Forum Settings
Forums

Top Anime: Adjustment to the Weighted Ranking System [Waived]

New
Feb 1, 2016 6:46 AM
#1
Offline
Mar 2014
2421
My concerns about MyAnimeList's ranking system were outlined in a brainstorming thread last summer; to paraphrase, I made the case of giving all entries an equal chance at the top chart. Newer and more popular anime have an undeniable advantage over long-forgotten titles, so it is therefore hard to argue that the current system accurately represents the opinions of all users.

Max said:
Maybe it would be best to weigh a user's votes depending on their mean score? So if a user's mean score was 7 and they rated something 7, it could be counted as a 5 or something. Kind of on a curve.

On July 11th, I posted this. It was initially a throwaway idea in a sea of many, but it began to resonate with me over time. Maybe it was a plausible hypothesis, though I did not have any means of conducting the experiment. Being unable to exemplify its merits frustrated me to no end.

I can't seem to recall what happened next. Either upon request or at her own will, a friend of mine began working on the system in her spare time. As it's nothing more than a hobby, it took her a few months to get to a reasonably presentable stage. You can check it out here, in all of its unfinished glory:

http://bitpuritans.eu/mal/anime.php
(This may take a moment to load.)


Most of the current top anime hold their general ground, like LotGH, Gintama, FMAB, and HxH 2011, yet less popular titles have managed to make their way into the mix. It is very interesting to see, and much more impartial in function. Do you think MyAnimeList would benefit from displaying its top lists in this manner?


Mod Edit: Modified title for clarity and/or easier searching.
_Ghost_Mar 1, 2016 2:20 PM
This topic has been locked and is no longer available for discussion.
Pages (5) [1] 2 3 » ... Last »
Feb 1, 2016 7:19 AM
#2
Offline
Jun 2009
21
heyo~! i'm the gal running this project; i've been referred on to give a quick brief on how the site actually works:

• the HELL PATH and HEAVEN PATH rate the lowest-rated and highest-rated anime respectively; shows with more dislikes than likes wind up in the left list, and more likes than dislikes on the right. i'm often fond of a lot of the shows on the HELL PATH so having it be at the *top* of a list for a change is refreshing
• after that, the ratings *only* count the average of scores normalised to below 5 (hell path) and above 5 (heaven path); this is because i've noticed that liking or disliking an anime tends to be quite binary or knee-jerk at the extremes~ rating this way ensures that radically good shows (or bad shows with a few ironic likes) don't get their ratings muddled by outliers. some notable examples of how it improves polarising, but critically acclaimed shows:

shoujo kakumei utena (284th -> 80th)
tenshi no tamago (1422nd -> 157th)
mind game (668th -> 121st)
yojouhan shinwa taikei (49th -> 8th)

as well as popular, but otherwise over-hyped ones:

shingeki no kyojin (56th -> 196th)

• the coloured bars show a sampling of the average spread of ratings, coloured by normalised decile; black shows a voter rated an anime in the bottom 10% of all their anime (regardless of score), going upwards to brown, red, orange, yellow, green, blue, indigo, violet and bright violet (for top 10%).
• numbers directly to the left and right of the bars represent how many voters disliked (score under 5) or liked the show (score over 5). the coloured (again, by decile) bold numbers on the rightmost of each list show their average "rating" on that list out of 1000
• i haven't fixed the place numbers for the hell path yet, but i'm getting to it!

any thoughts~?
Feb 1, 2016 7:25 AM
#3
Offline
Jul 2015
1009
I like this. MODS, DO IT! NOW'S YOUR CHANCE TO GIVE THIS SITE CREDIBILITY.
Feb 1, 2016 10:52 PM
#4

Offline
Jun 2014
24
I'd actually like to see
*braces for pitchfork wielding masses*
All ratings from inactive accounts longer than a year become non-contributions for anime overall ranks. Allow the currently active members of the site to determine where anime stand, rather than certain anime having a ridiculous number of ratings by member, but 50-60% of those dont ever visit the site anymore.
It would be fun to see how anime balances out after that, especially with new anime to stand comparison against the long standing titles in the top 20. The amount of contributing ratings would hold less disparity between anime, giving them more freedom to fluctuate based upon newer active users.
Well, the mob is at the door, thanks for your time kind internet frens ^^
I suppose my last thoughts as the rattling of garden tools and yells of an angry mob draw near is this: Let the users of the list determine the list. Otherwise it becomes a stagnant resource and sooner or later serves no purpose other than to stand there and shout "You kids these days dont know how good anime used to be"
Feb 2, 2016 3:42 AM
#5

Offline
Dec 2011
8943
s1rnight said:

• numbers directly to the left and right of the bars represent how many voters disliked (score under 5) or liked the show (score over 5).


