Not a member? Register now!
Like us on Facebook, follow us on Twitter! Check out our new tumblr page. :)
Manga News: Toriko has come to an end; but be sure to swing by and share your favourite moments from the series!
Manhwa & Manhua News: Check out the new Manhwa and Manhua shout-out (March 2016).
Event News: One Piece Awardshas concluded. Check the results Here!

Translator Level System is a Go (+ Quick FAQ)

+ posted by njt in Quality Control on Jan 25, 2010 21:09

Alright it's been a week since I introduced the plans for changing the translator's ranks and after a week trial of trying it out - it seems to be heading in the right direction. We've been able to get two S rank translators helping out, Serizawa and Molokidan (Perhaps 1 more, Nihongaeri?) and with their help we've been able to rank 16 translators! Not bad! If you think you're qualified and could help out send me a PM. Also don't forget to submit your translations here to receive your rank.

Mind you, we're still polishing off the best way to go about assigning levels and it'll take a while to get through the current active translators - but from the responses and replies from our initial rankings things are heading in a promising direction!

I'll go into the questions that have been brought up so far but first off, let me list the current changes to the new system. (It'll be listed more in detail below)

  1. I've added Difficulty to the previous 4 factors that were used in deciding your rank.

Here are the main concerns brought up by the people that posted in the news post.

Q) What about international translators?

A) We'll run the new system for a few months, once things are extremely smooth and we have the time, we'll consider taking the system to the international translators. We'll go into detail more on that later when we can.

Q) I translate in a few languages, where does that put me?

A) For now we'll still need to grade your ability to translate from Japanese into English and you'll get a rank according to how you do. In the future when we implement the above international translators ranking system we'll be able to show how you do in the other languages as well.

Q) I'm not satisfied with my rank, is there something I can do about it?

A) We aim to be fair and trustworthy with the levels we give you. If you feel that your skills are much higher than the level you've been assigned you're free to contact another, different S level member. If they agree with you and your level being too low they can then bring it to the attention of the person that originally graded you. Depending on how talks go after that you will either be graded again, or told to try again in 3 months. Seeing as we want this system to succeed we will put in the effort to make sure it is fair.

Q) I'm not happy with being a ~ level translator, can I hide what I am?

A) Sorry, no. This system is to help promote quality and reliability of translations. If that's not apparent then this system is a waste of my, and the other translator checker's time. The original creator of the manga cares about his story, I care about the accuracy of such story. So we're trying to encourage you to do everything in your power to make sure that story stays accurate as well.

Q) How will I be graded?

A) We're finding out as we're doing this, that no single grading scheme can cover every manga. So while we request you to offer a translation, we most often check another translation just to be sure. Things that we check for are as following:

1) Accuracy

While we understand various translators have their own style, we also believe that there is a difference between style and, well, just being wrong. Upon finding mistakes I've been pointing out what was missing, and what was needlessly added or how they interpreted the Japanese wrong. Sometimes translators get the wrong speaker speaking, and sometimes the conversation was continuing from one bubble to many others but not translated as such. We also consider translations that make hardly any sense to the reader inaccurate. (This often happens due to very poor flow)

2) Flow

The Japanese used in manga is natural to the Japanese person reading it, why can't the translation be in natural English if an English person is reading? With that in mind, we're trying to put a stop to the English translations that make characters sound like robots, have speaking problems (unless that is evident in the original, that is) and well, anything else that really ruins what the author is trying to go for. Seeing as this is different from translator to translator in these areas we ultimately give this little weight in deciding the final grade.

3) Grammar

I felt this, while also covered in "Accuracy", was important to have a spot on its own. If the person translating into English doesn't know English well enough it should be noted.

4) Punctuation

While it seems minor, it could actually be the cause of a lot of misunderstandings! Thus we are encouraging the right use of punctuation :).

5) Difficulty

We've always considered this but never mentioned it in the list. When deciding on the final grade, based on the level of manga. Now we realize there's an increase in difficulty to read manga when going from Shounen to Seinen, but that's not always the difference between an easy and difficult manga. So we also take note of the type of Japanese used and other factors that would make it either easier or harder to translate.

Now I'm sure the lot of you will not want to read all of that, so here's an easier to understand break down.

