Feature Request Replace the 5-stars rating system!

Discussion in 'Suggestions & Bug Reports' started by Dino Translation, Aug 20, 2017.

  1. Dino Translation

    Dino Translation Active Member

    Joined:
    Jul 25, 2017
    Messages:
    74
    Likes Received:
    115
    Reading List:
    Link
    Warning, this post is very long.
    Note, this post is written from the perspective as a translator not a reader.


    Replace 5-stars ratings system, why?
    This part is mostly ranting, if you don’t care about it, skip to the Proposal section!

    As you can see, the 5-stars rating system right now is utterly broken. Even if NU policy does say that one should rate the story and not the translation, but it is obvious that most the ratings are about the translation.

    People are saying that separating the translation rating and story rating will hit the motivation of the translators. Honestly, I agree. But looking at things right now, is there really any different? People will give low ratings for bad translations anyway, and that eliminates the translators just as fast as separated ratings. It makes absolutely no different.

    Translators who quit will quit.
    And once they do, only a dead series remains.
    Not only it doesn’t keep the translators from getting flamed, the series they were translating fell victim to it as well.

    Even if you don’t give the option to evaluate the translators, the readers will just take it out on the series anyway, and the series ended up dead because of it. The low ratings will prevent other translators from picking it up, giving it no chances for recovery. Even if some translator did have some interest, they will probably just read it themselves without translating it. After all, even if they did, no one would read it anymore because of the low ratings. There’s no second chances.

    If you think translators don’t care about ratings or such. You’re wrong. Translators are often readers as well, and there are some who decide on their new series using the Pick Up Request forum, and a low rating series doesn’t have much chances of being requested. If you think that some translators don’t care about ratings, you’re not wrong, but that’s simply wishful thinking on your part. You’re hoping that things would go well without doing anything about it even though you can.

    If you think that there are still plenty of novels out there, losing a few wouldn’t hurt. You’re not wrong but not right either. Why must we deliberately limit ourselves if we don’t have to? Isn’t it better if we have more freedom? If things could be better, then why not try to improve on it instead of remaining stagnant? Don’t be a little frog in a well! (i.e. narrow-minded)

    In the first place, I think that the very premise of separating the story and translation itself is ridiculous. After all, a novel is basically a story conveyed to the readers through writing.
    The author is responsible for both the story and writing.
    Whereas the translator is responsible for only the writing, in place of the author.
    In other words, they’re responsible for half the novel, if the novel itself gets rated, they cannot escape from being involved.
    Some people may ignore the bad translations and evaluate only the story. But again, that’s simply wishful thinking. If readers want to give 1-star for a bad translation, what can you do to stop them? Nothing.

    Finally, why is it that we have to take care not to hurt the translators’ feelings? Are we afraid that they will quit and we’ll have no novels to read? Heck, if so, why are we doing nothing? Instead of that let-them-be attitude of ours, why don’t we actively find a way to encourage the translators to do their best instead? Why care about hurting their feelings when we should be thinking about how to make them more motivated? If we like novels, shouldn’t we support the translators instead of simply let them be?
    Of course, there are already people who thank the translators. But newbie translators with limited skill won’t get thanks so easily, it’s easy to lose motivation. Experienced translators won’t be affected, but there’s a reason why so many newbie translators dropped out before becoming full-fledged, and I blame the 5-stars rating system for that.

    Well, that’s a long rant. In any case, back to the topics. I believe it would be better to replace the 5-stars rating system with something more translators friendly.

    Below is my proposal. Considering how difficult things are, as well as the many problems that will arise from it, it probably won’t get accepted.
    But rather than expecting it to be accepted, this is more to raise awareness of the problems of the current system.



    Proposal: the +1, +2 method
    Objective of this proposal:
    • Remove negative responses.
    • Translator centered evaluation system.
    • Allow re-evaluation of the series when a new version (by a different translator) is released.
    • Allow the evaluation of translators, again, not in a negative sense.
    How to do that?
    1. Replace the 5 stars ranking system with +1, +2 buttons, similar to Facebook, Google Plus. This is the same reason why the forum don’t have a dislike button.
    2. Add a dislike button to the series as a whole, disliking a series will hide it from the user but it doesn’t affect anything else. You can call this “Ignore List” in contrast with “Reading List”.
    3. Evaluation is done on a {Series} + {Translator} pair basis. That is to say, when evaluating a series, you’ll see a list of translators working on it and give them your evaluation individually.
    4. Evaluation of the {Series} is the average of all chapters, with every chapter has their own evaluation which is the MAX value from all the versions released by different translators.
    5. Evaluation of {Translator} is the sum of all the Likes (Points) they received from the {Series} they translate, multiplied and adjusted by the number of chapters and average length of each chapter.
    6. Quality Evaluation is calculated by Points / Readers. However, this value should be hidden from readers. It could be used to show a recommendation mark or ranking without giving them the gritty details. Let readers decide for themselves where is the cut off for good and bad series. See Coverage below.

