Forum SettingsEpisode Information
Forums

What is happening with the score (this is just technical question, not a question like "why is the score so high?")

Pages (2) [1] 2 »
Post New Reply
#1
Jun 17, 1:09 PM
Offline
Joined: May 2016
Posts: 332
Today, I saw the score went up 9.20, with ~72000 votes. But know the score dropped down to 9.16 again, and there is only ~58000 votes. Suddenly ~14000 votes disappeared?

What happened?

This is just a question about how MAL's score system works, please do not share any opinions about the score itself (it doesn't encourage discussion).

I just want an explanation to this strange phenomenon.
 
#2
Jun 17, 1:16 PM

Offline
Joined: Apr 2019
Posts: 543
i didn't even notice that there was a sudden increase/decrease in votes
interesting

basically something has to have changed in that time frame for that to happen
maybe they are modifying the score system? probably not
very curious
You son of a .. turtle

 
#3
Jun 17, 1:21 PM
Offline
Joined: May 2016
Posts: 332
Nim0174 said:
i didn't even notice that there was a sudden increase/decrease in votes
interesting

basically something has to have changed in that time frame for that to happen
maybe they are modifying the score system? probably not
very curious


Unfortunately I don't have screenshots, but I'm sure that I saw what I've said...

Or maybe it didn't happen, I just went crazy xD
 
#4
Jun 17, 1:24 PM
Offline
Joined: Jun 2016
Posts: 20
SheevPalpatine said:
Nim0174 said:
i didn't even notice that there was a sudden increase/decrease in votes
interesting

basically something has to have changed in that time frame for that to happen
maybe they are modifying the score system? probably not
very curious


Unfortunately I don't have screenshots, but I'm sure that I saw what I've said...

Or maybe it didn't happen, I just went crazy xD


No, the score definitely changed as you said. I pointed this out on another forum:

https://myanimelist.net/forum/?topicid=1726590&show=50
 
#5
Jun 17, 1:28 PM
Offline
Joined: Jun 2019
Posts: 2
What you saw did happened for a while and I was also surprised. It could simply be some technical issue, since no of votes went from 58k to 72k and then back to 58k.
 
#6
Jun 17, 1:30 PM

Offline
Joined: Apr 2019
Posts: 543
SheevPalpatine said:
Nim0174 said:
i didn't even notice that there was a sudden increase/decrease in votes
interesting

basically something has to have changed in that time frame for that to happen
maybe they are modifying the score system? probably not
very curious


Unfortunately I don't have screenshots, but I'm sure that I saw what I've said...

Or maybe it didn't happen, I just went crazy xD


no no it makes sense
i saw it go to 9.2
now i know why
very very strange
You son of a .. turtle

 
#7
Jun 17, 1:32 PM

Offline
Joined: Jan 2015
Posts: 1805
https://imgur.com/a/vGzvRda

I got a sc, but the members are not included.
 
#8
Jun 17, 1:37 PM

Online
Joined: Oct 2013
Posts: 10312
Its an error it happens sometimes, if snk wasn't getting this much 1/10 it probably would already be at number 1 already.

 
#9
Jun 17, 1:39 PM
Offline
Joined: Jan 2009
Posts: 46126
server glitch happens duh lol pay for MAL supporters https://myanimelist.net/membership to help them buy more needed servers XD
 
Jun 17, 1:49 PM

Offline
Joined: Dec 2008
Posts: 504
SheevPalpatine said:
Nim0174 said:
i didn't even notice that there was a sudden increase/decrease in votes
interesting

basically something has to have changed in that time frame for that to happen
maybe they are modifying the score system? probably not
very curious


Unfortunately I don't have screenshots, but I'm sure that I saw what I've said...

Or maybe it didn't happen, I just went crazy xD


I have one here......
 