1) Accuracy
  1. * Adding / taking out words
  2. * Wrong interpretation (Wrong person speaking / ended a bubble when it was still going, etc)
  3. * Wrong meaning
  4. * Very poor flow (to the point where it makes no sense / interpreted wrong)
2) Flow
  1. * Characterization
  2. * Reads naturally (Flow from panel to panel, bubble to bubble)
3) Grammar
  1. * Correct usage
4) Punctuation
  1. * Correct usage
5) Difficulty
  1. * Shounen vs Seinen
  2. * Type of Japanese used
  3. * Setting

Now that you have an idea of what we look for and getting back to the original question, how you will be graded, we take the errors of each of the 4 areas and based on how many of them give you your ending grade. Seeing as how each translation is of different lengths we can't give you a solid "if you made ## number of errors you fall into this category." Instead, based on the amount of errors, the number of pages, the difficulty of the Japanese that caused such error, and the severity of them we make a judgment. That's the gist of it, but we do keep this in mind when we decide on the final outcome:

If by the end of our going over your translation we feel you destroyed the original work, then we will most likely mark you as a trainee, and encourage you to continue studying the language before you take translation seriously.

If the story is still quite off (less than 50% right), but the overall "idea" of the story is there then you'll most likely be marked a D.

If the story is right for the most part (about 70% right), but still wrong in enough areas that it would cause a lot of problems for those reading it to really understand the story, then we will most likely be marked a C and encourage you to actively seek a proofreader or translator checker based on your end result.

If the story is right for about 85% then we will most likely mark you a B, give the ok for scanlations and encourage you to work on the areas that you're weak.

And finally, if the story is right for about 95% or more we'll most likely mark you an A and encourage you to seek scanlators and to translate more series, or to even help out fellow translators :).

Once the translator has been ranked they'll have their account modified to show that as well as get one of the following badges: (Pending rank check and Trainee still to be made)

The badge will show in the right column where it says "About the author". Also, the placing of that image, and the hiding of Trainee translations, should be coded by the end of next week. I'll also be organizing a thread to link translators with things that will help them in the areas they were marked low in so as they can increase their likelihood of upping their rank.

I'm extremely sorry for the length of this post, but I was trying to be as thorough as possible. If you have any more questions, comments, and concerns let me hear em! This is for you, so the more I hear your thoughts the better system we can make!

Thanks :).

Have you shown your appreciation today? Click the thanks button or write your appreciation below!

0 members and 2 guests have thanked njt for this news
#1. by lucifell ()
Posted on Jan 25, 2010
Sounds good to me, autobots roll out!
#2. by unok-kun ()
Posted on Jan 25, 2010
... And what if trainee translators get their TLs checked before posting?
#3. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Jan 25, 2010
And they purposefully lie that it's their work? Insta-ban :(.

I guess I should comment telling them that it needs to be their work or an honest representation of their true ability (not reliant on someone checking)
#4. by unok-kun ()
Posted on Jan 26, 2010
I didn't mean that -_-"

I meant after the approval, if a trainee translator does all the work of translating a chapter with his awful skills and then he gets the translation checked before posting it, it won't show anyway... Then why would trainee translators try to improve if no one can read their tries even if they're checked?

That's not encouraging translators to improve... In that case, that's just discouraging. I don't think that's the purpose you had in mind with all that system, right?
#5. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Jan 26, 2010
Oh, did I neglect to mention an OK by an upper rank translator to push it live? But that is in the extreme cases that the translators go through the entire chapter and fixes everything - which, from my own experience, is a lot of work :(. But yeah, the trainee is to give them a chance to show their work and get it to us to monitor their progress to hopefully reach level D - where it *does* go public :)
#6. by BokoLife ()
Posted on Jan 26, 2010
I see you went with the e-badge idea...
#7. by kewl0210 ()
Posted on Jan 26, 2010
Seems fine.
I really don't think being Shonen or Seinen level makes a difference, though. A lot of seinen mangas are so severely action-based the translations are short and simple. Just the fact that most of them don't have furigana (some like Steel Ball Run do) doesn't qualify them to be harder, but I suppose that's factored in when grading.