    Example
    Series Name: {Awesome Novel}
    Average Points: 86.6
    {Translator A}
    +1 +2 (80 Points) (10 chapters)
    Released chapter 1 -> chapter 10
    (100 readers) (80%)
    {Translator B} +1 +2 (70 Points) (05 chapters)
    Released chapter 11 -> chapter 15
    (60 readers) (116.7%)
    {Translator C} +1 +2 (90 Points) (25 chapters)
    Released chapter 12 -> chapter 36
    (85 readers) (105.9%)
    • +1 Button
      
Give 1 Points to {Translator ?} evaluation, Good.
    • +2 Button
      
Give 2 Points to {Translator ?} evaluation, Very Good.
    • Points

      Sum of all +1, +2 receives.
    • NUM chapters
      
Number of chapters {Translator ?} has released for this {Awesome Novel}.
    • Readers (hidden)
      Number of users who have read at least 25% the chapters of {Awesome Novel} released by {Translator ?} or +1/+2 the translator. This should be hidden from users.
    • Coverage (hidden)
      
Points / Readers. The possible range of value for Coverage goes from 0% to 200%. This should be hidden from users.

    Sorting, Ranking
    There are two kind of data used for sorting in this proposal:
    • Points
    • Coverage

    Series Points
    The Average Points of all chapters. Calculation method is as followed:
    Group chapters, different versions of chapters are group together, including parts (c10 part 1, c10 part 2) and multiple chapters (c1-c3). Releases of multiple chapters (c1-c3) are put in multiple groups.
    The evaluation of each chapter is determined by the MAX value of Points of the translators who have made a release for that chapter.
    Then the average is calculated for all chapters.

    In the example above:
    SUM of chapter 1 -> 10 = 80*10 = 800
    SUM of chapter 11 = 70
    SUM of chapter 12 -> 15 = MAX(70, 90)*4 = 360
    SUM of chapter 16 -> 36 = 90*21 = 1890
    AVERAGE = (800 + 70 + 360 + 1890)/36 = 86.6
    This series gets 86.6 Points in overall

    Coverage
    Same as Points, each chapter’s Coverage is determined by the MAX value of all the translators who made a release for that chapter.
    Then the average is calculated for all chapters.

    In the example above:
    SUM of chapter 1 -> 10 = 80%*10 = 800%
    SUM of chapter 11 -> 15 = 116.7%*5 = 583.5%
    SUM of chapter 16 -> 36 = 105.9%*21 = 2223.9%
    AVERAGE = (800% + 583.5% + 2223.9%)/36 = 100.2%
    This series gets 100.2% Coverage in overall

    Points and Coverage should then be used with whatever data which is already existed to rank the series (Reading List, Popularity, Activity).
    Coverage are meant to replace the order by rating option.
    When a chapter have multiple version, the one with higher Coverage value has priority.

    Translators
    Translator Points

    Similar to Series Points. A Translator receives Points from all the series they translated for, adjusted by the number of releases and the average length of each releases. This is a little tricky since each series has a different word count per chapter.

    In any case, Translator Points will always increase as long as you keep releasing new chapters (Assuming you have at least 1 point to begin with that is). This is to encourage translators to keep translating, even if they didn’t do a good job of it. Though the gesture is very small, Points could be interpret as the readers appreciating their efforts.

    Translator Readers (hidden)
    The number of readers who have read the translator’s work.
    If a single user read from two different series of the same translator, they’re counted as 2 readers.
    This value should be hidden from the users.

    Translator Coverage (hidden)
    It’s calculated by Points/Translator Readers. Points is the unadjusted value of the total of Points from all series the translator is involved in.
    This value should be hidden from the users.

    Translator Rankings
    Nothing come to mind, but we should make use of Translator Points and Translator Coverage. As for whether there should be a translator ranking or not is up to debate.

    Remind people to give their thumbs up
    While a user has read 25% to 75% of the chapters that {Translator ?} has released for {Awesome Novel}. Put up a reminder for them to do so.

    Migrate from 5-stars system
    Turn 4-stars to +1, applied this to all translators of the novel
    Turn 5-stars to +2, applied this to all translators of the novel
    1-star to 3-stars are regarded as readers.
    Anything else is discarded.
    Reviews don’t show any rating.