Why are these numbers so low? There are things on there that have 50 times as many ratings on MAL as the sum of those figures.

And lol at all the links leading to an illegal streaming site... And a crappy one at that...
There is no such thing as shit taste. Only idiots who think everyone should have the same taste as they do.
Feb 2, 2016 3:49 AM
#6
Offline
Jun 2009
21
ah, i haven't finished processing all the users yet~ there's quite a lot and it lags the server while it's running so i'm hoping to get some thoughts on the system before i get back to it

canaduh: i like that idea! i'll see if i can find a simple way to only count ratings from users in the last year or so
Feb 2, 2016 6:47 AM
#7

Offline
May 2015
3235
This is an awesome idea. I'm so sick of new shows getting in the top 100 before they've even gone past two episodes...do it!

Would it be possible to take airing anime out of consideration for rankings until they've been fully aired, and maybe to weight scores based on how many people have rated said anime? It always bugs me when an airing show with only 35-50,000 people watching beats out shows that have been scored by more than 100,000 people, just because those few who were watching as it aired scored high based on first reactions.
KaoruMatsuokaFeb 2, 2016 6:50 AM
Feb 2, 2016 10:00 AM
#8

Offline
Oct 2010
470
Very interesting, that sounds like a great idea too. It would be interesting and maybe a more exact way to use the rating system.

Tumblr |  Anime List | 🌷🌼

Feb 4, 2016 8:31 AM
#9
Offline
Jul 2015
881
This is a great idea, but I think that if they do that, they shouldn't eliminate the actual system.
 
Feb 4, 2016 9:44 AM
Offline
Mar 2014
2421
Shinjisus said:
This is a great idea, but I think that if they do that, they shouldn't eliminate the actual system.

What I'd like is for this to be made the default alongside a checkbox that shows the page unweighted.
vegetablespiritFeb 4, 2016 9:50 AM
Feb 5, 2016 12:11 AM

Offline
Nov 2011
9206
Max said:
Shinjisus said:
This is a great idea, but I think that if they do that, they shouldn't eliminate the actual system.

What I'd like is for this to be made the default alongside a checkbox that shows the page unweighted.

That would be perfect.

Full support.
Feb 5, 2016 3:31 AM

Offline
May 2014
488
Like the idea. I can't "prove" that this system is better and fairer, but deep down it makes sense to me. Plus the results are more than encouraging. And it's not like good popular show's ratings became bad. Good job.

parazeFeb 5, 2016 3:35 AM
Feb 5, 2016 3:52 PM

Offline
Feb 2010
34597
That's just perfect. I'm loving that idea. Though I'm already happy to just be able to have a look at the improved rankings on the site you linked.
I probably regret this post by now.
Feb 6, 2016 2:37 AM

Offline
Apr 2009
5711
Interesting idea, it would def. counter people who rate while priorizing only one side of the spectrum and I am all for giving older anime a chance.

Canaduh said:
I'd actually like to see
*braces for pitchfork wielding masses*
All ratings from inactive accounts longer than a year become non-contributions for anime overall ranks. Allow the currently active members of the site to determine where anime stand, rather than certain anime having a ridiculous number of ratings by member, but 50-60% of those dont ever visit the site anymore.
It would be fun to see how anime balances out after that, especially with new anime to stand comparison against the long standing titles in the top 20. The amount of contributing ratings would hold less disparity between anime, giving them more freedom to fluctuate based upon newer active users.
Well, the mob is at the door, thanks for your time kind internet frens ^^
I suppose my last thoughts as the rattling of garden tools and yells of an angry mob draw near is this: Let the users of the list determine the list. Otherwise it becomes a stagnant resource and sooner or later serves no purpose other than to stand there and shout "You kids these days dont know how good anime used to be"


I don't see the problem with counting "inactive users". As Ping Pong, HxH, ERASED and many other series on the Heaven path show, good series will always have a chance. They just won't be in the Top 100 anymore by virtue of being slightly above average in a crap season alone or when they are so called "entry series" that attract tons of high ratings because of new anime fans who vastly overrate their new experience in comparision to other shows. Disregarding inactive users is a massive neglect of past reception, which is something that will once again massively screw ratings over and render the point of this system moot as - sadly - a lot of new users are not willing to watch older shows, partially resulting in the problems we have now.