I'll work on my stuff till I get to S...
#8. by fanatik ()
Posted on Jan 26, 2010
Yeah, I've got a small suggestion regarding the thread you mentioned. Since you're gonna go through A LOT of material and you'll have to explain errors anyway, why not make another thread, where you and other people with close to a native speaker proficiency in Japanese would collect examples of difficult to understand or interpret correctly cases with a subsequent correct translation. No basic or average stuff, but only those things that have the level of difficulty from above average to highest and, thus, tend to cause possible errors with pretty high possibility. A master-class or, uh, a model lesson of sorts, you know. Some short explanation on the particulars of the difficulty, presented in a phrase, would be nice, but not necessary - to minimize your workload just the right translation would suffice. I realize it will still give you even more work, but that will ultimately be helpful to not just one person, but to a lot of people who really strive to get better in understanding the workings of Japanese.
#9. by serizawa (ならぬことはならぬ)
Posted on Jan 26, 2010
To add to njt's post, I would like to make the following comments:

1) The profile for each level will be more or less like:
- level D and Trainee: people who are beginning to study the art of translation. Every saga has a beginning! :)

- level C: this is the average translator. People who have a decent understanding of Japanese and English, yet still fumbles in some complicated Japanese structures.

- level B: this is a good translator. People who can handle most series without making any major mistakes.

- level A: this is the crème de la crème. People who can handle ANY series without making any major mistakes.

Translation-wise, level S is (so far) just like level A.

At first, it is expected to have a Gaussian distribution, with most people in level C. First because the average translator does make some mistakes regarding some complex and/or unusual Japanese structures, and second because having two levels to grow means that there is still plenty of room for study and development. Hopefully, this will be a motivation to improve one's skills.

It must be noted that the translation checkers have absolutely NO interest in giving people a low grade/level.

The reason is quite pragmatical: the lower the level of the translations, the more work there is for the checker to point out all the passages needing improvement. The dream of any checker is to go through an awesome translation which needs no corrections nor comments other than "brilliant!".

So, if the checker gives the translator a level lower than expected, that means that the checker REALLY wants the translator to improve.

It is a BIG MISTAKE to see this level check as an exam, where getting a lower level would equal to "failure". The translator should see this level stuff as:

1) A chance for having your work evaluated and getting an assessment of your current translation skills FOR FREE, WITHOUT NEGATIVE CONSEQUENCES (like the loss of a job or having a tainted resumé) and WITH TIPS FOR IMPROVEMENT.

These are no negligible advantages - those who work professionally as a translator are surely well aware of this. In a pro world, you are doomed if you provide a less than perfect translation. Here, you can do mistakes and people will even point them to you and offer advice.

2) A motivation for HONING ONE'S TRANSLATION SKILLS. As a fan translator, whenever you translate a series, it is fairly usual to see people thanking you for providing a translation, which certainly gives satisfaction for the translator.

It is just too easy and tempting to indulge in this satisfaction and stop improving, believing that the work is good (after all, people are thanking!) and the harnessed knowledge is already enough, which is a lie. This is a dangerous trap.

This level system, especially the levels below B, is a reminder that yes, there is room for improvement. It is NOT vexing to have a low level. But it IS vexing to have a low level and DO NOT STRIVE to improve. This would be a lack of respect for:
- the author of the manga, since his/her story won't be conveyed correctly;
- for the readers, since they won't get a work of quality;
- for the translator himself/herself, since he/she is not being as good as he/she can be.

Surely the level system itself isn't perfect. Also, the system implemented in MH can have room for improvement, which depends on the active feedback of translators, readers and fans alike.

But this meritocratic system is the best MH could come up with, in order to
- Motivate translators to improve;
- Give due recognition to the brilliant translators that are here in MH.

I guess that's what I had to say :)
#10. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Jan 26, 2010
you like? :3

Read this a bit more carefully:
5) Difficulty

We've always considered this but never mentioned it in the list. When deciding on the final grade, based on the level of manga. Now we realize there's an increase in difficulty to read manga when going from Shounen to Seinen, but that's not always the difference between an easy and difficult manga. So we also take note of the type of Japanese used and other factors that would make it either easier or harder to translate.

Ideally I'd like to make a weekly column about common errors, as well as really advanced Japanese and good ways to tackle it. We'll see though ;).

Awesome additions :). Thank you very much :).
#11. by BokoLife ()
Posted on Jan 27, 2010
I prefer mine in red.
#12. by HisshouBuraiKen ()
Posted on Jan 28, 2010
Nice icons, they have kind of an old-school DDR vibe XD
#13. by Unproductive ()
Posted on Jan 29, 2010
The criteria is still a bit off. It's better than the first set where all translators were pigeon-holed into lock-steps of English and Japanese proficiency. That was really really unrealistic and gives awful feedback to translators whose proficiencies are wildly unmatched.

