Discussion Novel ratings are skewed

I'm not advocating for a change to the way the system is presented. The reality is, this change is never going to happen. But I want to point out that the ratings of novels aren't indicative of the overall quality of the novel. The major flaw comes from ratings based on a chapter by chapter basis. If you scroll through many of the popular novels, you will notice that a lot of the good reviews come people who haven't entirely finished the book. In fact, for people who've completed novels and rated them, the average rating is lower. The rating you see on the main page is an average of all the ratings. There are no metrics set to weigh rating categories with less reviews. It's a very simple system. Clearly, this is flawed because there are less people giving wholesome reviews of completed novels which makes the reviews biased towards readers who are still reading at translation speed. It's clearly a money making scheme, by both the translators and the authors. I'm merely here to shed light on these details to the uninformed majority(I'm sure many of you already know this fact but there are a lot that don't). As for the fix, an easy way would be to allow readers to filter completed reviews from ongoing reviews.

 

No shit sherlock

 

Its common sense dude that the ratings is bias.
Just having a harem tag or netorare tag will instantly get 1star rating from most of the internet fags.

Thats why i dont even try to look at ratings.

 

You foolishly and arrogantly underestimate the True Nuffians.

Besides, nobody pays attention to the specific rating, just the map of the 1-5 stars. Goalposts means it's either a QI novel or a very polarizing one (yaoi, rape=love, dumbass-antihero, etc)
Slope down means crap, slope up means maybe decent, and mostly 5's means wide appeal in it's genre.

 

Ratings and reviews dont mean shit, just ignore them.

 

yes. skewed.
the only reference you can get from the ratings is how many peopke liked it or hate it in general~

 

Not in danmei

 

The only times where I visit reviews section of a novel is to find spoilers. Other than that it’s pretty damn useless.

 

There is some parameters missing : not only there is the bias about the rating before the end, but there is another one:
Every category of novels (not only harem) has it's own public, and notation occurs only amongst it, fans ir haters of this category, so there is no objective vote about quality, or very few.
Moreover, the pool of reviewers (a fraction of nuffians) is very restrain, if you take in consideration all the people who use novelupdate without inscription.

 

well, my personal suggestion is to disable rating if chapter released is below 50
it doesnt make sense a novel has a 1 star rating when there is only prologue released

 

It's pretty obvious that the ratings are skewed, in addition to that, people actually dont know how to review things properly(even me). They mostly don't address things like character interactions, plot structure, dialogue(basically the "meta" of a novel, the stuff that makes up hamburger rather than the hamburger itself(the book)), etc. Only some of these things are actually mentioned and when looking over reviews, the ones that actually try to look at some of these things get a like from me. Everyone else just sounds like "this is the best because i enjoyed it" and "this is shit because i didnt like it". Badly written things can be enjoyed( Invisible Dragon for one) but it won't change the fact that it is bad in terms of quality.

Also, translators aren't really making money due to ratings(it helps though to get new readers). Rather, they make money off donations/ad revenue on the bigger sites. Not to mention Qidian's paywalling. For small groups or lone translators, i think that they are hardly making/dont get money at all.

Nowadays i hardly give a crap. You're looking to entertain yourself, if the novel you're reading isn't doing that then drop it and go somewhere else. Though if later on something you dropped feels like something you would want to read again/feel entertained by then read it.

 

Overall my novels are rated at aroung 3 or so as an average but honest and not, they're better than crap rated 4+.

 

me too

 

I think the average score is good indicator of the quality of the novel. There are a lot of exceptions, but more aren't. To understand whether a novel is an exception, I look for thorough reviews by people who seem to have read a paper book before and now what good writing is: style, structure, character development, ... Some cheesy novels have higher ratings than I'd ever give them, some well written novels are seen as boring. But that's where reviews come in, at least for me.

I fail to understand your money making argument.

 

I think that even the average rating score is pretty useless. One of the biggest problems is that these are usually given by people who have read very little of the book - often less than 10% of it. With so little relevant exposure to the novel in question, just how valuable are their opinions? And yet they're given just as much weight as someone who has read the whole thing and is able to give out a proper impression of it.

 

If the novel is chinese and by qidean then you can expect the rating to be inflated. Just take any of their rating and cut it by half so it remains true. If you need to wait 100 chapters for it to be good then you are an idiot who has wasted too much time on the novel so you try to delude yourself into believing it got better. Same goes for crap that goes on further than 400 chapters because in my experience any chinese web novel over 400 chapters turns to crap no matter what.