Overall this system is pretty good. A question about people with rather low average scores, though: Let's say the average score of someone is 4.something, with tons of titles being in the 3-5 region, but not a lot of 10's and 1's, to indicate that the quality of the 1's is way below this user's standard 4.something rating, yet with equalized votes, it would count as a 2 or 3. Likewise, with such an average (or an even lower one), 8s,9s and 10s would be virtually the same when the system weights the score for this particular individual. I assume this is intended?

Another point of thought is an additional weight being put on entries belonging to the same franchise, to get rid of the sequel favouring trend, but that might be a very flawed addition. I am open to discussions about it, though.

I wonder: How would a weighted manga list look like? Not counting LN (which should really be their own category anyway), I find it a lot more accurate than the anime top list. I'd be interesting to see the effect here as well.
NidhoeggrFeb 6, 2016 5:27 AM
Steel Ball Run anime when?
Feb 6, 2016 3:44 AM

Offline
Jan 2010
104
s1rnight said:
• the coloured bars show a sampling of the average spread of ratings, coloured by normalised decile; black shows a voter rated an anime in the bottom 10% of all their anime (regardless of score), going upwards to brown, red, orange, yellow, green, blue, indigo, violet and bright violet (for top 10%).

Do I get it right, that score's weight given by a user is based on ONE's list of anime/manga?
If it is so, then it's downright stupid. Because normal person won't watch the show that he despises. So let's take 2 people: one watches only good quality anime, and gives them scores from 6 to 10, oter watches everything and gives a wide-spread ratings of 1-10. So, taking into account their bottom 10% a show that they both rated 7 receives weight of first person equal to show that second person rated 2?
If it's not based by the entries added only in person's list - then could you explain a little more detailed as I didn't catch how you calculate things then.

s1rnight said:
• numbers directly to the left and right of the bars represent how many voters disliked (score under 5) or liked the show (score over 5). the coloured (again, by decile) bold numbers on the rightmost of each list show their average "rating" on that list out of 1000

If you divide those in 2 fields (like/dislike) then you should add score 5 to dislike sum, as it's a negative score.
Feb 6, 2016 7:50 AM

Offline
Nov 2010
656
I agree with Whiterock, a long time ago I also thought that the score is wrong, but 1 fact changed it, if 1 person prefers to watch only good shows, their score wouldn't go lower than 5.
And since most people prefer to watch only good shows, I don't think their score should have different meanings that other people. Because if someone never rate anything below 5, it means that score of 8 would probably decrease the rating of anime, though he meant to give it an 8.
Unless I also didn't understand how it's been weighted.

And one more thing:
"as well as popular, but otherwise over-hyped ones:

shingeki no kyojin (56th -> 196th)"

There are many shows which are overrated, but you can't say that shingeki no kyojin is bad and have to go down in the top list, you maybe didn't like it, but most of people did, so there's nothing wrong about it being at the top. Moreover, you can't argue with people's taste, some would say it's good and some would not, but it's an personal opinion and you can't decide for them if it's good for those people or not, which basically this system does.

So overall I don't really want to accept this system. Sorry. I also tend to watch good shows only, which should result in high ratings, and I don't want a system which would think that I'm scoring my anime badly. Yes, there are people who only give 8-10 even if the shows sucks, but they are few and not worth changing all the system.
Feb 6, 2016 8:00 AM

Offline
Dec 2012
24356
I really don't agree with this system at all. It pushes the mentality of elitism and pushes down popular shows while elevating cult ones.

If you look at the scores, it only makes sense that most people don't use scores below 5 that often. Because why would you keep watching shows and completing them with a low score on regular basis?

For most people or myself at least. There is a big difference between 7, 8, 9, 10. But there is practically no difference between 1, 2, 3 and 4.'

and like @FunnyGames said, it seems like it's forcing one idea of what is correct scoring and what isn't. Who has the authority to just say that SnK is an overhyped series that should be lower in the ranking? why should people just accept what is obviously a subjective notion.
tsudecimoFeb 6, 2016 8:04 AM
Feb 6, 2016 9:05 AM
Offline
Mar 2014
2421
@FunnyGames @tsudecimo The fact of the matter is that if you are not using the full scale, your scores should be adjusted to fit within it. This changes the value of your input in the process, but you are still having a fair say on the matter.

You claim there's no correct scoring system, though there technically is one provided by the website. This suggestion allows people to use their own scoring systems whilst adjusting them to the site's standard. A good comparison would be editing guidelines; while there are many ways to say or spell a word, a news source would only accept one pseudo-correct answer for the sake of consistency.