    I don’t believe this proposal is perfect, but simply think of it as an interesting concept. In the first place, this is written with translators in mind, so readers may not like it.
    Feel free to disregard it as you will.


    ———

    Last but not least, @Tony I’m not saying you should do this or that, but in the future when you’re making big changes to NU (Say NU 2.0), I would really appreciate it if you spend some thought on how to improve the current rating system. Because having a single rating for a novel that multiple translators work on doesn’t make senses, that method wasn't designed with novel translations in mind. Not giving an option to give feed back to translators will just make readers find a different way to throw their frustration (in this case, toward the novel). That’s just looking away from the problems hoping it would disappear.

    Well, that’s the gist of it, it has been a long boring trip, but thanks for reading.
     
    Valven, Illisaide, sharlnina and 9 others like this.
  2. mrttao

    mrttao Well-Known Member

    Joined:
    Aug 23, 2016
    Messages:
    2,653
    Likes Received:
    1,532
    Reading List:
    Link
    On the one hand, I feel this is not ideal. on the other hand I feel this is an improvement over the actively harmful 5 star system.
     
  3. asriu

    asriu Well-Known Member

    Joined:
    Jan 9, 2016
    Messages:
    14,921
    Likes Received:
    14,774
    Reading List:
    Link
    that's good~
    still if NU become translator oriented I prefer it make entire new section or better new site just for translator~
    my view is despite translator have great part for bring a title~ I must divide translator quality and story quality as two separate thing~
    despite one can effect each other
    example ren and Coiling Dragon compare with teh ping and Hikaru ga Chikyuu ni Itakoro……
    I agree with translator rating that can used to show
    • single / team translator quality grammar~ story flow well something like that
    • pacing~ release speed ~ maa not really mind about this cuz I'm used waiting for long time
    but to affect a novel rank? totally disagree
    if someone give bad rate cuz it not like translator IMO that ridiculous~
    that similar with reader ask faster release disregarding translator problem~
     
    Slayerwolfx2 and mrttao like this.
  4. kenar

    kenar ヽ(`・ω・´)ゝ

    Joined:
    Nov 2, 2016
    Messages:
    2,876
    Likes Received:
    4,543
    Reading List:
    Link
    How much is the burden to the server? And how about people rating ntr novel as 1 star despite not reading it at all?
     
  5. kenar

    kenar ヽ(`・ω・´)ゝ

    Joined:
    Nov 2, 2016
    Messages:
    2,876
    Likes Received:
    4,543
    Reading List:
    Link
    yeah, some people starting to complain about translation speed. It is a bit troubling since most of the translator isn't as fast as a certain translator
     
  6. Dino Translation

    Dino Translation Active Member

    Joined:
    Jul 25, 2017
    Messages:
    74
    Likes Received:
    115
    Reading List:
    Link
    Ideals are what we always aim for but could never reach...
     
    AkeShiro, pinkrainmi and mrttao like this.
  7. Yukkuri Oniisan

    Yukkuri Oniisan 『Procrastinator Archwizard Translator and Writer』

    Joined:
    Oct 24, 2015
    Messages:
    5,399
    Likes Received:
    9,168
    Reading List:
    Link
    Well... As a translator myself, don't really care about rating and stuff...
    I only translate what I think is interesting and I want other people to read it too......
    If it had low rating then it only mean that there are not many people enjoy the series like me........

    I mean, I even find this title interesting:
    A Little Girl (Loli) and A Butler in Another World/ Youjo to Shitsuji Isekai de/ 幼女と執事が異世界で
    It got BL, it got Yuri, it got genderbender, it got an adult male reincarnated into little girl.
    I bet it will be instant 1 in NU Rating.
    I think it final score will hover between 2 to 3.
    But if I got my hands free, why not? I translate since I want to share the story to other people.

    Also rating.
    At least for me, I read something, not because of the rating or the review, but because I like the premise / story / theme / etc, which might be different from other people. People like Naruto/OnePiece/Bleach? Nope, I like Gintama and Negima better. I like Bobobobo-bobo-bobobo. I read Cromartie High School. I watch RyoKoBu and Tenshi no 3P. As you can see, my taste is not mainstream. So mainstream rating has no any relevance nor meaning for me.
    I mean EDF 4.1. got a 64 score in metacritic, many criticizing the plot, gameplay or the clunky graphics. But it was a solid 90 for me, I mean who cares about plot or high end realistic graphic? I find shooting bugs and giant robot fun! I even can punch Kaijuu with Giant Robot Pacific Rim-style!

    So yeah, if translator only pick a story based on the rating then I find it sad. Fantranslator should translate what they like, not dictated by their surrounding...