My first reaction here was to laugh at the 50%/70%/85%/95% level of correctness. At a glance, it seems hard for 50% or 70% to hold a story together and in my opinion, it probably would be better to keep the levels entirely subjective as in poor/fair/good/excellent. Numerical scoring is usually a poor evaluation of an activity that is more art than mechanic. I can't help but feel that you're missing the entire forest for the trees.

Beyond that, there are still a couple very important features missing from this ratings system. Those should not be mentioned until more translation checkers come on board, so I won't bring them up now.

It is interesting to see another translation checker volunteering in the effort.
#14. by bright_kingdom ()
Posted on Jan 30, 2010
nice idea :DD
#15. by Vixi ()
Posted on Jan 31, 2010
Right, when I first heard about and read about this, I thought it was a bad idea, after seeing your criticism of a translation, it's laughable.

I'm a not JP > English translator, I am a translator and learner of Gaelic languages and a native British English Speaker also reading English Lit. at degree level.

I'm also trained in translation and subtitling and this is a pointless exercise. Professionally, it's the overall meaning which is important, not every word in the sentence.

Firstly, from looking at your "corrections", they're Americocentric and quite frankly you're nitpicking.

"You're troublesome?" should be turned to "You're a pain in the ass" - give us a break. We aren't all American, we shouldn't have to use language like that, why can't we speak like we do?

And what right have you got to judge a subjective work the way you are? Just because you wouldn't translate it a certain way, doesn't mean it's incorrect! A translation doesn't have to be (and shouldn't be) a word by word account. Leaving the odd word out is perfectly fine if the translator deems it necessary. It's like a work of art.

Also, what gives you the gall to do this in the first place? To do this you must have an amazingly inflated ego. This is just an ego trip, a self satisfying exercise to make this site and scanlation something it's not.

Why should volunteers be graded as if in school, what reward is there? These are fans working for fans, and if another fan doesn't like it they can rescanlate / take over.

Why don't we have tests for editors then? Typesetters? SFX & clone ninjas? It's a natural progression isn't it?

Your corrections and suggestions are often in fact incorrect.

Oh and another thing, your sfx corrections are also, unreasonable and again subjective.

Right that's all I'm going to say ~ This really disgusts me.
#16. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Feb 1, 2010
@Vixi - Thanks for your comments ^^ From disgust we can fix up things and make the system better. For most of the nitpicking we're merely offering suggestions, but we'll try to make that more apparent though~. Like making sure to use "perhaps" or "Maybe try" and the like on each one, but yeah, I was thinking it was a bit obvious ~ guess that's what happens when you assume though. The whole "ass" out of "u" and "me" thing, eh? ;)

If I'm being "Americocentric"(never heard of that word before o.o), then the person should realize this and make their own judgment. Heck, the rating as stated above is hardly affected by the "flow". Those are by only means suggestions to perhaps make the reading experience a bit better, nothing more and nothing less.

And I quote the above:


Seeing as this is different from translator to translator in these areas we ultimately give this little weight in deciding the final grade.

Anyway, there are times where they are completely wrong in the meaning and thus get marked accordingly. There may be times where we make a mistake as well, but we try to avoid such things. I'm assuming you're talking about Bright_kingdom's translation checking? This was her first one, so I offered to help as she got the hang of it. I'll talk with her about her level of nitpicking and we'll come to some kind of solution.

This is also not an "ego-booster" of any kind. More like I've seen tons of manga get butchered and if it's going to get released on the site that I run, why not provide a bit of quality control to help at least "fix up" common errors or extremely mistranslated manga.

Unless you find it wrong to encourage others to improve and find their own style?

"Your corrections and suggestions are often in fact incorrect" I'd like to think not, but by all means, point out those if you would^^.

If it disgusts you then feel free to suggest a better method. Since this system was brought up to help make things progress for the better, negative comments *do* help. They won't fall on deaf ears ;).

Personally, seeing the lack of want/desire to advance while having no type of quality control is worse than motivating people to improve, imho.