As is common on the Suggestions board, rebuttals to improvements are often made by those at fault—in this case, the overwhelming majority of people that exclusively use scores of 6+. You are protecting the status quo because it directly benefits you, a blind eye paid to the many problems it causes.

To be concise: This proposition punishes everyone, whereas the current system just punishes people with a mean score below 6. It is objectively fairer for both sides to be reprimanded.

Still, I would like to make the view optional, so I don't think it would affect you.
vegetablespiritFeb 6, 2016 9:42 AM
Feb 6, 2016 10:24 AM

Offline
Nov 2010
656
@Max Once again, I can't say anything as I don't know the math behind it.

But lets try an experiment:
- Assume I have a list of 10 anime - Death Note - 10, School days - 9, Durarara - 8, Akira - 7, C3 - 6, C - 5, Colorful - 4, blood-c - 3, nami - 2, some anime - 1.
- And another person has this list - SnK - 10, Death Note - 10, DTB - 8, Angel Beats - 9, Bleach - 10, School Days - 9, Ranma 1/2 - 10, One Piece - 10, Naruto - 9, Naruto: Shippuden - 10.

Both of us have 10 anime in our lists. I managed to watch 10 different anime with different scores, as I enjoyed those less, but the other person made it to watch the best of them.
You said that in the end it would be fair?
So if I made it to make an average score of my list to 5.5 (each anime has unique score number), the other person made an average of 9.5, which is bad and wrong average, the system will reduce the score of other anime, basically saying that DTB - 8 - would become 5 or 6? But it was good, and Angel Beats -9 - would become 6 or 7? But he really enjoyed it?
It's not his fault that he watched the best of them... So why does he have to suffer this loss when he knows that he probably wouldn't enjoy C, Colorful, nami or such anime and won't start watching them.
So it's already wrong. Remember: most people tend to watch the best of anime.

And you also can't expect that people would give score equally between 1-10, as 1 is very low and it's very rare for an anime to be so bad, every anime has something special in it. Also, you can't say that there's an anime which is exactly an average anime, for each person it will look better/worse. It's individual opinion, if I don't like it, it doesn't mean you won't like it.

I think that this system is forcing everyone to score something they wanted to score to something else, which is not a good idea. If a person don't know how to score, it's their problem, not everyone.


And you can use this system in external website, like anime.plus with theirs achievements and favorites systems.

Sorry for being so harsh, but I'm trying to change your opinion like I did some time ago. You can't use math to decide people likes and taste, as it's not function or fact.
Feb 6, 2016 12:16 PM

Offline
Jan 2010
104
@Max
Yes we know that there are people who score things 6-10 where they consider 6,7,8 as bad anime.

But that doesn't mean that we should hurt people who doesn't want to waste their time to watch truly bad anime, as well as those who watched very little of bad anime.

There are not that much of people like me who would keep watching horrible anime and especially sequels to use the whole scoring system. It made me puke, but I finished watching Uta no☆Prince-sama♪ Maji Love Revolutions (3rd season) which I gave score 1, and previous season 3, and I still gonna watch season 4 when it comes out, because I am crazy completionist with enough spare time to waste my life :)

If there is anything that needs a fix - then that's only the mentality of people who can't admit they watched an average or bad show and give it a score 4-6 it deserves, and not 8.
Feb 6, 2016 1:31 PM

Offline
Jun 2014
24
I guess if you think something deserves 4-6, once again that is your right of opinion, one we all have, and there's no accounting for taste. If someone gave something an 8, maybe they just like the series?
I guess how good an anime is will always be debatable between its watchers, but if someone rates something higher than you, you can't claim it doesn't deserve it :o

If you want to purposely watch things you don't enjoy, that's your prerogative, not everyone WANTS to watch an anime that deserves a 1. -4 just to feel like they are using the lower spectrum, and personally I think that topic is in the wrong forum.

My original point was not to not use only information for the last year, but just to eliminate data from accounts that have not been on in more than a year, as the list is of no use to them. as I stated up there, let the users of the list determine the list :>

If Boku Dake becomes rank 1 with only 200k votes from users who are watching it atm, let that be what happens. The list will become boring and stagnant if there is no room for fluctuation by the people who are only just now getting into anime, and perhaps if ever there was to be introduced a system to hold a record on each animes page to tell where its highest standing of all time was, the new viewers could go back and find animes that had been higher on the list before and go back and watch them.

