 I'm not normally one to start a conspiracy about the legitimacy of MAL scores; however, based on the average score distribution plot of shows on MAL, there is no way this can be correct. The amount of 1 ratings is simply not feasible. It's not just compared to the way the lower scores are gradually decreasing - just scroll through the pages of ratings and literally nobody is giving it a 1. This had to have been done all at once. Just take a look at the Neverland score distribution:  Even though there is definitely a noticeable increase in people that gave it a 1 vs a 2 (a 5.31:1 ratio), it's not nearly as dramatic as for Kaguya (a 9.25:1 ratio). Thoughts? Also, is there an admin here who would be able to investigate this? |
phiraethMar 16, 2019 4:58 PM
| I think there have been other threads about this topic, particularly that it seemed like there were times where a large number of accounts would rate the show a 1 at the same time. I'm inclined to think someone just really does not like this show and has a lot of time. Definitely seems sus to me, but I also love this show to death |
 |
| Yeah, it could be worked on soon once things kicked in in the ''higher-up'' - section, take a look at the ''Illegitimate Accounts'' thread in Support. |
| When a series concludes airing if someone has the series either watching at less than a 3rd of the series or has it dropped at less than a 3rd then those scores are not counted towards the overall series scorehence why alot of series especially with weaker starts tend to have a massive boost in score when they end, so a decent amount of these bot scores will be negligible (this happens alot for just about all popular shows especially when they start reaching a top ranking). Also series that dont have it as complete are given less weight to the aggregate score as well so most the bots also lose weight there On that note if this stuff bothers you, you can painstakingly go search through the users on the site and report the bot accounts, they are always really obvious rating only with 1's or 10s only to go inactive immediately, before it finishes airing you can try and track down the ones that have a ongoing show listed as completed to which there is about 50 some right now, those are most likley bot votes to either inflate or decrease the score. |
Immahnoob said: Jizzy, I know you have no idea how to argue for shit, tokiyashiro said: Jizzy as you would call yourself because youre a dick The most butthurt award goes to you And clearly you havent watched that many shows thats why you cant determine if a show is unique or not Or maybe you're just a child who likes common stuffs where hero saves the day and guys gets all the girls. Sad taste you have there kid you came up to me in the first place making you look more like a kid who got slapped without me even knowing it and start crying about it to me |
| Yet the score keeps increasing |
 |
LiedElfen said: it had to of been a glitch cause it went from like 8.32 to 8.50 in literal minutes. Fucking weird but i like to imagine it was some dude bot voting it to hell or some shit so there was another guy with too much time on his handsThe score just went from 8.32 to 8.50 with one episode, i don't know how that worked but holy shit. |
Immahnoob said: Jizzy, I know you have no idea how to argue for shit, tokiyashiro said: Jizzy as you would call yourself because youre a dick The most butthurt award goes to you And clearly you havent watched that many shows thats why you cant determine if a show is unique or not Or maybe you're just a child who likes common stuffs where hero saves the day and guys gets all the girls. Sad taste you have there kid you came up to me in the first place making you look more like a kid who got slapped without me even knowing it and start crying about it to me |
LiedElfen said: The score just went from 8.32 to 8.50 with one episode, i don't know how that worked but holy shit. This has happened before. It usually ends up fixing itself back to what it normally was within a few hours. No idea how mal fucks up like that randomly. |
JizzyHitler said: When a series concludes airing if someone has the series either watching at less than a 3rd of the series or has it dropped at less than a 3rd then those scores are not counted towards the overall series scorehence why alot of series especially with weaker starts tend to have a massive boost in score when they end, so a decent amount of these bot scores will be negligible (this happens alot for just about all popular shows especially when they start reaching a top ranking). What happens if a bot rates a show a 1 and marks it as completed? |
ycleped said: they tend to get deleted and caught cause they are among only a small few, and after a show finishes the throwaway accounts are usually so minuscule compared to the actual ratings its negligible. Plus these types of 1 bombs usually dont continue if the show finishes unless you are a gintama entry or in the top 10.....or a gintama entry in the top 10JizzyHitler said: When a series concludes airing if someone has the series either watching at less than a 3rd of the series or has it dropped at less than a 3rd then those scores are not counted towards the overall series scorehence why alot of series especially with weaker starts tend to have a massive boost in score when they end, so a decent amount of these bot scores will be negligible (this happens alot for just about all popular shows especially when they start reaching a top ranking). What happens if a bot rates a show a 1 and marks it as completed? |
Immahnoob said: Jizzy, I know you have no idea how to argue for shit, tokiyashiro said: Jizzy as you would call yourself because youre a dick The most butthurt award goes to you And clearly you havent watched that many shows thats why you cant determine if a show is unique or not Or maybe you're just a child who likes common stuffs where hero saves the day and guys gets all the girls. Sad taste you have there kid you came up to me in the first place making you look more like a kid who got slapped without me even knowing it and start crying about it to me |
| Score inflation happened to Bunny Girl Senpai too last Fall. |
 |
| Score boost has been fixed, apparently. |