I, as others, am willing to give our time to help others improve. If the people take our comments any other way than us trying to help them, then there needs to be a change in how we comment/suggest things. As well as an understanding from the translators that we're encouraging them to get better and strive for the top ;).
#17. by Gomenasai ()
Posted on Feb 1, 2010
Yeah, I was a little late to the game, and posted this in a section that's now closed... n e way, here's what I said b4:

Man, it took me a while to read through all of that... and I didn't read it all either :p

The one thing I have to say is: Will this really accomplish what you're trying to do? Because as I'm sure you know, aside from Naruto, Blech (not a typo) and One piece, there are almost NO other series with multiple translators. In other words, if there's a D lvl translator doing series A, the series will b stuck with D lvl translator.

Another problem you'll likely run into (I REALLY hope you don't), is the matter of taste, or sometimes even culture. I would translate 高校一年生 (koukou ichinensei) as a high school freshman. However, there are some places that don't call 9th graders freshmen. What if the checker calls the person wrong because of that? or the other way around, the person translates it as a first year (yuck!), and is told that they're too stiff or their flow is bad? Then there's the matter of slang that changes within one's own country. I'm sure you get what I'm trying to say without the long explanation.
#18. by serizawa (ならぬことはならぬ)
Posted on Feb 1, 2010

the numerical percentage is only to give a general idea of the quality demanded to achieve a level. If you look at comment #9, you'll see essentially the same idea, using other wording and *no* percentages! ;)

No doubt that numerical scoring is a poor evaluation of something that has a good deal of subjectivity, like translation. But again, that was only to give an idea - and it seemed to work, since you yourself used it to your advantage ("At a glance, it seems hard for 50% or 70% to hold a story together and in my opinion").

But rest assured that in the checking, there will be no numerical data. Those who have received the results of my checking can attest to that.

Again, the whole idea behind this translation checking is to motivate people to improve and to give due recognition to the good translators here at MH. :)

As for another checker volunteering in the effort, you are talking to one :) If you visit the translation check thread, you will see four people there to help with this initiative.

Nobody is perfect and people can make mistakes. If a translator feels that he/she didn't get a level reflecting his/her abilities, then he/she should say so, by all means. I think njt has already stated this.

Last, but not least, if you think that there are features missing from this system, please tell what they are :) As I wrote in comment #9, there is still room for improvement. So, if you feel something is amiss, then please enlighten us, so that the work can be improved. :)

#19. by serizawa (ならぬことはならぬ)
Posted on Feb 1, 2010

hey, how can you comment on something if you haven't read it all? :p Go read it first! <shoos Gomenasai> :D

Jokes aside, this translation check is *NOT* about deciding who is the best translator. It is about checking whether the translator is already fine or still needs some (or much) improvement.

In your example, if there is only a D level translator doing XYZ series, this series will indeed be stuck with him.

BUT, at least people WILL know that the translation isn't perfect.

ALSO, if a translator with a higher level (say, level A) likes the XYZ series, then perhaps he will be motivated enough to try to translate it, seeing that there is room for more accurate translations for that series.

Until now, unless the person actually browsed and checked the translation, he/she had *no* means to know about its quality. With this translation level system, the reader/fan will have a very convenient and handy way to know how trustworthy the translation is.

I think this alone is a good thing for the fans and readers, not to mention the benefits for the translators, as I stated in comment #9 :)

Matters of taste and culture have already appeared, especially taste. But it is up to the translator checker to be fair and impartial enough to understand and accept those differences - including slangs and idiomatic expressions, PROVIDED THAT the translation is correct/accurate. As you can see in njt's post, taste is something that is NOT regarded in the check.

And if a translator feels that the checker isn't being fair regarding this, then he/she is more than welcome to say so, again, PROVIDED THAT the translation is correct/accurate. That is, no use trying to convince the checker if the meaning of the word(s) being used is wrong! :)
#20. by Unproductive ()
Posted on Feb 1, 2010

Perhaps you aren't using the numerical scoring method for correctness. As far as I can tell in your TL checks, you're using the grading method that assumes that Japanese and English proficiency exist and improve in lock-step.

I don't see how I have taken advantage of the system.


There are better ways to write to make characters sound more in character. There are better ways to write to make it sound natural and native. There are better ways to write to make the passages carry the correct mood. Along the way, you might also find that most rules of grammar and spelling can and will eventually be subverted. But in your example of an "arbitrary" correction, I would have to say that NJT's suggestion is probably an improvement even for non-Americans. (arse anyone?)