Change is a good thing fren :>
Feb 6, 2016 1:58 PM

Offline
Apr 2009
5711
The problems is that when you only ever use only 2 or3 different scores at max you are having an objectively less differenciated score than someone who uses the entire scale and your numbers obviously mean less than the numbers of someone with more refined rating. If your entire opinion just consists of "everything is awesome" or "it's not so great, but still good", your opinion is completely worthless when you try to dig deeper - and this is exactly what the 8-10/10 raters are. It's completely superficial, mostly arbitrary, disregards more nuanced lists or entry points for discussion and is ultimately just forced positivity for whatever reason (A possible lack of interest in critical discussions, the need to validate everything for no reason besides existing, fan bias, etc.).

Once again: The weighted chart shows that the "pushes the mentality of elitism and pushes down popular shows while elevating cult ones" argument is complete bs: The newest Gintama entries, HxH 2011, the Monogatari series, ERASED, Shigatsu, etc. They are ALL there and those are pretty new and popular entries that are enjoyed by many more casual aniem watchers as well! So if your series is good both user groups with a restricted and a full rating system will get their votes considered and equally weighted in this new system, whereas in the old one people will essentially be punished by having their impact on their ratings reduced for not rating everything under the sun 8-10/10.
On the other hand, shows like Ashita no Joe (which is basically one of most important anime entries of all time) actually get the same "acclaim" they get in the scholary world and in Japanese anime history. So really everybody profits except for the people who want every show they watch to have a publicly displayed rating above 8 because they themselve rate everything at least with an 8. And if those people really need to resort to such drastic measures to confirm their taste or rating, then THEY have a problem and should learn to either better articulate their opinion or accept that one can personally like stuff even though it's actually pretty bad. You know, like most normal people do with movies like Sharknado.

Canaduh said:

My original point was not to not use only information for the last year, but just to eliminate data from accounts that have not been on in more than a year, as the list is of no use to them. as I stated up there, let the users of the list determine the list :>

If Boku Dake becomes rank 1 with only 200k votes from users who are watching it atm, let that be what happens. The list will become boring and stagnant if there is no room for fluctuation by the people who are only just now getting into anime, and perhaps if ever there was to be introduced a system to hold a record on each animes page to tell where its highest standing of all time was, the new viewers could go back and find animes that had been higher on the list before and go back and watch them.

Change is a good thing fren :>


By elimating the votes from older people who might get inactive you essentially massively favour new shows over older shows which is the current problem in the system along sequel bias, rating inflation and incompability between the two main rating systems most people use. Let's be honest: What happens when all the people who watched TTGL are offline in 10 years because they moved on? TTGL is one of the better and def. one of the most influential mecha shows in the 2000s. Their votes would get discounted.For what reason? On what basis? Because it's old? That's hardly an argument. By the same reasoning, why should we consider new shows to be worthier of being on the top than older ones? Because they are new? It's without reason in both cases, true reception of art is crafted from gathering the responses of all parts of the community, it's crafted from critical discussion on varying levels that try to find consensus. Even the past reception shapes our reception of today. This is why history classes exist and why we as society are able to move forward while facing challenges.

And as I said before: There is a lot of room for fluctuation seeing as a good chunk of the entreis in the new system are post-2010. And unlike in the old system where this is also the case, this one would have less of a rating inflation and combine the two rating systems that users on the site normally use. This system promotes change while also preserving a sense of past reception.
And honestly: When you are an actual fan and really interested in the hobby, you should want to know about game changers in the history of anime. You should be at least partially interested in the devlopment and history of the medium to understand why certain things are what they are, what molded them into their current form and what their pros and cons are. This is how you form an educated opinion, it enables you to actually have better discussions about the things you should have a lasting interest in. This is what being a fan is all about imho.
If instead all you care about is being validated by having MAL show that every seasonal you watch and like is rated a "very good" or higher, then this just shows your immaturity as you feel the need to constantly confirm your own subjective opinion via flawed systems no matter if it is logical or not. And why should we - as a site with a concern for facilitating critical opinions tied to a fair rating system- nurture this childish behaviour? It's not like your 8s or 9s mean less to you (and really, in your list this is all that should matter), it just puts your objectively less differenciated opinion on par with people who use other systems. Max made some very good points and I agree with most of them.
NidhoeggrFeb 7, 2016 12:57 AM
Steel Ball Run anime when?
Feb 6, 2016 2:48 PM

Offline
May 2013
613
This seems like a pretty good way to fix the rating inflation and make the top list actually useful again. I hope this gets pushed.


