| I notice a lot of people rate some animes really low without actually watching all of it. I don't know if it is already calculated in this way (If it is, feel free to ignore what I am about to propose). Why don't we weigh the contribution of each rating based on how much they have completed the anime , so whoever finishes the whole series get 100% contribution to the final rating. Assume we have N users with rating r_1, r_2, .... r_N The avg score = (r_1 + r_2 + ... r_N) / N = (1/N) r_1 + (1/N)r_2 + ..... So the contribution of each user's rating is 1/N which is even. I propose the weight for each contribution to be changed from 1/N to c_i / C where C_i is the ep completed by that user, and C is the sum of all eps completed by all N users. For example, if we have 2 users, user 1 watched 9/10 ep and gave it a rating of 9/10, user 2 is a hater that watched 1/10 ep and gave it 2/10. The average score would be ( 2 / (2 + 9)) * 2 + (9 / ( 2 + 9)) * 9 = 7.73 instead of 9/2 + 1/2 = 5 As for updating the score, suppose we have N users and the anime now has score S_N for a new rating given by user N+1, r_(N+1) S_(N+1) = [ S_N * C_N + c_(N+1) * r_(N+1) ] / [C_N + c_(N+1) ] where C_N is the total ep finished by N users. c_N is the number of ep finished by the N-th user. I think this can be effective against haters, because they would have to update the number of episode every week since they can't mark it as completed. |
dragonshadeFeb 7, 2017 9:18 PM
 |
| I would honestly love for it to be weighted like this. There is a weighting system involved, but it's honestly kind of dumb. It's called the "one fifth rule" which is why seasonal anime get a boost when they finish airing. When an anime is currently airing all votes are considered equal and when it ends votes from people who have seen less than 1/5 of it are given a weight of zero and people who have seen at least 1/5 are giving a weight of 1. This not only means people who didn't like the anime so much that they dropped it early are ignored completely making the ratings slightly inflated but they also mean that even people who plan to finish a long anime aren't considered to have valid opinions until they've seen quite a bit of it. You need to watch 73 episodes of Bleach before your rating is taken into account. That is far too long. |
  |
zombie_pegasus said: I would honestly love for it to be weighted like this. There is a weighting system involved, but it's honestly kind of dumb. It's called the "one fifth rule" which is why seasonal anime get a boost when they finish airing. When an anime is currently airing all votes are considered equal and when it ends votes from people who have seen less than 1/5 of it are given a weight of zero and people who have seen at least 1/5 are giving a weight of 1. This not only means people who didn't like the anime so much that they dropped it early are ignored completely making the ratings slightly inflated but they also mean that even people who plan to finish a long anime aren't considered to have valid opinions until they've seen quite a bit of it. You need to watch 73 episodes of Bleach before your rating is taken into account. That is far too long. Oh wow, I didn't know about that. yea, I think one fifth rule lack continuity with it's sharp cut off. |
|
| The only problem I can see with this is that it's possible to set series at any number of episodes watched whilst they are still airing, and there is no action taken against those who do so. Which means that someone could (and people would) manipulate the score by claiming to have watched substantially more episodes than exist for the series. |
| There is no such thing as shit taste. Only idiots who think everyone should have the same taste as they do. |
| No. Doesn't help against hater. They can just wait until the show finished airing and then set it to completed and claim they watched it (you can't prove they didn't watch it) and score it low. Maybe it would help for shows that air endlessly (One Piece or stuff like that) - but they usually already have a score which is too high cause of the 1/5 rule. Also I'd like have the scores of people that only watched a low amount of episodes counted. If a shows is bad and people drop it early because of bad pacing I'd like to know it. With only the scores of the hardcore fans counted (or getting more weight) the score will seem to high while there might be people that don't like it (and might drop it as well after picking it up because of a high score ... but then getting bored cause after 3 episodes it is still boring). |
| I think we don't need this one. I mean if you like a show or hate it why you should care the mean score of it? |
kuuderes_shadow said: The only problem I can see with this is that it's possible to set series at any number of episodes watched whilst they are still airing, and there is no action taken against those who do so. Which means that someone could (and people would) manipulate the score by claiming to have watched substantially more episodes than exist for the series. I am surprised there is no check for that. SAD ! |
|
kuuderes_shadow said: The argument could also be made that people who would set a rewatch as 48/24 completed rather than 24/24 + 1 rewatch would have twice the voting power, but I think there would just need to be an upper limit on what will be counted (100% completed limit and no weight bonuses after 24 episodes watched of a currently airing anime).The only problem I can see with this is that it's possible to set series at any number of episodes watched whilst they are still airing, and there is no action taken against those who do so. Which means that someone could (and people would) manipulate the score by claiming to have watched substantially more episodes than exist for the series. |
|
| I don't fully understand the calculations, but it's definitely better than the current system, as it makes it more fair. |
 |
More topics from this board
» Reply button should show text that will be quoted.Euthymia_Gerv - Apr 15 |
8 |
by Alexioos95
»»
7 hours ago |
|
» @ sign spam/attackkuroneko99 - Apr 16 |
3 |
by Jhiday
»»
Yesterday, 11:11 AM |
|
» Preview for "About Me" sectionNubFix - Mar 24, 2022 |
5 |
by NubFix
»»
Apr 16, 10:16 PM |
|
» Could MAL match users with the most similar watched anime and ratings?mur_koshka - Apr 7 |
4 |
by mur_koshka
»»
Apr 16, 1:44 AM |
|
» Users who haven't finished an anime should be restricted on rating itThruddingsfly - Feb 15 |
39 |
by 3miL
»»
Apr 14, 11:07 AM |