| It's a known issue. Kaguya, Shield Hero and Neverland all have a varying amount of fake 1s on their scores, and the mods don't seem to have a way to catch them since the guy puts in extra effort to hide himself from the public rating stats table so the public can't report him in the illegitimate accounts thread. Assuming 0.4% out of those 1.6% 1's are legit (meaning 1.2% are fake), Kaguya's score should be : 8.33 (current score) = 0.988*x (real score) +0.012*1 (fake 1s) x=8.42 The effect will be a lot lower once the show ends and everyone rates it, but the botter will still be bringing it down by 0.04-0.05. It used to be a 0.2 difference early in the season, so I guess we can be happy with the fact that the guy wasted probably days of his life only for a few hundredths of the score. |
 |
BaronBrixius said: It's a known issue. Kaguya, Shield Hero and Neverland all have a varying amount of fake 1s on their scores, and the mods don't seem to have a way to catch them since the guy puts in extra effort to hide himself from the public rating stats table so the public can't report him in the illegitimate accounts thread. Assuming 0.4% out of those 1.6% 1's are legit (meaning 1.2% are fake), Kaguya's score should be : 8.33 (current score) = 0.988*x (real score) +0.012*1 (fake 1s) x=8.42 The effect will be a lot lower once the show ends and everyone rates it, but the botter will still be bringing it down by 0.04-0.05. It used to be a 0.2 difference early in the season, so I guess we can happy with the fact that the guy wasted probably days of his life only for a few hundredths of the score. Definitely frustrating but it is what it is...why someone would even waste their time doing this just boggles my mind |
ycleped said: I'm not normally one to start a conspiracy about the legitimacy of MAL scores; however, based on the average score distribution plot of shows on MAL, there is no way this can be correct. The amount of 1 ratings is simply not feasible. It's not just compared to the way the lower scores are gradually decreasing - just scroll through the pages of ratings and literally nobody is giving it a 1. This had to have been done all at once. Just take a look at the Neverland score distribution: Even though there is definitely a noticeable increase in people that gave it a 1 vs a 2 (a 5.31:1 ratio), it's not nearly as dramatic as for Kaguya (a 9.25:1 ratio). Thoughts? Also, is there an admin here who would be able to investigate this? I've looked the same way you did and I'm convinced someone managed to add some 1 without accounts OR with a mean to hide those accounts from the updating members list. Nevertheless a bit more than one month ago, it seems the admin took measures against this one guy and it seems to me that's the reason the score went up again despite the fact that number of 1 doubled in the same time :) I'm not technically versed enough to know how to cheat with such a scale but I can tell you that even some best girl polls had distorted results and to me, it looked like the guy involved probably distorted the result WITHOUT fake accounts. I don't know how it's possible to do something like that though but if someone is knowledgeable in web-development, maybe he can find a way. |
| There are some 'more intelligent' fake accounts appearing lately. They randomly rate many anime between 2-10 as diversion and will only rate stuff that they want to downvote as 1. You can only predict that they are alts by their last online date. They are only online till a few days after creating the account. It honestly surprises me how these people spend days creating these downvote/upvote accounts just to influence the rating. Like they don't have anything else to do? There is a good way to easily prevent the influence of these accounts. If MAL includes an algorithm to ignore the ratings of users who have never been online for say, 6 months after creating their account, many of these alts will be rendered ineffective. It won't hurt real users as their list is intact and their score is considered for average once they come back online. But for alts, it is impossible to maintain these accounts for so long. |
Chrome_Falcon said: There is a good way to easily prevent the influence of these accounts. If MAL includes an algorithm to ignore the ratings of users who have never been online for say, 6 months after creating their account, many of these alts will be rendered ineffective. It won't hurt real users as their list is intact and their score is considered for average once they come back online. But for alts, it is impossible to maintain these accounts for so long. There are many more possibilities for the kind of algorithms that could "defeat" the fake accounts but it's a real job (data base analyst) and you need to be relatively clever for that too. Furthermore, since I'm a bit paranoid, I thought that many sites who rely on advertisement to survive didn't really fight against that since it's easier to sell ads when you have "officially" more accounts/users than you really have. Let's say it's certainly not always a priority to get rid of the fake accounts. In the contrary, they help to earn money to a certain degree. |
| Now that the show is listed as finished the 1 votes really dont affect it all that much |
Immahnoob said: Jizzy, I know you have no idea how to argue for shit, tokiyashiro said: Jizzy as you would call yourself because youre a dick The most butthurt award goes to you And clearly you havent watched that many shows thats why you cant determine if a show is unique or not Or maybe you're just a child who likes common stuffs where hero saves the day and guys gets all the girls. Sad taste you have there kid you came up to me in the first place making you look more like a kid who got slapped without me even knowing it and start crying about it to me |
More topics from this board
Poll: » Kaguya-sama wa Kokurasetai: Tensai-tachi no Renai Zunousen Episode 3 Discussion ( 1 2 3 4 5 ... Last Page )Stark700 - Jan 26, 2019 |
296 |
by Intter
»»
Apr 14, 11:26 AM |
|
» Misattribution of arousalegosumjohn - Apr 7 |
11 |
by Aruhaya
»»
Apr 14, 11:13 AM |
|
» I laugh as hard in this series as gintama š¤£Jannie7 - Mar 31 |
38 |
by xDemi
»»
Apr 1, 11:38 AM |
|
Poll: » Kaguya-sama wa Kokurasetai: Tensai-tachi no Renai Zunousen Episode 6 Discussion ( 1 2 3 4 5 )Stark700 - Feb 16, 2019 |
246 |
by DevilsMayMew
»»
Apr 1, 1:32 AM |
|
Poll: » Kaguya-sama wa Kokurasetai: Tensai-tachi no Renai Zunousen Episode 5 Discussion ( 1 2 3 4 5 ... Last Page )Stark700 - Feb 9, 2019 |
260 |
by DevilsMayMew
»»
Apr 1, 1:04 AM |