Feb 6, 2016 3:00 PM

Offline
Jun 2012
378
There's no reason why an update like this shouldn't go live; everyone could benefit from a more balanced top anime list, especially newcomers to the medium/community
Feb 6, 2016 3:06 PM

Offline
Apr 2012
2241
I fully support this decision and have absolutely no issue with it being implemented.
"The more inexperienced you are, the more you want to show off."
- Oreki, Hyouka
Feb 6, 2016 3:23 PM

Offline
Apr 2009
5711
The only thing this system might have problems with are the people who either only rate 1 or 10. Not many of those, but I suspect this will become a popular trend for people who try to outsmart the system. And then there are troll accounts or votebomb accounts, made just to downvote or hype a few certain series.


How would this system counter these cases?
Steel Ball Run anime when?
Feb 6, 2016 7:43 PM

Offline
Sep 2013
22818
Rating on mean score is pointless since about half the 10/10 rating system will be useless.

And a 5/5 system is not good enough, a 3/5 is average and 4/5 is good? almost every single anime will have the same score and rankings will be pointless.
Feb 6, 2016 9:17 PM

Offline
Jun 2014
24
Firstly I would ask that you stop quoting me please, I don't much care for arguing points back on forth in a forum unrelated.

Ok, so last post on this, I like the idea of having a toggle on the main list to show this system, I would probably be against it being fully replaced without changes to the outdated ratings of unused accounts and I think maybe a tag of last position it was on the list, plus highest ever placement would be nice as years roll by, so you could see the direction any particular anime was heading.
Feb 6, 2016 10:59 PM

Offline
Nov 2011
9206
Canaduh said:
Overall this system is pretty good. A question about people with rather low average scores, though: Let's say the average score of someone is 4.something, with tons of titles being in the 3-5 region, but not a lot of 10's and 1's, to indicate that the quality of the 1's is way below this user's standard 4.something rating, yet with equalized votes, it would count as a 2 or 3. Likewise, with such an average (or an even lower one), 8s,9s and 10s would be virtually the same when the system weights the score for this particular individual. I assume this is intended?

This is intended afaik. The system is meant to standardize ratings when they are calculated for rankings, so extremes in either direction would be balanced. In the same way that 1-3 would all mean "trash" on a 6-10 rater's list, 8-10 all mean "treasure" on a 1-5 rater's list; although the difference may be significant to the individual, the difference is not significant globally.



Nidhoeggr said:
The only thing this system might have problems with are the people who either only rate 1 or 10. Not many of those, but I suspect this will become a popular trend for people who try to outsmart the system. And then there are troll accounts or votebomb accounts, made just to downvote or hype a few certain series.


How would this system counter these cases?

You're right that this system can't counter that well, but this will virtually never happen outside of intentional abuse (such as the troll/votebomb accounts you mentioned). Not only is this an unpopular method of rating, but I also doubt that many users would who do would have an equal or nearly-equal number of 10's and 1's on their list. It is likely to skew mostly to one or the other, in which case the system would handle it appropriately.
Feb 7, 2016 12:52 AM

Offline
Apr 2009
5711
Canaduh said:
Firstly I would ask that you stop quoting me please, I don't much care for arguing points back on forth in a forum unrelated.


And this is the problem with many people who don't use the full rating scale. Why should we validate this kind of superficiality and indiffirenciated attitude by basing the entire ranking about them when we can just as well please both systems while still giving them a say (albeit with slightly reduced influence) and get rid of some longstanding problems?

TripleSRank said:

You're right that this system can't counter that well, but this will virtually never happen outside of intentional abuse (such as the troll/votebomb accounts you mentioned). Not only is this an unpopular method of rating, but I also doubt that many users would who do would have an equal or nearly-equal number of 10's and 1's on their list. It is likely to skew mostly to one or the other, in which case the system would handle it appropriately.


Fair enough. It's worth trying out anyway and once the system is implemented one can gradually add or adjust variables.
NidhoeggrFeb 7, 2016 12:55 AM
Steel Ball Run anime when?
Feb 7, 2016 1:10 AM

Online
Jan 2009
92287
i support this, i hope MAL developers will take a look at this seriously
Feb 7, 2016 3:12 AM

Offline
Jul 2008
421
I support this as well. People will never properly use the entire scale (including myself to some degree) and this seems like a pretty fair solution.


