how good the thing is (or how the larger population does/would think about the thing), run into the same 'do I trust this source?' issues that previously existed with movie critics and similar since who knows when. Maybe the best lesson is that a 1-5 or percentage metric is a bad measure of an entertainment product. I've always preferred reviews that were written essays or videos discussing what someone thought about a movie or show. Even if I don't trust Richard Roeper or Doug Walker or the like to be unbiased, if they have to make an argument for their position, I can view that and decide if this time I agree or not.