Translation accuracy doesn't include writing ability beyond the bare minimum of making coherent sentences. If your only desire is a judge of translation accuracy, then there isn't much need in paying attention to any advice about writing. On top of that, there aren't many series that require good writing skills nor do I believe the translation checking team is all that skilled in writing or judging writing.

Purely pursuing accuracy won't lose meaning, but it might lose everything else. Yet on the other hand, good writing can occasionally feel like translation inaccuracies or merely creative writing. Quiet a dilemma, right? My thoughts are to follow your own path, gauge your own abilities, and set your own priorities even if it gets you stuck at a certain grade on somebody else's scale. If you are confident enough, you shouldn't care.
#21. by Vixi ()
Posted on Feb 1, 2010
Right, I was angry last night ~ I'll try and sound less angry this time ~

@ Unproductive

What is wrong with saying "It's troublesome" or "You're troublesome"? Nothing. There isn't a reason for it to be changed, especially when it's common to say it. Maybe not in America, but certainly here we use it along with shortenings like "You're a trouble". Why should we have to say things like "ass" in our work?

Also, a good translation is not only accurate and coherent but also a pleasure to read. All translation requires writing skill, sometimes in order to make things easier to read in English you may change sentence order, or restructure it completely. This isn't inaccuracy. In fact it's encouraged when working in a professional environment. My experience is in subtitling, but it makes no difference about sentence order, or even if you leave bits out as long as it makes sense. It's slightly different with scanlations since you don't have timing percentages to work with, but the idea is the same.

@ njt

If someone who has been approved on your site and translating a lot for a year and a half, has a good university degree in Japanese and yet qualifies for a low C borderline D, I'd say that your marking system is broken. Especially since the things which were pointed out / marked wrong were never marked wrong in a university environment, by professionals who teach Japanese for a living.

A C/D translator if not fit to translate manga on their own, it was not only suggested that the translator got a proofreader but also that they read more manga? To oh, understand "nuances" more? ... In my opinion that's rude, I'm sorry, but the checker has no idea how much the translator reads, or whether they read Japanese out side of manga, or how much Japanese they listen to on a daily basis, or speak, for all they know, they may read more than them. "Nuances" again are subjective.

I don't think it's right to judge fans like this, people who are practicing, or just giving their time up so that others may enjoy manga. It's not anyone's place to grade people like this. If people really want to progress, they are quite able to do so by their own motivations, without having a label pushed upon them which will no doubt be a stigma.

My friend has said herself that she is demoralised by the comments she got, mainly because there were no positive remarks, just a long list of wrongs, without explanation or justification. And these visible grade labels will be seen as a stigma by the lower grades, it really is unfair.

Striving for the top is one thing, but punishing and in effect demoralising people at the bottom is another.

As I said before, when someone has studied Japanese full time for 4 years, spent a year in Japan and spent a further year and a half translating for free at quite a rate and at good quality... I don't see this as C/D quality. I don't see HOW someone with this experience could be considered this way.

It's just wrong. I may be speaking up because of a friend, but this concern does extend broader than that. I know how I'd feel and how my friend felt and I certainly don't want to imagine anyone else feeling as bad.

I know I was severely riled and frustrated yesterday as what I saw and see as a futile exercise.

Every other sentence was "wrong" and the tone of the checker very much came across as if they believed that it was appropriate at giving a C/D rank, and that without a proofreader, they'd be a D, which means the translation would be pending and not readily available on the site...

Again, there's something wrong with this don't you think? Since she's already been through the whole process of being given the "okay" a year and a half ago. She's improved since then, why the regressive "grade".

Anyway, that's me done for now... =/
#22. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Feb 2, 2010
My thoughts are to follow your own path, gauge your own abilities, and set your own priorities even if it gets you stuck at a certain grade on somebody else's scale. If you are confident enough, you shouldn't care.
Very very nicely put :).