    But I agree about rating system.
    It should be simply:
    1. You should read this.
    2. You should not read this.

    The rating to end em all.
     
    AkeShiro, pinkrainmi, Lovely and 8 others like this.
  8. mrttao

    mrttao Well-Known Member

    Joined:
    Aug 23, 2016
    Messages:
    2,653
    Likes Received:
    1,532
    Reading List:
    Link
    what about a "meh" option? I read it but i am kinda neutral
     
    Slayerwolfx2 likes this.
  9. Yukkuri Oniisan

    Yukkuri Oniisan 『Procrastinator Archwizard Translator and Writer』

    Joined:
    Oct 24, 2015
    Messages:
    5,399
    Likes Received:
    9,168
    Reading List:
    Link
    Then.
    1. Interesting.
    2. Not interesting.
    A meh is not interesting.
     
    Lovely likes this.
  10. Little Evil

    Little Evil >.> I see you...

    Joined:
    Jul 27, 2017
    Messages:
    233
    Likes Received:
    4,104
    Reading List:
    Link
    5 star system is ok. You just need to look if it has NTR, if yes, you should not take it into consideration as many rate one even without reading after they see that tag. Generally, novels above 4.2 are readable and above 4.5 are really good. Also, as someone said here, translation is part of what makes reading novel enjoyable. If it is sub par, it should affect novels ranking. That way, new readers know what it is that awaits them.
     
  11. Yukkuri Oniisan

    Yukkuri Oniisan 『Procrastinator Archwizard Translator and Writer』

    Joined:
    Oct 24, 2015
    Messages:
    5,399
    Likes Received:
    9,168
    Reading List:
    Link
    Oh... Also about translation quality:
    1. Readable
    2. Unreadable
    Simple! Since some people even bothered by the presence of miscapitalization and over or under-punctuation and deemed it as unreadable, while some people can't even differentiate good grammar English from bad grammar English since they barely able to speak English at all. Like some even don't know the reason why English narrative is past tense (historical past).
    Please remember not everyone uses English as the first language.
     
    Pythagoras Theorem likes this.
  12. Dino Translation

    Dino Translation Active Member

    Joined:
    Jul 25, 2017
    Messages:
    74
    Likes Received:
    115
    Reading List:
    Link
    That's because @Yukkuri Oniisan is already a veteran translator. I'm a newbie so I'm still bound by my common sense as a reader. If I continue to translate, perhaps I may reach the same enlightenment. But that's only if I can last until then.
    Beside, even as as a reader, I find it annoying that people would give a low rating for non-mainstream stuffs.

    On a side note, I like Gintama and Negima too, those two are good. Of course, I also like Naruto, OnePiece and Bleach, but I simply don't like the hype people make them out to be. I haven't read or watch the others one you mention so I can say anything about them though.
    Generally, I don't rank the series by which one is better but simply interesting or not. And I read them according to my mood. Because I believe how much you enjoy a series depend on what you want to read/watch at the time. Even a series you thought as crappy may become great when you read/watch it later.
     
  13. mrttao

    mrttao Well-Known Member

    Joined:
    Aug 23, 2016
    Messages:
    2,653
    Likes Received:
    1,532
    Reading List:
    Link
    Meh is neutral. There is something between negative and positive.
    "This is bad" and "this is good" cannot encompass all things. some things are niether bad nor good
     
  14. Westeller

    Westeller 『I'm Home!』 Staff Member

    Joined:
    Apr 3, 2016
    Messages:
    7,250
    Likes Received:
    21,430
    Reading List:
    Link
    The 5-star rating system is superfluous, heavily abused, and just plain useless. It tells you absolutely nothing about the novel - not even on average how many people liked it, which is what it's supposed to do. Almost anything would be an improvement.

    But, uh, man. Ratings should not all be about the translator. Unless the translation is literally impossible to read, you don't rate the translation - you rate the novel. I do get that a lot of people see poor engrish translations and immediately rate the translation instead of the novel, but just as many see things like harem or yaoi tags, silly titles, a first chapter they don't particularly like, and immediately rate a novel 1-star. That has absolutely nothing to do with the translator, and it's just as big, if not bigger, of a problem. We need a more constructive and informative rating system than a generic 5-star blegh system, and it shouldn't be all about the translator. The novel is just as, if not way more, important.
     
  15. J-Mitch

    J-Mitch ⚖ Tipping the Scales of the World

    Joined:
    Mar 11, 2016
    Messages:
    1,913
    Likes Received:
    3,737
    Reading List:
    Link
    Nah, that was not long at all.