 

Addressing your final point, the inflated average ratings are proof of this. If you compare the review of someone who has finished the entire novel next to the average rating, there's usually a remarkable difference of around 0.5-1.5 stars. Qidian and novelupdates haven't taken any measures to give readers a sorting option that would allow readers to sort completed reviews from uncompleted reviews. Typically, the higher ratings come from readers who have only read a portion of the novel(the beginning of the novel is always better because the author wants to attract readers). My argument is that if readers were given an accurate rating based off completed reviews, there would be less readers due to the unattractiveness of a lower rating. However, the caveat to this is that there aren't many people who write completed reviews. The rating system is inherently flawed, giving these companies a slight edge when it comes to earning money.

 

Okay, now I get what you mean. I still don't think this is intentionally done to make money off of. How many money would they actually make off of that? And it doesn't take much effort to look through reviews on NU anyway. I don't think this is a big issue. It's maybe a nice feature request for NU though, I suggest you post it on the relevant forum instead of complaining about it here. If it hasn't been done already.

I think saying that the beginning is always better is offensive. You're saying all writers are cashgrabs. That's not a phenomenon unique to webnovels, and there are writers trying to write quality stuff too ... You're doing a disservice to authors.

What you can see in the first 10%: writing style, how well the beginning part is written, beginnings of plot and character development. It can be surprisingly indicative. It might be noisy data for some novels, but not irrelevant data. And there's the law of large numbers, although a significant part of novels probably don't reach a large enough sample size of reviews. Anyway, despite all the noise on the data, the average still provides some metric for the appeal of the novel. Data can be really noisy and still contain useful information. But actually, it's not even that bad on NU.

Personally, I ignore <3.5, I consider >4.0 and strongly consider reading >4.5. I add in tags, synopsis and some reviews and then I am, let's say, 80% sure whether I am going to enjoy the novel or not. This system worked out very well for me. See? Some people have positive experiences!

So you are one of those people. You're part of the problem. Ironically, your post is liked by OP which makes it even funnier. It's chinese? Bad. It's long? Bad. It's from Qidian? Bad. That's ridiculous, you judge novels by stereotypes. Further than 400, it turns into crap? Where are you pulling that one from. In my experience with Chinese novels, if it's bad I notice it within 50-100 chapters, not 400. Take Against the Gods for example. It's not very good, but it's never really been exceptional even from the beginning.

I can't say I enjoyed a lot of Japanese novels (or Korean ones for that matter). The better ones I've read, I remember Rakuin no Monshou, were good adolescent novels but were average in writing style. Like Horowitz's Alex Rider, they're nice reads very strong on the action part. I can dig that.The rest? Cringy writing style and dialogue with stutter, just as much infested with tropes as Chinese gutter novels.

On the side I'm reading the (paper) novel Tigana by Guy Gavriel Kay - it's a pretty long one, 750 pages or so, and the first 250 pages were world and character building with a little stint of action. You say you'd be bored out of your mind. Does that mean it's bad? No, it means that's not your cup of tea. Some people do have patience for a story to unfold, you know. It's like you've only ever read series like the Ranger's Apprentice or Alex Rider (adolescent action novels) and say that Dostoyevski's novels are bad. It doesn't appeal to you, but others do want world and character building. There are actually a few writers trying to write good novels, even in webnovel format! Imagine that! And they can be Chinese! Conveniently ignore those, sure, go on.

In your experience - well, that's just your experience. No less, no more. Your post reads like you just read through all translated chapters of Martial God Asura.

I dislike the dismissive tone of this thread. Average ratings are just another tool in the box. We often read incomplete translations. The ones that do have read entire novels often did it in the original language, but they're scarce. You do need some sort of rating system, right? What exactly are you trying here? Making a point? Well, everyone already realized it's imperfect. You're saying it's a capitalist scheme? Throw your tinfoil hat in the garbage bin. Suggest a solution by adding a feature suggestion on NUF, that would actually be productive. I'm out of this shithole of a thread. I didn't see a solid, well constructed argument anywhere ...

 

I feels you man, i dont know about the money making scheme cause well, its make my head hurt to think about complicated stuff. But well when i read the novels and the content is not matching with the synopsis, rate and the review, i can only shake my head then say "The heck, this is a scam"

 

Ouch did i touch a nerve? I took out the chinese part and said any web novel because you cannot maintain good consistency over such a long period of time. Web novels and actually novels are two different things because one actually involves planning with editors while the other is written on the fly. I said ratings by qidean novels are inflated due to alt accounts and biase. Also I'll be hear if anyone wants to complain to me about my opinions because I'm not a kitty cat who runs from people talking bad about me or my novels. *Im talking about LoseByLeaving