Jun 17, 2:03 PM
Offline
Joined: May 2016
Posts: 332
Maybe it was just a bug afterall
 
Jun 17, 2:07 PM
Offline
Joined: Jan 2009
Posts: 46126
SheevPalpatine said:
Maybe it was just a bug afterall


same thing happened with One Punch Man just look at this thread there https://myanimelist.net/forum/?topicid=1732747

so ye like kerragaming said this is a known error
 
Jun 17, 2:16 PM

Online
Joined: Feb 2017
Posts: 1001
deg said:
SheevPalpatine said:
Maybe it was just a bug afterall


same thing happened with One Punch Man just look at this thread there https://myanimelist.net/forum/?topicid=1732747

so ye like kerragaming said this is a known error


It also happened to Season 1 some months ago. It jumped to 8.60 then back to 8.48. Also before it jumped to 8.60, it went down to 8.33 I think.
 
Jun 17, 4:39 PM
Offline
Joined: Dec 2015
Posts: 232
It's the invisible hand. ;)
 
Jun 17, 6:24 PM
Offline
Joined: Feb 2018
Posts: 68
This happens a lot, very curious to see what happens ..
Signature removed. Please follow the signature rules, as defined in the Site & Forum Guidelines.
 
Jun 17, 6:39 PM
くみこ

Offline
Joined: Feb 2019
Posts: 923
Flawed rating system OP
 
Jun 17, 6:56 PM

Offline
Joined: Oct 2013
Posts: 1104
@ChrnoTodd , woah, so it really happened! Impressive. I thought 9.17 was the maximum score this entry got so far. Nice, if this trend continues, then after episode 10 we might see stable 9.20 score. No need to say that if hypothetically SnK reaches 9.25 stable score, then first place among animes with the best score will belong to "Attack on Titan". :P Season 3 Part 2, but still "Attack on Titan".
 
Jun 17, 11:32 PM
Offline
Joined: May 2017
Posts: 107
Adnash93 said:
@ChrnoTodd , woah, so it really happened! Impressive. I thought 9.17 was the maximum score this entry got so far. Nice, if this trend continues, then after episode 10 we might see stable 9.20 score. No need to say that if hypothetically SnK reaches 9.25 stable score, then first place among animes with the best score will belong to "Attack on Titan". :P Season 3 Part 2, but still "Attack on Titan".


This was a glitch tho. It never actually reached 9.2
 
Jun 19, 7:48 AM

Offline
Joined: Oct 2013
Posts: 1104
blackiemma22 said:
This was a glitch tho. It never actually reached 9.2

Oh. Didn't known that. Well, but nevertheless I'm curious whether after 10th episode the score will go above 9.20 ;p
 
Jun 19, 7:57 AM
Offline
Joined: May 2016
Posts: 332
Adnash93 said:
blackiemma22 said:
This was a glitch tho. It never actually reached 9.2

Oh. Didn't known that. Well, but nevertheless I'm curious whether after 10th episode the score will go above 9.20 ;p


It will drop. The score is high currently because only the hardest fans give a 10/10 score to a show which is not finished airing.
 
Jun 19, 8:09 AM

Online
Joined: Oct 2013
Posts: 10312
SheevPalpatine said:
Adnash93 said:

Oh. Didn't known that. Well, but nevertheless I'm curious whether after 10th episode the score will go above 9.20 ;p


It will drop. The score is high currently because only the hardest fans give a 10/10 score to a show which is not finished airing.


Not true, the snk seasons has never been this high, the fact that season 3 part 1 is rated 8.48 while part 2 is currently rated 9.17 means that a lot more people are giving this season a 10/10.

60k is a good sample size to basically predict what other people will give this season and since this season will have a good ending. I expect the score to go up once it finishes.

Now will it keep that score is a different story.

I have gotten close real life friends that always tell me how much snk sucks, but to my surprise they are saying this season is far better than the previous seasons. It means that this season is getting more positive score than negative, which is why it is so high to begin with.

Hardcore snk fans fail to make the previous seasons of snk even sniff the 9 range, so this high score must be also from persons that are liking this season more.
Modified by keragamming, Jun 19, 8:13 AM
 
Jun 19, 8:13 AM

Offline
Joined: Oct 2015
Posts: 701
keragamming said:
SheevPalpatine said:


It will drop. The score is high currently because only the hardest fans give a 10/10 score to a show which is not finished airing.


