Introducing Natively - Now with Movies & TV Shows, Korean, Spanish, German and More!

I think the JLPT level approximants are vitally important for when you are submitting your book. You use your current level, and decide if the book is on the lowed, high ed, etc. Otherwise, you’d just see a massive amount of numbers and no real context. Like, is it needed once you’ve selected a book already on the platform? No, not particularly, BUT anyone who is trying to add their book for the first time, and have not read any of the other books listed on the site will have no idea at all what to do.

5 Likes

Personally, people are weirdly obsessed with what the JLPT tests people on. like why are you so obsessed with “oh every word on the N3 doesn’t appear in this book, and it has some words form N2 so you can’t put it with those levels”. Like, that argument is so weird. You’re not memorizing a dictionary, you are giving yourself an estaminet level of Japanese based on how much grammar/vocab you know. Its about how much you will struggle with the material, not how much vocab/grammar you will know thats in the book. As someone who has passed N3, I place the books I’ve read mostly in the N3 bracket, as I tend to read book at around my level. Thats all there is to it. Like yeah, I didn’t know 600 words from Yakusoku no neverland, THAT DOESN’T MATTER. It certainly wasn’t that hard of a book, so obviously its not a N2 book. Was it harder than 少女終末旅行(21)? yeah, so it got level 25. Like, thats all there is to it. The whole rating thing is about estimating stuff so that people can find books to read. Until we get multiple people reading the same books (which like 5 of the books have this on the site so far) we won’t have super accuate rating. Like, it literally says so on the site, so I don’t see how adding ‘N#’ increased the ’ false sense of precision and accuracy ’ like, the whole thing isn’t anywhere near accuract yet except for like the top 5 books. You can compain about the preception of accuracy after the amount of ratings increases so there is even something to be ‘accurate’ to judge on. Currently, almost all the awesome books on this side has 1 graded estimate. Thats only 1 judgement from a random person on the internet. How accurate do you think it can be?

2 Likes

I completely disagree. Relative levels is the only thing that matters. I submitted around 20 manga last week. You know how I choose the initial level? By comparing them to books already on the site.

And this is why the initial level should be optional. If you don’t know the level, don’t guess. The site’s levels won’t line up with the JLPT levels anyway, since as we’ve been saying, the JLPT ranges on the site are arbitrary.


EDIT:
As I’ve mentioned before, I think you shouldn’t be asked for an initial level directly. Rather, you should supply a few books of similar difficulty (and maybe easier and harder ones if that makes sense). This is more in line with how the site works anyway, and it removes the question of how someone should pick the initial difficulty (manually comparing vs JLPT level, etc.). And again, this should be optional in case you haven’t read other books on the site, or in case you haven’t even read the book you’re adding yet.

7 Likes

You’re right @NicoleIsEnough that the books are not tied at all to the specific things the JLPT tests and in that sense it certainly isn’t accurate and a very fair criticism. I guess my initial thought around the levels was more around @MonnieBiloney’s thought - it’s simply using JLPT as a general proxy for Japanese proficiency and the filters are purely a “here’s what we think other people around the N3/N2/N1 level are reading and what you might find interesting”.

If people don’t find those levels N3/N2/N1 filters useful, of course I’d remove them but my impression is that they do! N3 filters to mostly manga and some of the more accessible novels / light novels. N2 filters to novels, harder manga and light novels. N1 to hard novels. Now, I personally can’t speak to the actual difficulty of these things, but generally those results seem to make sense to me.

Of course, JLPT levels do not not directly relate to proficiency in reading fiction, but I do think it’s still the common parlance when discussing general Japanese proficiency and very frequently used in regards to recommending reading material even to people not focused on the specific JLPT tests.

You’re criticism though about a “false sense of precision and accuracy in my opinion” I do agree with and would be the primary driver if I were to remove it. And perhaps there’s just a better way to go about it - maybe, like you said, I have more focus on general descriptors (beginner, intermediate… etc) and have the ‘approximate mapping to JLPT’ simply in the ‘our grading system’ page where I can more fully add disclaimers.

I don’t think there’s much disagreement here - it is better to set initial levels based on levels on the site, not JLPT levels. But, while it’s better to set a good initial level, I don’t think it’s that big of issue. The temporary rating system can change the level quite significantly (5-6 levels easily) and I can see if a book looks like it didn’t reach a good initial non-temporary level if a book only got rated as ‘more difficult’ or ‘easier’ for all 5 gradings.

So while there have been people setting based on JLPT levels when they request books, it hasn’t really caused large issues I don’t think. I do agree with @MonnieBiloney that it is beneficial for this use case currently, but definitely agree with @seanblue that his system would be more ideal and intuitive. I do need to have an ‘unrated’ mechanism, which really wouldn’t be hard to create either.

5 Likes

I should add that I think using comparisons when requesting a new book would solve another problem. Remember that when I added a bunch of books and commented that many settled around level 20, I was mostly concerned that over half of their comparisons were to one another (basically to other temporary level books), which seemed odd to me. They basically each reinforced whatever bias I had when giving them their initial arbitrary rating. If you had to pick similar difficulty books when requesting them, it wouldn’t be possible for this to happen (at least not when requesting several at once).


Also… :laughing:

11 Likes

I have my library sorted by level, showing all volumes. In general, the volumes are shown from high to low, but for some reason ヒナまつり is slightly out of order.


EDIT: A second one out of order (this one split by another series).

1 Like

I’m also getting a 500 error when I try to edit my review for こみっくがーるず 1, even though I literally just saved a few for another series a minute earlier. The text just says:

Cute series with great characters and great comedy. Highly recommended to anyone who likes slice of life comedy, and especially to those who enjoy the “cute girls doing cute things” sub-genre.