I know it sounds hard to believe that a person that has studied Japanese at a university, lived in Japan and has translated a good while could get ranked a C, or a D even. But it can and I'm sure will happen. It could be the translator checker was brand new at this and was picky at every single thing, it could just be that the translator made careless mistakes, or, it could be any number of things. The system isn't perfect(what system is?), but we aim to be fair. As we've stated in the above and I quote:


A) We aim to be fair and trustworthy with the levels we give you. If you feel that your skills are much higher than the level you've been assigned you're free to contact another, different S level member. If they agree with you and your level being too low they can then bring it to the attention of the person that originally graded you. Depending on how talks go after that you will either be graded again, or told to try again in 3 months. Seeing as we want this system to succeed we will put in the effort to make sure it is fair.

I'll be talking with the translator to make sure to add positive remarks in there, as I, and I know Serizawa does. (I'll talk with Molokidan as well)

But there should be no reason why the translator should just stick with a poor rank if she believes she's done better. We set the above thing up for those such reasons.

As for what you said to unproductive - some of the marks in there were trying to do such things. Moving the translation around so it becomes more of a pleasure to read. As for a proofreader, even the best of translators have them and since there were a few grammar mistakes here and there I don't see any harm in suggestion she uses one. Along with further reading Japanese more - again, this should be a no brainer and if the translator is already doing it all the better. But you'd be surprised how many translators just translate without doing any extra reading of any kind. Thus why failure to carry over nuances occasionally happens.

Anyway, the "tone" of the translator can be fixed, it was her first try after all. We all get better over time. If your translator friend has any further problems then have her contact us. There's no reason to call the system broken if they don't even try to utilize it properly.

I mean each translation check takes a good few hours (depending on the length, and depending if the initial check was enough to decide on a proper ranking or not) so we're not doing this for "fun" or for some kind of power trip, we're doing this to genuinely try to promote them to get better as well as provide some kind of awareness at what to expect. If there is any sort of dissatisfaction with it, methods should be taken to remedy it, not just griping about it.

also in regards to D:


Level D: A Level D translator has the basics down, but has too many mistakes to be approved for scanlations. The translations can go out to the public and can be used to get an idea of what the manga is about, but should not be used in scanlations.
#23. by Vixi ()
Posted on Feb 2, 2010
@ njt -

Griping about it would be bitching about it and not doing anything about it, but I decided to actually sit and spend time airing my grievances.

Anyway, that's all I really have to say on the subject.
#24. by njt (Last Boss ♪~( ̄。 ̄))
Posted on Feb 2, 2010

Right, but it's stopping there. We've pointed out how things can be remedied (that we are aware of) I mean, without other suggestions there really can't be any improvements or they'd have already been made.

So if the translator is not satisfied try getting her to state such things^^ and ask for another person to check her translation~.
#25. by serizawa (ならぬことはならぬ)
Posted on Feb 3, 2010
That "grading method" you see is just a general notice. As you can surely notice, I am just copying the way Molokidan uses to show the result of his evaluation.

There are more things going on under the surface. :) If there are points to be improved, I point them out to the translator and if there are points that I find particularly good, I let the translator know as well. But everything is done privately, so it's no wonder that a third person doesn't know about that.

Also I don't believe that Japanese and English skills are tied up together. Edgar Allan Poe is no Mori Ogai and Matsuo Basho is no T.S. Elliot.

But curiously, in the checks I did so far, those skills seems to be tied up, and that's why I'm using those notices. People who have a good English tend to have better skills understanding the Japanese language.

I don't know why this happens. Perhaps it's because translators who care about writing in good English do take it seriously to have a good understanding of any languages they work with, including Japanese.

Either way, if there is a large discrepancy between the Japanese skills and the English skills of the translator, I will modify the notice to reflect that.

As for "taking advantage of the system", perhaps I can answer this with a question: If you reckon (correctly, by the way) that numerical scoring is a poor evaluation, then what made you think that 50%-70% of correctness is hard to hold a story together? :) See what I mean? :)

No doubt that any attempt of quantification will meet with failure. However, to give the translation candidate an idea of the meaning of the levels, the use of percentages is valid, though not quite accurate.
#26. by Name-Undecided ()
Posted on Feb 3, 2010
All of a sudden I really really want to read the translation Vixi is talking about. It seems a no-brainer to me that someone can be 100% fluent in Japanese and still be bad at translating said Japanese into proper English. If you doubt that, go to Japan and ask the average Japanese man on the street to translate chapter XX of manga YY for you. That's one reason why I'm usually skeptical when people go "Oh, I got a native Japanese to translate this, it must be perfect!"