Not true, the snk season has never been this high, the fact that season 3 part 1 is rated 8.48 while part 2 is currently rated 9.17 means that a lot more people are giving this season a 10/10.

60k is a good sample size to basically predict what other people will give this season and since this season will have a good ending. I expect the score to go up once it finishes.

Now will it keep that score is a different story.
I also believe we have a potential 'winner' here. Attack on Titan is obviously at this point going to be remembered and talked about as that one show everyone said was 'edgy' and relied on being 'mindless rule of cool' and went on to become one of the most tightly written stories in anime with the biggest amount of foreshadowing since day 1. It makes total sense that many many more people are giving it a 10 now.
 
Jun 19, 8:17 AM

Online
Joined: Oct 2013
Posts: 10312
Nostalgik said:
keragamming said:


Not true, the snk season has never been this high, the fact that season 3 part 1 is rated 8.48 while part 2 is currently rated 9.17 means that a lot more people are giving this season a 10/10.

60k is a good sample size to basically predict what other people will give this season and since this season will have a good ending. I expect the score to go up once it finishes.

Now will it keep that score is a different story.
I also believe we have a potential 'winner' here. Attack on Titan is obviously at this point going to be remembered and talked about as that one show everyone said was 'edgy' and relied on being 'mindless rule of cool' and went on to become one of the most tightly written stories in anime with the biggest amount of foreshadowing since day 1. It makes total sense that many many more people are giving it a 10 now.


I always said it even since the first season that people are just not seeing the bigger picture and is viewing snk at face value, I will admit after season 3 part 1, that snk will be the series that is misunderstand by many.

Thankfully this season is the turning point, from here on now, people will see just how good and well written this season really is.

I can just imagine when season 4 comes around.
 
Jun 19, 8:30 AM
Offline
Joined: May 2016
Posts: 332
keragamming said:
SheevPalpatine said:


It will drop. The score is high currently because only the hardest fans give a 10/10 score to a show which is not finished airing.


Not true, the snk seasons has never been this high, the fact that season 3 part 1 is rated 8.48 while part 2 is currently rated 9.17 means that a lot more people are giving this season a 10/10.

60k is a good sample size to basically predict what other people will give this season and since this season will have a good ending. I expect the score to go up once it finishes.

Now will it keep that score is a different story.

I have gotten close real life friends that always tell me how much snk sucks, but to my surprise they are saying this season is far better than the previous seasons. It means that this season is getting more positive score than negative, which is why it is so high to begin with.

Hardcore snk fans fail to make the previous seasons of snk even sniff the 9 range, so this high score must be also from persons that are liking this season more.


It will drop to between 9.00 and 9.10. That's my prediction.

"I have gotten close real life friends that always tell me how much snk sucks, but to my surprise they are saying this season is far better than the previous seasons"

It doesn't mean that they will give 10 ;) I also think that this season is probably the best, but I won't give a 10, because of some issues.

But this is offtopic here, this thread is about the technical problems of MAL.
 
Jun 19, 8:38 AM

Online
Joined: Oct 2013
Posts: 10312
@SheevPalpatine

Just for clarification, do you mean as soon as it is completed it will drop to 9.00 - 9.10? after a day or two? Or do you mean it will drop to that after a Month or more?

The reason why I am saying that is, normally when a series ends for example mob psycho 100 season 2, the score goes up and then after a while it goes down.

Unless the ending is bad, it would go down quickly, but most cases the scores normally goes up and then slowly goes down.

I predict that when it ends there is a good chance it will surpass fmab, but most likely will go down, especially with the 1/10 votes. Also if snk had a 0.3% 1/10 vote like fmab it would be at number 1 already, because it currently has the highest 10/10 percentage on mal.

But we will see who will get it right in the next two weeks. ;)
 
Jun 19, 8:43 AM
Offline
Joined: May 2016
Posts: 332
keragamming said:
@SheevPalpatine

Just for clarification, do you mean as soon as it is completed it will drop to 9.00 - 9.10? after a day or two? Or do you mean it will drop to that after a Month or more?

