Ratings - too many or too few stars

So basically like the title says, I feel like the ratings of a lot of novels are unreliable because people only really use the 1 or 5 stars when they rate things. At this point I'd prefer if they simply removed the stars and just gave us the option to say "good" or "bad" since that's the only thing we seem to be doing anyway.
It's not really a problem for the really good or really bad novels since they're obviously getting the ratings they deserve, but the mediocre novels don't seem to be rated as such untill there's a lot of ratings and it balances out.
Personally I find it annoying when I see reviews saying "I rated it X stars because I don't think it deserves such a low/high rating". Kind of beats the point of a rating system when people don't give the rating they truly think it deserves.

 

I don't think it really matters because I never read reviews or check ratings before reading a series.

Just have to follow some set rules.
1) never read jap novels.
2)you rarely go wrong with the romance without harem genre
3)stay the fuck away from any harem genres.

 

its not really a problem if people are only rating 1 or 5 stars, the proportion of people who dont like the novel and love the novel is what matters.

The real problem is that we have some people who rate 2/3/4 stars and another group which favors binary rating, and what happens is that the ratings of this site may be slightly distorted depending the ratio of people's rating guidelines. Then again the ratings would be consistently distorted so all that matters is not to compaire this site rating's too literally to other websites without the context.

 

Ratings are subjective. They are supposed to be. Regardless of what you do, there will be bias and there will be idiots. What we have now is simple and effective.

Use ratings as a threshold to classify novels, but use your own judgement as well.
Eg: If it has a rating >3.5, it is more likely to be good. But, if it's not, don't read it. It's as simple as that.

 

Which we can then conclude that based on the majority of reviewers, most people who rate stories are bipolar. Extreme fandom/extreme hatred xD
Still you can somehow trust the rating system. If you have a peculiar taste on kinds of stories, yours will probably belong to the lower tiered ones. If you have no specific taste and just want to read something to pass time, then most high tiered are acceptable form of entertainment.

 

I believe the ratings should be-
1.Don't even think of touching it
2.Very bad
3.Ok-ok
4.Readable only if you are ok with the tags
5.Good
6.Awesome
7.Awe wait for it some

 

I understand your intention and the reasons behind your argument, however I believe that the cause of such actions if human nature itself.
The 5 star rating system is reasonable in theory and the ability to add a commented review doesn't leave much to be desired.
It's the lack thereof of unbiased reviewers and increase of smug, overconfident and prejudiced individuals that cause the current 5 star review system to be seen as incapable.
Any different review/rating system would be interesting to see but I don't think a "good/bad" rating system would be too viable, at least not more-so than the current.

 

Its fine as is.

 

Anime Seasons has Dislike, Like, Love. 3 options, and it always falls into the 77-84% rating. (They convert the choices into %)
Just like everything falls between 3.6 and 4.4 on this site.
So even if you change the rating system, you get similar results, because you can't change the people giving the ratings.

@Tony has been super busy adding stuff for us due to drama, don't go trying to make more work for him.

 

From experience what I've find out, is that people mostly rate on feelings. If you put a tag that people don't like, you can guarantee people would rate 1 without reading it.

 

Another post about rankings..... All reviews are biased. What you need to know is that not everyone will like your favourite novel. You just need to read their review.

 

By step 1 you'll miss so many good novels. Also most of CN novels here are just copy paste so, if you want to play like this then add CN here too ;p

 

Blob ftw!! right, @Westeller?

 

I still think it should be mandatory when posting a review to make your past reading list or at least the standards with which you were reviewing the novel by... Otherwise being an anonymous online novel reviewer... unless you do it for a living and have a following, I couldn't really care less and form my own opinions lol

 

Don't trust the rating. Your taste will never truly be the same as someone else's taste, so read the first 20-40 chapters to determine if you like the story or not.

 

Its fine as is but it could be better

 

I think 1 and 5 star ratings are ignored by system and avg. rating calculation exclude those.

 

Regardless of what system you implement, it will still be the same people who will rate it...
I think it's kinda pointless to change it. I don't use it either way.

 

This is pretty similar to the case with YouTube. It has a star rating system then change it to like and dislike when it becomes apparent that the users either give it one or full stars. Youtube also ditch the "average" and just show the numbers of like and dislikes.

Also, rating systems really needs large numbers of "samples" to actually create a reasonable result. Basic statistics. Sites actively ask for ratings to their products, or place the rating mechanism in top for ease of use, for this very reason. Something that NU doesn't get. It actually makes you scroll down to the very bottom to do that.

Also, bias. Again another problem when taking statistics. As you note, people would give it 5 stars if they like, or 1 star if they dislike it, if they think the average rating is too low or too high. That is they actively skew the result. This is why Amazon and the like has taken countermeasures like taking the reviewer's history into account to create a more reliable result. Again, NU is bad at this. There is the "report" feature... But really, how reliable is that? You "report" these people and nothing happens.

So to have a reliable result, you need to take a large sample and to remove bias. This means a very active participation on the NU's part, something that the people of NU is not capable of.

In short: Just ignore the rating. It is pretty much useless. And don't expect improvements. There won't be.