Report hacker09's Profile

Statistics

All Anime Stats Anime Stats
Days: 167.8
Mean Score: 6.37
  • Total Entries7,477
  • Rewatched77
  • Episodes12,285
Anime History Last Anime Updates
Tokyo Revengers: Tenjiku-hen
Tokyo Revengers: Tenjiku-hen
55 minutes ago
Watching 1/13 · Scored -
Coffee Kousou
Coffee Kousou
2 hours ago
Completed 1/1 · Scored 5
Tokyo Revengers: Seiya Kessen-hen
Tokyo Revengers: Seiya Kessen-hen
2 hours ago
Completed 13/13 · Scored 7
All Manga Stats Manga Stats
Days: 0.1
Mean Score: 7.00
  • Total Entries2
  • Reread0
  • Chapters1
  • Volumes1
Manga History Last Manga Updates
Naruto
Naruto
Jun 12, 2021 1:31 PM
Dropped -/700 · Scored -
Mario
Mario
Jan 6, 2020 12:05 PM
Dropped 1/1 · Scored 7

All Favorites Favorites

Anime (10)
Character (10)
People (10)

All Comments (1184) Comments

Would you like to post a comment? Please login or sign up first!
ShaggyZE May 26, 4:23 PM
sorry for the late reply, I'm good, but been busy, noticed you have been too, with school.

I never really used Endless Mal because of the issues and it's not compatible with my dark mode theme/userstyle plus there is probably a lot of edits I'd have to do to make it work right with it.

I will try to look at it soon, when I have time and just disable stylus.
_nelt May 21, 11:32 PM
I won't be able to make a % enjoyment score as I don't do data analysis on users, I'll let you know if I implement the score normalization in the future.

Images come directly out of the MAL api, some times they have weird resolutions but there's nothing I can do.

If you need bigger ones you can put an "l" before ".jpg": ".../anime/12/50345.jpg"> ".../anime/12/50345l.jpg", but this will
increase the size and not fix the aspect ratio
url.replace(".jpg", "l.jpg)


Also the broken images are a MAL api problem, as they some times move the images. they get fixed eventually but it takes some time.
_nelt May 21, 4:05 AM
That works for a random sample of users, but this one is with users having a high % similarity to you, so, in theory, their average score doesn't necessarily tend towards the MAL mean score and should be representative of your taste.

The actual problem in this case is when there are too few users and the score leans towards their scoring habits (like giving everything a 9 or 10),
that would need some sort of score normalization, that I may implement in some distant future.

Let me know about your app!
_nelt May 20, 11:56 PM
I don't have the type in my database :/

The entry score is the average of the similar users scores, it's not exactly an enjoyment score, but it should do.
_nelt May 20, 10:14 PM
Yeah I guess you could say so
_nelt May 20, 10:08 PM
https://myanimelist.net/anime/ + id makes the mal url

it's the anime from the most similar users ordered randomly (actually: random() * 100 + score),
the anime you've already watched, sequels and side stories are excluded
_nelt May 19, 2:17 AM
_nelt May 19, 1:06 AM
Ok I'll try something and let you know
_nelt May 19, 12:41 AM
I could make it to be something similar to maki, but why don't you just use that instead?
_nelt May 18, 4:45 PM
I guess I could make a /random endpoint to let you get random anime from similar users, but you may end up with terrible recommendations
_nelt May 18, 4:33 PM
Yeah that'd be very hard to do with the similar users concept, you'd need something more randomized
_nelt May 18, 4:22 PM
Yo

The main purpose of this api is to find similar users (/hacker09/similar), while recommendations are simply the anime you haven't seen from those users lists.
Its very structure makes it so that recommendations change only when your list or other users lists change (or new users are matched).

To overcome this issue you can get up to 20 pages of recommendations (for performance reasons the page size is fixed to 16)
You could request 2 or 3 pages together without getting rate limited, but the ideal way would be to request one page at a time on some user input (like a "load more" button)

The batch offsets the similar users that provide the recommendations,
for example, batch=1 may find users with similarity 95-85%, batch=2 similarity 85-80%, and so on
However changing batch doesn't necessarily mean to get fresh recommendations.

You can try it here, the sidebar on the left contains the list of users that are providing the recommendations,
also you can click on a user to load that specific user anime you haven't seen.
You can change the users batch using the < 1 > buttons on top of the sidebar.

What do you mean by making it similar to maki-web?
Kasea Apr 28, 3:22 AM
Hi, just found out about all the tampermonkey scripts you've made, haven't started using them yet but thank you for making a bunch of QoL functionality and making the site better/easier to use. Greatly appreciated since the developers will do anything to avoid changing the site themselves.

(●'◡'●)
Sethoitae Apr 19, 4:59 PM
MMEEP MEEPHIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
Sethoitae Apr 19, 4:38 PM
MMRRREEEEEEPPPPPPPPPPPPPPPPP
It’s time to ditch the text file.
Keep track of your anime easily by creating your own list.
Sign Up Login