The reason why I am saying that is, normally when a series ends for example mob psycho 100 season 2, the score goes up and then after a while it goes down.

Unless the ending is bad, it would go down quickly, but most cases the scores normally goes up and then slowly goes down.

I predict that when it ends there is a good chance it will surpass fmab, but most likely will go down, especially with the 1/10 votes. Also if snk had a 0.3% 1/10 vote like fmab it would be at number 1 already, because it currently has the highest 10/10 percentage on mal.

But we will see who will get it right in the next two weeks. ;)


I mean after some weeks, not immediately.
 
Jun 19, 10:22 AM

Offline
Joined: Apr 2019
Posts: 543
Nostalgik said:
keragamming said:


Not true, the snk season has never been this high, the fact that season 3 part 1 is rated 8.48 while part 2 is currently rated 9.17 means that a lot more people are giving this season a 10/10.

60k is a good sample size to basically predict what other people will give this season and since this season will have a good ending. I expect the score to go up once it finishes.

Now will it keep that score is a different story.
I also believe we have a potential 'winner' here. Attack on Titan is obviously at this point going to be remembered and talked about as that one show everyone said was 'edgy' and relied on being 'mindless rule of cool' and went on to become one of the most tightly written stories in anime with the biggest amount of foreshadowing since day 1. It makes total sense that many many more people are giving it a 10 now.


well said
in fact i was not a big fan until i read through the next arcs
at this point i've rewatched and reread all of it multiple times
isayama pays so much attention to tiny details
You son of a .. turtle

 
Jun 19, 5:24 PM

Offline
Joined: Oct 2013
Posts: 1104
@SheevPalpatine , I must disagree with you and completely agree with what @keragamming wrote earlier. Many people have already given their votes since the show began airing, not only those who are the hardest fans. I will just look at myself. I like this anime, even I like very much, but I'm not a hardcore fan of the franchise as a whole, since I haven't read the manga (and I do not plan to, at least in the nearest future) and I am not crazy each time I hear or see "Attack on Titan reference". Despite this, I gave 10/10 score to this title shortly after I began watching it. Why? Because it is such a great work and it satisfies my sense of quality in every aspect: visuals, music, characters etc.

We have only 2 episodes left and the score, instead of maybe decreasing a bit, increased and that's the fact. I'm positive about possiblity of its score going up once "Attack on Titan: Season 3 Part 2" finishes. By then many people who haven't already given it a score might do it, and I think there are many people who like to give their score to animes once either they finish watching them or after the last episode is released. Mostly it's strongly connected, 'cause if somebody watches an anime regularly during the period it is being aired, then it's obvious that he/she will be rather in touch with all episodes as the schedule goes.
 
Jun 19, 6:38 PM

Offline
Joined: Mar 2013
Posts: 464
this season should stay number 2 spot! I really love this series.
 
Jun 19, 6:51 PM
Offline
Joined: Dec 2015
Posts: 232
Cloude said:
this season should stay number 2 spot! I really love this series.



At this point, I very much believe it will. The next episode will likely bump the score beyond 9.20, that will be more than enough room to compensate for the outbound hype train.
 
Jun 19, 7:33 PM

Offline
Joined: Jul 2016
Posts: 502
With booster account post in illegitimate forum. I see that the amount gonna slow and stagnant at the end. And the downvoters cleaned. We will see the true score to fall about a month after the anime ends.


 
Jun 19, 7:46 PM

Online
Joined: Oct 2013
Posts: 10312
chriskor022 said:
With booster account post in illegitimate forum. I see that the amount gonna slow and stagnant at the end. And the downvoters cleaned. We will see the true score to fall about a month after the anime ends.


Those 1/10 voters are doing a lot more damage to the series, the thing is most of them are not bots and are just haters so their account can't be deleted and thus the troll vote will stay. I honestly believe they should ghost the 1/10 votes from the rating system, since 99% of the time it is use for down voting a series score.
 
Jun 20, 4:30 AM

Offline
Joined: Aug 2016
Posts: 42
keragamming said:
I honestly believe they should ghost the 1/10 votes from the rating system, since 99% of the time it is use for down voting a series score.