Feb 7, 2016 8:25 AM

Offline
Apr 2009
5711
I also don't think that "average" is in any way a negative thing. It's just that: Average. Not bad, but also not particulary good. you can watch it, it can be a decent timewaster, etc. Why people find the thought of some of their favourite anime being only average so apalling is beyond me. I like a lot of stuff that some people will consider average or bad and I don't feel bad about it just because most people dislike it... But as galimx said: If you think it's bad, but still rate it "good" it's just plain dishonest and you abuse the system by rating in a way it wasn't intended to be used. A correction via weighting sounds like the most normal thing to do in such a situation.
Steel Ball Run anime when?
Feb 8, 2016 2:15 AM

Offline
Jan 2010
104
Just to note why I said that score 5 is negative - because MAL uses 1-10 ranking system. mean score is 5.5, so everything below 5.5 is negative, everything above is positive, aka 1-5 and 6-10.
5 would have been neutral if we had score "0" or used 1-9 scoring system. Simple math.

And as in personal terms - I define average for shows that don't have strong bad points, yet doesn't have special trait to it. "Assembly line" type of anime. With occasional exception to shows where bad traits and good traits pull down the show in both directions. In those cases I leave the score based on personal feelings.
Feb 8, 2016 6:07 AM
Offline
Mar 2014
2421
Whiterock said:
Just to note why I said that score 5 is negative - because MAL uses 1-10 ranking system. mean score is 5.5, so everything below 5.5 is negative, everything above is positive, aka 1-5 and 6-10.
5 would have been neutral if we had score "0" or used 1-9 scoring system. Simple math.

Sure, but this is irrelevant to the current implementation. If MAL says 5 is 'average', then 5 is average. Math doesn't matter.
Feb 8, 2016 9:07 AM

Offline
Jan 2015
116
This doesn't make any sense to me, if someone only watches things they enjoy, forcing an average of 5 onto their list would completely mess up the ratings. I personally have an average of 7.6, and if a 5 average was forced onto my list I would have to add a bunch of random entries and give them 1's to balance out the average and make my ratings accurate again.

A 5 average should only exist for someone who chooses shows completely at random and never drops anything. If you have any selection process whatsoever, you will have a higher than 5 average.
Feb 8, 2016 9:18 AM

Offline
Jan 2015
116
Max said:
Whiterock said:
Just to note why I said that score 5 is negative - because MAL uses 1-10 ranking system. mean score is 5.5, so everything below 5.5 is negative, everything above is positive, aka 1-5 and 6-10.
5 would have been neutral if we had score "0" or used 1-9 scoring system. Simple math.

Sure, but this is irrelevant to the current implementation. If MAL says 5 is 'average', then 5 is average. Math doesn't matter.


So you use 5 - average to mean your own personal average? I don't think this is a useful way to use the scale, because it entirely depends on what you have watched. If you have seen only 1 anime, then by this logic, it must be rated 5 regardless of what you thought about it, right? This is an extreme case but one person's list will always be a small subset of anime, and their ratings will depend on what they have watched.
Feb 8, 2016 9:24 AM

Offline
Nov 2014
406
Phann said:

So you use 5 - average to mean your own personal average? I don't think this is a useful way to use the scale, because it entirely depends on what you have watched. If you have seen only 1 anime, then by this logic, it must be rated 5 regardless of what you thought about it, right? This is an extreme case but one person's list will always be a small subset of anime, and their ratings will depend on what they have watched.
No, on the MAL scoreboard 5 is called "average", that is obviously bullshit. For all your information the REAL average on MAL is 7,23 atm. Fun fact that there's hardly anyone who rate anime like the scoreboard say, but then there are ppl like him who do it and try to "force" others to do it too.
Feb 8, 2016 9:24 AM

Offline
Nov 2010
656
@Max, can you explain the math and how I should read your website?

I understood the first column which represent at what position the anime is, second is name, but other?

I would like to give it a try, I just want to see how it will look compared to today's top anime.

And one more question, will the list show the score like today? Like Steins;gate 9.17? (I believe it would have a different number than 9.17)
Feb 8, 2016 9:30 AM

Offline
Nov 2014
406
FunnyGames said:
can you explain the math

If I understand what he want to do is when you rate the anime:

Your average*(the score you give/10)

This means if your average is 7,69 like mine and you give an anime score 10, then your actual rate will be 7,69
if you give 5 then it will be 3,845
Feb 8, 2016 9:33 AM

Offline
Apr 2014
4399
why?

a 7/10 for me is the same 7/10 for u

i just find most anime 7/10

doesnt mean i think 7/10 is a 5/10
Feb 8, 2016 9:34 AM

Offline
Nov 2010
656
Delenai said:
FunnyGames said:
can you explain the math

If I understand what he want to do is when you rate the anime:

Your average*(the score you give/10)

This means if your average is 7,69 like mine and you give an anime score 10, then your actual rate will be 7,69
if you give 5 then it will be 3,845


Wait, then how an anime will get a 10?