    Not sure I agree with most of the idealism behind your application for reform, but I do understand the need for it.
    I say that because it is a translator's obligation to present the story appropriately. A lot of translators out there have crap grammar. Their translations are no better than mtl. Some are just that: Mtl. and for those that decide to take the novel seriously later on, the ratings suffer due to their predecessors.

    So, yes, I do love the idea for a reset for when a new translator takes over a series (who starts from the beginning). I also like the idea to rate the translation quality. But a complete reform seems daunting.

    There are some good ideas, but this is complicated. I can't imagine much of anyone actually reading the instructions should this be implemented if the manual is just as long as your thread.

    I didn't even read all of it. I started skimming and complete skipping before you even mentioned 2 ranking systems.

    Besides, there are groups that use multiple translators for a series. So, if you do this, it would be group based, not translator specific. And have you seen what some novels look like before they are looked at by an editor? In the translation community, translators are rated by multiple factors. From your system, this will likely not give an appropriate representation of a translator. More, the group as a whole (Translator + Editor).

    For me, right now, I think the best parts to take from this are two things:
    1. Separate ratings based on Group.
    2. Rate translation quality (divided into accuracy + Grammar).
     
    Last edited: Aug 21, 2017
  16. saekicchi

    saekicchi (≧◡≦) ♡

    Joined:
    Oct 22, 2015
    Messages:
    404
    Likes Received:
    544
    Reading List:
    Link
    I'd rather have something like this:
    Readability/Translation Quality (maybe add an option to indicate translation group): ☆☆☆☆
    Story Line/Plot: ☆☆☆☆☆
    World Building: ☆☆☆☆
    Characters: ☆☆☆
    +others
    Overall: ☆☆☆☆☆
    And a section where they can elaborate more on their rating.
     
    Summer and pinkrainmi like this.
  17. Pythagoras Theorem

    Pythagoras Theorem Horse (?)

    Joined:
    Nov 27, 2016
    Messages:
    730
    Likes Received:
    1,759
    Reading List:
    Link
    I'm not sure what you mean here by "veteran"? Don't people translate a novel because they love it and want to share it with everyone? :blobunsure:

    Are you saying that the reason you translate something so that a lot of people will read it and praise how amazing you are? :hmm:

    For me, I think you should just translate a novel that you love because if you love that novel, then there are bound to be others who would love it too. Yeah, you might get less views and yeah, you might get hate. But so what? I won't stop reading my favourite novel just because some jerk told me the novel was shit and called me a nerd with bad taste. And you shouldn't too.

    As long as you're having fun, what does something like rankings and reviews matter?
     
    pinkrainmi, AliceShiki, kenar and 3 others like this.
  18. Dino Translation

    Dino Translation Active Member

    Joined:
    Jul 25, 2017
    Messages:
    74
    Likes Received:
    115
    Reading List:
    Link
    Yeah, I agree. At least my reason is. But I would still feel bad when people badmouth my favorite novels. Emotion and Reason are different things. Even if I know myself I shouldn't care about it. It still bugs me.
    That's why I'm inexperienced. I haven't transcended to a state where I can just disregard all the people who would speak ill of me. That's why I said @Yukkuri Oniisan is a veteran, in that he has that unshakable belief. Hopefully, I can reach that state soon.

    EDIT: Let's give an example. This didn't happen but I feel that something like this could happen anytime soon.
    1. "Heck, what's this guy saying, how dare he insult my novel?"
    2. Then I spend the next 15 minutes cursing him.
    3. "Phew, I'm tired. Novels? I have enough for today. That reminds me, I haven't play that game for sometime."
    4. Then I spend the whole night playing games. And the releases are delayed...
    5. "The releases are too slow, what the heck are you saying, I've been doing my best here!"
    6. Go to step 2.
     
    Last edited: Aug 21, 2017
  19. AkeShiro

    AkeShiro Blanket Master

    Joined:
    Oct 20, 2015
    Messages:
    1,172
    Likes Received:
    808
    Reading List:
    Link
    I think points would fare better?
     
  20. gggo

    gggo Well-Known Member

    Joined:
    Aug 9, 2017
    Messages:
    513
    Likes Received:
    331
    Reading List:
    Link
    I think that what the OP's proposed rating system is doing is to only accumulate positive response (likes). I think this is inspired by social media because this is commonly used to calculates a post's appearance in the newsfeed.

    I think this could probably work for the recommendation algorithm of the novel, but I think it's a terrible replacement for the current ratings system because it not only hides the no. of readers who read a novel, it also attempts to deliberately mislead users into thinking that a novel is good even though only a few people have read it.
     
    mrttao likes this.