Or just cap 1/10 votes to be less than or equal to 2/10 votes, except in cases where the average is very low.
 
Jun 20, 4:32 AM
Offline
Joined: May 2016
Posts: 486
keragamming said:
chriskor022 said:
With booster account post in illegitimate forum. I see that the amount gonna slow and stagnant at the end. And the downvoters cleaned. We will see the true score to fall about a month after the anime ends.


Those 1/10 voters are doing a lot more damage to the series, the thing is most of them are not bots and are just haters so their account can't be deleted and thus the troll vote will stay. I honestly believe they should ghost the 1/10 votes from the rating system, since 99% of the time it is use for down voting a series score.

agreed, it's insanely petty how the series' detractors are willing to go that far just because the series has hit critical success.
 
Jun 20, 5:25 AM

Offline
Joined: Oct 2013
Posts: 1104
Sanic_de_Hegehog said:
keragamming said:
I honestly believe they should ghost the 1/10 votes from the rating system, since 99% of the time it is use for down voting a series score.



Or just cap 1/10 votes to be less than or equal to 2/10 votes, except in cases where the average is very low.

It wouldn't work to be honest. But my that, the score of 2/10 would eventually become a new "tool of downvoting" used by trolls.
 
Jun 20, 6:30 AM

Online
Joined: Oct 2013
Posts: 10312
Adnash93 said:
Sanic_de_Hegehog said:



Or just cap 1/10 votes to be less than or equal to 2/10 votes, except in cases where the average is very low.

It wouldn't work to be honest. But my that, the score of 2/10 would eventually become a new "tool of downvoting" used by trolls.


That is why I was saying ghost the 1/10 without persons knowing that the 1/10 score is excluded from the rating system.

Just like how someone can get shadow or ghost ban, like on forums where they can comment, but only they them self will see their comment.
 
Jun 20, 6:36 AM

Offline
Joined: Oct 2013
Posts: 1104
@keragamming , but wouldn't it make 1/10 score pointless in general if it was ghosted? Plus let's not forget about people who are not haters, but can't give a score over 1/10 to one anime they really disliked and that looked terrible even in the eyes of people who were not its audience.
 
Jun 20, 7:15 AM

Online
Joined: Oct 2013
Posts: 10312
Adnash93 said:
@keragamming , but wouldn't it make 1/10 score pointless in general if it was ghosted? Plus let's not forget about people who are not haters, but can't give a score over 1/10 to one anime they really disliked and that looked terrible even in the eyes of people who were not its audience.


They can still give the 1/10 score on their list, it would just not be calculated on the overall rating without their knowledge.

Majority of the time the 1/10 votes are use solely for rating down a series score, the most annoying part about it is that majority of those 1/10 are from persons that haven't even watched the series.

Ok, take a look at snk part 2 https://myanimelist.net/anime/38524/Shingeki_no_Kyojin_Season_3_Part_2/stats

Look at the trend and look how the percentage of the scores goes down as the score lowers, but for some reason 1/10 scores has a higher percentage than the scores starting from 5/10 to 2/10. If we are going by the trend the 1/10 votes should even be lower than the 2/10 votes.

And this is basically the case for all the top series, some to a lesser extend, but it happens to all.

1/10 scores at this point is solely use to downvote a series, only the really bad series that are rated in the 6 range and lower will most likely have legit 1/10 scores since there would be any real reason to downvote them since their score is already low and isn't competing with their favorite anime.
Modified by keragamming, Jun 20, 7:24 AM
 
Jun 20, 7:28 AM

Offline
Joined: Feb 2018
Posts: 298
keragamming said:
Adnash93 said:
@keragamming , but wouldn't it make 1/10 score pointless in general if it was ghosted? Plus let's not forget about people who are not haters, but can't give a score over 1/10 to one anime they really disliked and that looked terrible even in the eyes of people who were not its audience.


They can still give the 1/10 score on their list, it would just not be calculated on the overall rating without their knowledge.

Majority of the time the 1/10 votes are use solely for rating down a series score, the most annoying part about it is that majority of those 1/10 are from persons that haven't even watched the series.