Nothing sticks out as something that might cause server issues.

1 Like

It’s ordered by date added if the levels are the same and it looks like you did add them in that order (2 before 1)… but I do understand why it’s a little confusing as I only show the date added whereas I order by the microsecond added. I probably should only order by date added, not microsecond… and order by series & series order after that. Although it does still get a little strange once it’s around midnight, but overall would be more intuitive, thanks for pointing out.

I could perhaps order by something other than date added if the levels are the same… by series and then by date added probably makes more sense upon thinking on it more.

I’ve looked into this and it looks like you just were unlucky - every time I release a new update, there’s a one or two second window where a request can fail. At some point I will fix this, it’s really not hard at all :slight_smile:

2 Likes

I blame you for that one. :stuck_out_tongue:
(You added volume 2 to my account instead of volume 1 when I requested it for some reason).

I would definitely find series order more useful than date added. In fact, I find date added to be one of the least useful things to sort by (since date finished is better than date added as well).

Thanks for letting me know. Looks like the change was saved on one of my attempts anyway.

2 Likes

@sweetbeems
I sincerely appreciate you including the Google Play Books link I sent along with my latest submission – but can I ask for the link to be a smidge prettier? :sweat_smile:
image

This is, of course, non-urgent. :slight_smile:

1 Like

heh - yes will do. There are a few I need to do (aozora, google play, app store)… thanks for reminding me :slight_smile:

Edit: @pocketcat prettified!
Screen Shot 2021-07-07 at 11.29.08 PM

2 Likes

Wow! Cool! Now all I need to do is get to level 15 and start reading!

3 Likes

PRODUCT UPDATE

This is a small one for user features, but a big one for me! As a general note, I’ve had a lot of life things in the past 2 weeks, which is why development has been a bit slower. Looking to get back to the grind :upside_down_face:

Summary:

  • privacy policy has slightly changed (added my error alert service sentry, which I forgot about & did add more disclosure)
  • reimplemented cookie popup (it’s been fixed) & cookie policy
  • if you want to update your cookie preferences, you can do so on the cookie policy (reimplemented google analytics collection, which is only collected based on your preferences)
  • overhauled my book addition process - very fast and mostly automated now! Feel free to request as much as you want.
  • you’ll now see 650 unique series on the site now (over 100 additions in the last week)
  • you will see that you’re being prompted for gradings now (this is an old update)

Additional Notes

  • I’ve tested the cookie popup pretty throughly, but let me know if you hit issues
  • I’m working on a general email wrt these updates, just need to figure out how I want to do it :slight_smile:

Upcoming in the next week

  • Adding flagging system for errors on the website
  • Adding explicit content tag
  • Improve grading display
  • Author search / improved search auto complete
8 Likes

Now that the process is mostly automated, would it make sense to just add the top X most popular manga and light novels (according to AniList or MAL)? You’d presumably want to add them without a difficulty score since you’d otherwise just be guessing, but I can see reasons this would be good and reasons this would be bad.

If you don’t do this, some people might get annoyed that a popular book isn’t on the site. Then they’d have to request it (or they might simply not bother), and give it a tentative difficulty rating (assuming they’ve read it already and want to give one). On the other hand, given that the difficulty ratings is the main appeal of the site, others might get annoyed at seeing an unrated book on the site.

I don’t think you can please everyone on this, which is why I’m mentioning it and curious if you were planning to do something like this.

2 Likes

Like you note, I’d have to add an ‘unrated’ functionality before this which should be soon too, but yes once I do that I think that makes sense! I probably will only do that occasionally.

At some point, I do want to allow Amazon search similar to book walker, which would streamline things but there will always be a little bit of manual work involved unfortunately … Natively needs editions & series grouped tightly together whereas Amazon only does it as a convenience and their API doesn’t handle this well. I’ve got it very efficient, but not zero work :slight_smile:

So a long winded way of yes but won’t be a super massive volume. I’ll still mostly depend on user requests I think.

1 Like

It’s going to be quite embarrassing going through my additions and flagging them. :sweat_smile: This is what I get for choosing random books of eBay I guess.

Just don’t forget the Unsubscribe option! :stuck_out_tongue: I’m sure the people in this thread like updates, but the first thing I do when I get an email from any new company (outside of introductory ones of course) is look for the unsub option and mark it as spam if they don’t have it.

4 Likes

Haha no judgement from me! I am thinking I only start with strong sexual content, but we’ll see. It’s a tricky issue, especially since I’m not the one tagging it but relying on users. And there’s also an issue of book covers which show a lot of skin and yet people have said they’re PG. That last use case is especially tricky as one of the main use cases of the tag is prohibiting from search… :thinking:

Heh yes this is why I haven’t sent any mass emails… don’t want my domain tagged as spam!

1 Like

I think the concept of PG varies a lot. Some countries (and individuals) are relatively unfazed by nudity or showing a lot of skin so long as there’s no actual sex and consider it child friendly. Others consider too much skin as problematic in and of itself. Might need to break out sexual themes and nudity? That’s more of an issues for manga, though. With the books I’m referencing it’s more cut and dry.

I wonder if something like ‘Mild sexual themes’ and ‘Strong sexual themes’ would help or confuse… :thinking: I guess a tooltip could be used to clarify but not sure everyone clicks on those.

2 Likes

anyone ever suggest putting video games on here too?

1 Like

本好き went up from level 30 to 31, so a new comparison between it and 狼と香辛料 (level 36) appeared. That was an easy vote. :laughing:

I’m a little sad that I’m only currently reading sequels to things I’ve already added to Natively, because it means no new comparisons for me even when I mark these as finished. :cry:

4 Likes