But I digress. What I was going to say is, translation is a highly subjective thing, so translation checking is highly subjective as well. As long as the MH team is open to reevaluation and there's a system for addressing complaints, I think it will all work out eventually.
#27. by Unproductive ()
Posted on Feb 3, 2010

If you called a rhetorical argument "use to your advantage," I would suggest that you review the implications that "use to your advantage" brings with along with its use. As a matter of rule, numerical evaluations will fail eventually - usually sooner than expected. It's still possible to use them as heuristic to test whether some hard rules make sense.

In this case, the rule as is described fails. Either the numerical rule is grossly incorrect, it is being described incorrectly, or it's not even being followed by the checkers. This is before considering the systemic problem of using numerical scores to judge translation ability from individual samples.


Might as well not beat around the bush. I thought it would be quite obvious. The highest priority in any grading system would be establish some level of credibility in the graders. At this point, all the rest of us know is that the graders have been selected by njt for some reason or another, level S be damned.

At this point, one translation checker is as good as another. Nobody knows much about the TL checkers. If one translation checker is bad, all of them could be bad. They may not be bad in the same ways. That depends on how the translation checkers were selected. Maybe they were selected for being the best overall TL checker; in which case, their flaws might all be different. Maybe there was a huge blind-spot in the selection criteria; in which case their flaws might all be the same. Or it could be some combination. Who knows.
#28. by Vixi ()
Posted on Feb 3, 2010
@ Unproductive

I'm so glad someone else thought of this too ._.;
My sentiments exactly.
#29. by fanatik ()
Posted on Feb 25, 2010
Well, everybody probably just forgot about this innovation now, but as I said earlier I was gonna wait and see how it'd go. And now... it's possible to start making some conclusions. Some unpromising conclusions.

Vixi above was unhappy that the level given was lower than expected - can't say if in this particular case it was deserved or not - but apparently the opposite is true as well... You're giving out levels higher than deserved. A guy can't see that 方 reads not as 'kata', but as 'hou' (I told it was a bit harder, when there's NO furigana readings, didn't I?) and is an element in a modified construction 'hou ga ii' (as basic as it gets, for goodness sake...), and, naturally, makes up some phrase totally unrelated to what was really said, confusing everybody, and that among lots of other things that distort the very plot of the manga over and over again, and you give the guy a B-Level?... YOU give them B, YOU think this is what B-LEVEL should be?... You're kidding me? You're friggin KIDDING me? I'm downright scared to think what D-level is, then.

What was the point of all of this if this is what we get as a result? Mistakes, clear and plain mistakes (not in some stylistics, but in the core meaning of sentences!) approved by a translator's level?! I'm sorry to say, but if this is the result, then this undertaking, this whole endeavor is a failure, complete and total.

About the Author:

Alias: njt
Rank: Last Boss ♪~( ̄。 ̄)
Registered: Feb 28, 2005
Message: PM |

Author's Contributions

Forum Posts: 4262

Latest Translations

Date Manga Ch Lang Translator
May 2, 2023 Baki Son of Ogre 10 tr McMaster68
May 2, 2023 Baki Son of Ogre 9 tr McMaster68
May 2, 2023 Baki Son of Ogre 8 tr McMaster68
May 2, 2023 Baki Son of Ogre 7 tr McMaster68
May 2, 2023 Baki Son of Ogre 6 tr McMaster68
Apr 21, 2023 Baki Son of Ogre 5 tr McMaster68
Apr 21, 2023 Baki Son of Ogre 4 tr McMaster68
Apr 21, 2023 Baki Son of Ogre 3 tr McMaster68
Apr 21, 2023 Baki Son of Ogre 2 tr McMaster68
Apr 21, 2023 Baki Son of Ogre 1 tr McMaster68

Latest Releases

Date Manga Ch Lang Uploader
Apr 14 MH Yearbook 2019 Mangahe...
Feb 15 MH Yearbook 2018 Mangahe...
Mar 1 MH Yearbook 2013 Mangahe...
Jan 19 MH Yearbook 2012 Mangahe...
Nov 14 Houkago 1 Osso
Nov 14 Oragamura 1 Osso
Nov 14 Kenka 1 Osso
Nov 14 101Kg 1 Osso
Nov 14 Murder 1 Osso
Nov 14 Doubles 1 Osso