Ok, take a look at snk part 2 https://myanimelist.net/anime/38524/Shingeki_no_Kyojin_Season_3_Part_2/stats

Look at the trend and look how the percentage of the scores goes down as the score lowers, but for some reason 1/10 scores has a higher percentage than the scores starting from 5/10 to 2/10. If we are going by the trend the 1/10 votes should even be lower than the 2/10 votes.

And this is basically the case for all the top series, some to a lesser extend, but it happens to all.

1/10 scores at this point is solely use to downvote a series, only the really bad series that are rated in the 6 range and lower will most likely have legit 1/10 scores since there would be any real reason to downvote them since their score is already low and isn't competing with their favorite anime.


Basically you need 10 10/10 to normalize a single 1/10
 
Jun 20, 8:08 AM
Offline
Joined: May 2019
Posts: 53
Shows how pettty a lot of fmab and stein's gate fans are and just how superficial their appreciation for their shows are by how much of it's tied to public validation.
 
Jun 20, 8:45 AM
Offline
Joined: Mar 2017
Posts: 1
keragamming said:
Adnash93 said:
@keragamming , but wouldn't it make 1/10 score pointless in general if it was ghosted? Plus let's not forget about people who are not haters, but can't give a score over 1/10 to one anime they really disliked and that looked terrible even in the eyes of people who were not its audience.


They can still give the 1/10 score on their list, it would just not be calculated on the overall rating without their knowledge.

Majority of the time the 1/10 votes are use solely for rating down a series score, the most annoying part about it is that majority of those 1/10 are from persons that haven't even watched the series.

Ok, take a look at snk part 2 https://myanimelist.net/anime/38524/Shingeki_no_Kyojin_Season_3_Part_2/stats

Look at the trend and look how the percentage of the scores goes down as the score lowers, but for some reason 1/10 scores has a higher percentage than the scores starting from 5/10 to 2/10. If we are going by the trend the 1/10 votes should even be lower than the 2/10 votes.

And this is basically the case for all the top series, some to a lesser extend, but it happens to all.

1/10 scores at this point is solely use to downvote a series, only the really bad series that are rated in the 6 range and lower will most likely have legit 1/10 scores since there would be any real reason to downvote them since their score is already low and isn't competing with their favorite anime.


Wheres the difference between people that give it a 1 and those who give it a 10?
This whole thread becomes the same thing it always becomes when it's about a high scoring anime.
People rate it 1 because they hate it, then those people who gave it a 10 before it even finished are flaming these people. They then flame them for giving it a 1 because they are just "kack haiders" while giving it a 10 being the same thing turned around. It is just anoying. Just et people hate it the same way you love it it doesn't matter anyway
Modified by xJublex, Jun 20, 8:49 AM
 
Jun 20, 9:05 AM
Offline
Joined: May 2016
Posts: 332
Please, stop the offtopic, this is a thread about a technical issue, not about the score itself. If you don't have any comment on MAL's bug, then leave this topic. Thank you!

Conversation about why so many 10/10 and 1/10 goes nowhere. So stop it, please, it has no sense!

(Btw, the main question of this topic was answered, so if a mod see this, can he/she close the topic? Thank you!)
 
Jun 20, 9:11 AM

Online
Joined: Oct 2013
Posts: 10312
xJublex said:
keragamming said:


They can still give the 1/10 score on their list, it would just not be calculated on the overall rating without their knowledge.

Majority of the time the 1/10 votes are use solely for rating down a series score, the most annoying part about it is that majority of those 1/10 are from persons that haven't even watched the series.

Ok, take a look at snk part 2 https://myanimelist.net/anime/38524/Shingeki_no_Kyojin_Season_3_Part_2/stats

Look at the trend and look how the percentage of the scores goes down as the score lowers, but for some reason 1/10 scores has a higher percentage than the scores starting from 5/10 to 2/10. If we are going by the trend the 1/10 votes should even be lower than the 2/10 votes.

And this is basically the case for all the top series, some to a lesser extend, but it happens to all.

1/10 scores at this point is solely use to downvote a series, only the really bad series that are rated in the 6 range and lower will most likely have legit 1/10 scores since there would be any real reason to downvote them since their score is already low and isn't competing with their favorite anime.


Wheres the difference between people that give it a 1 and those who give it a 10?
This whole thread becomes the same thing it always becomes when it's about a high scoring anime.
People rate it 1 because they hate it, then those people who gave it a 10 before it even finished are flaming these people. They then flame them for giving it a 1 because they are just "kack haiders" while giving it a 10 being the same thing turned around. It is just anoying. Just et people hate it the same way you love it it doesn't matter anyway


Big difference is that they haven't even watched this season. But I will stop here, since this is off topic.
 
Jun 20, 7:07 PM

Offline
Joined: Oct 2013
Posts: 1104
@keragamming , I understand clearly what you wrote and meant earlier about people using 1/10 score to downvote certain titles they dislike. I just wanted to point out that making 1/10 score ghosted is not a solution.

By that, 2/10 score will "receive" the position of a score used to downvote shows, as for it won't be ghosted and will be perfect for haters to use it against titles they want to downvote.

The problem with extreme scores and their measurment is that there are no good solutions to it and we need just to accept the situation as it is: some people will downvote "for fun" certain titles by giving them 1/10 score, and some people will give 10/10 scores just because of felt hype. Too bad lately there are more people downvoting to those who are just hyped or really enjoyed something, even if it could be objectively not that good.
And then comes the problem with minority... people who are fair and use their own individualised criterias to give certain scores to certain titles. For them giving 1/10 or 10/10 is not a mistake, "troll's playground" or another train on hypetrain's journey, but just an effect of thinking out everything and judging whether i.e. an anime or a manga deserves the highest or the lowest score.
 
Jun 20, 8:59 PM

Online
Joined: Oct 2013
Posts: 10312
Adnash93 said:
@keragamming , I understand clearly what you wrote and meant earlier about people using 1/10 score to downvote certain titles they dislike. I just wanted to point out that making 1/10 score ghosted is not a solution.

By that, 2/10 score will "receive" the position of a score used to downvote shows, as for it won't be ghosted and will be perfect for haters to use it against titles they want to downvote.

The problem with extreme scores and their measurment is that there are no good solutions to it and we need just to accept the situation as it is: some people will downvote "for fun" certain titles by giving them 1/10 score, and some people will give 10/10 scores just because of felt hype. Too bad lately there are more people downvoting to those who are just hyped or really enjoyed something, even if it could be objectively not that good.
And then comes the problem with minority... people who are fair and use their own individualised criterias to give certain scores to certain titles. For them giving 1/10 or 10/10 is not a mistake, "troll's playground" or another train on hypetrain's journey, but just an effect of thinking out everything and judging whether i.e. an anime or a manga deserves the highest or the lowest score.


But what I am saying is that, if I was apart of the coding team, we would have done it without announcing it to the public, that way persons continue to waste their time giving 1/10 votes when in reality it would not affect the series.

I think that is what Imdb does as well to counter the hate votes without making persons being aware of it. That is why snk part 2 episode 5 is still number 1 with 10/10 score, even though it has 5.9% of 1/10 votes. https://www.imdb.com/title/tt9906260/ratings?ref_=tt_ov_rt

I don't think their is anyway to make the rating system fair, but that was a solution when it comes to trolling.
Modified by keragamming, Jun 20, 9:02 PM
 
Jun 21, 1:00 AM
Offline
Joined: Feb 2018
Posts: 46
Zoroft said:
Shows how pettty a lot of fmab and stein's gate fans are and just how superficial their appreciation for their shows are by how much of it's tied to public validation.
i tend to disagree. Fmab and steins gate have many hate votes too and the absolute number of 1's they have is more than snk!! Its just that when the season will end and more people starts voting, the percentage of 1s will star falling to sometging around 0.4 percent which is the norm for high rated shows.
 
Jun 21, 6:48 AM
Offline
Joined: Mar 2018
Posts: 37
I want this and vinland saga to dominate tbh.
 
Jun 21, 9:25 AM

Offline
Joined: Mar 2015
Posts: 26
off topic, but i hope still useful to find a more objective score. i will stop here too ...

I only use this method personally, to check how objectively the score obtained by an anime/manga.

You have to calculate the normal score (between 3 and 8) all weighing 1.
While for the upvote score (9-10) and the downvote score (for example 2-1) their weight must be calculated differently.

Weight upvote (WU) = the number of people who give 9 votes divided by the number of people who give 10 votes.
Weight downvote (WD) = the number of people who give 2 votes divided by the number of people who give 1 votes.
The closer the WU and WD to weigh 1, meaning the score is more reliable (valid).

For example School Days :
School Days = 5.95 [MAL] / 6.02 ghosted score [excluding 1 score] / 6.01 ['objective']
(WU 1.08 - WD 0.89).
The score for School Days more reliable than AOT, FMAB and all animes in top most 10, because WU and WD points more close to 1. And the MAL score, ghosted score and 'objective' score not too much different.

For ranks (top 3) it will be like this :

Attack on Titan S3 Part 2
9.18 (mean) ↑ 9.26 (ghosting1) | 9.18 (MAL) ↓8.99 ('objective')
(WU 0.5- WD 0.08)

3 Gatsu no Lion S2
9.01 (mean) ↑ 9.03 (ghosting1) | 9.04 (MAL) ↓8.93 ('objective')
(WU 0.79 - WD 0.24)

Full Metal Alchemist Brotherhood
9.18 (mean) ↑ 9.20 (ghosting1) | 9.24 (MAL) ↓8.92 ('objective')
(WU 0.5- WD 0.2 ]


Shingeki no Kyojin S3 Part 2 more hated than FMAB (and some others series), because WD point = 0.08 (2 = 53 votes, 1 = 626 votes at the moment).
The score can be much higher, if those who give 1 votes are almost equivalent or less to those who give 2 votes.

Note : I use 'objective' (with quotes), because this is my own method - not official formula.
Modified by maem, Jun 21, 10:33 PM
 
Jun 21, 9:01 PM

Offline
Joined: Aug 2016
Posts: 42
maem said:

For ranks (top 3) it will be like this :

Shingeki no Kyojin S3 Part 2 = 9.17 [MAL] / 9.25 [ghosted 1] / 8.98 ['objective']
(WU 0.5- WD 0.08) ]
3-gatsu no Lion 2nd Season = 9.04 [MAL] / 9.03 [ghosted 1] / 8.93 ['objective']
(WU 0.79 - WD 0.24)]
Full Metal Alchemist Brotherhood = 9.24 [MAL] / 9.20 [ghosted 1] / 8.92 ['objective']
(WU 0.5-0.2 WD)]



Your math is out - ghosting the 1/10 vote should always make the score increase. Why go 3-gastu and FMA:B go down? The other problem with the method is that the upvote effect is much harder to quantify than the downvote since 'upvotes' are mixed in with legitimate votes. There is no way to determine how many are real. For 1/10 it is much easier to observe since there is an obvious spike going from 2/10 to 1/10. Assuming a spike from 2/10 to 1/10 being a downvote is a reasonable assumption to make statistically, but doing the same for a 9/10 to 10/10 spike doesn't really work since the show could just be 'that good'.

To develop a good 'objective' algorithm many different potential score distributions need to be analysed and taken into consideration. For example, you know something fishy is going on if the curve changes direction more than once (it shouldn't change direction at all in the first example).

Modified by Sanic_de_Hegehog, Jun 21, 9:38 PM
 
Jun 21, 10:29 PM

Offline
Joined: Mar 2015
Posts: 26
Sanic_de_Hegehog said:



To develop a good 'objective' algorithm many different potential score distributions need to be analysed and taken into consideration. For example, you know something fishy is going on if the curve changes direction more than once (it shouldn't change direction at all in the first example).


You're right. I forgot to put real mean score (without MAL formula).
I edit my post with current score , and yes; my algorithm still questionable .
 
Top
Pages (2) [1] 2 »

Preview MangaManga Store