If someone has an average of 5? He's score will never be counted as high?
And if someone has an average of 9? He have more effect on the anime.

Something is wrong...
Feb 8, 2016 9:37 AM

Offline
Jan 2015
116
galimx said:


That doesnt make any sense. How do you know if you will like something before you even watch it? Even if you read the synopsis and reviews and recommendations, its not always 100% you will like something, or dislike something. You need to experience it yourself and try to judge it in an appropriate way.


Between synopsis, reviews, and watching the first couple episodes, it is not too hard to find something you will probably like rather than going in blind.

And giving a little lower scores has nothing to do with not liking animes. Even scores of 5 and 6 mean average and fine, not bad scores. Some people like to look at anime in a more critical way and evaluate them with different scores. Im pretty sure if you try to look at your anime in a more critical way and differentiate between them, the scores would be a little more different. Some people are just too generous when scoring, others are more harsh and have higher standards. And another thing is, you watched only 100 shows, so a mean score of around 7 is pretty fine. But as you watch more and get more experience in this medium im sure your mean score will balance more towards 5 if you use the whole scale like it should be used.
But overall the scores doesnt matter at the end of the day. What matters are your reasonings behind liking a series, and these should be backed by valid criticism, not just pure "fanboying" or ranting about a series without any justifications. Personal experience and constructive criticism is the way to go.


I watch anime for fun, not to be a critic. I have absolutely 0 interest in critically judging a series, I rate based on how much I enjoyed it.
Feb 8, 2016 9:41 AM

Offline
Apr 2014
4399
@galimx thats completely bullshit to say the more anime u watch the more closer u reach a 5/10 mean score

ive watched as much TV series as u have and my mean score is 7.42

many ppl dont finish a show just to rate it a 3 or 4. just because u cant pick a show properly doesnt mean everyone will have low means.
Feb 8, 2016 9:53 AM

Offline
Nov 2014
406
@FunnyGames I didn't say it's good, even that should be fishy that the link he provided there's not a single score, only rankings (I guess).

Times like this is when that signature come into my mind that a mod (if I remember right he is a mod) uses: "When you are losing an argument post sites with doubtful reliability" or something like that
Feb 8, 2016 9:54 AM

Offline
Jan 2015
116
galimx said:

Enjoyment is the shows qualities clicking well together and so the viewer experiences a term that is called enjoyment. But this is not mutually exclusive with critical evaluation. Someone enjoys something because they thought it was good. There are a lot of factors that lead to enjoyment. Enjoyment is connected to story, characters, direction, overall mood, animation, pacing, etc,... This are ultimately factors that contribute to your overall enjoyment of a series.


I agree, which is exactly why I only rate based on enjoyment. Enjoyment encompasses all the factors of a show, both the factors that we typically see in critical reviews, and the factors that you can't explain or see. If you would give a show an 8 based on enjoyment, but give it a 6 instead after critically reviewing it, then I believe that is inaccurate, because that means there are factors that you either didn't understand or cannot put into words that contributed to your enjoyment.

The end goal of (most) anime is for the viewer to enjoy the show, so I think rating based on enjoyment is the most accurate way to rate something.
PhannFeb 8, 2016 9:59 AM
Feb 8, 2016 9:57 AM

Offline
Apr 2014
4399
galimx said:
moodie said:
@galimx thats completely bullshit to say the more anime u watch the more closer u reach a 5/10 mean score

ive watched as much TV series as u have and my mean score is 7.42

many ppl dont finish a show just to rate it a 3 or 4. just because u cant pick a show properly doesnt mean everyone will have low means.

Well I like to differentiate a little between the shows, so I use the whole scale according to what it says. Not all anime I've watched are 7 (good). Some are just fine (6) and a lot of them were average and generic (5), most seasonals that the industry is giving us nowadays comes into this range. There were also some pretty bad anime with a lot of flaws (scores 1-4). I can still enjoy them to some extend, but I cant overlook their flaws and generic overused tropes, while I judge the aspects of a certain show. Like I said, some people are just to generous while giving scores or dont put a lot of thought or effort in rating and others are more harsh and have higher standards. Either way, thats pretty normal.
see ur assuming i dont use the whole scale

i do but i just pick good anime sometimes there are flops and i give them bad ratings. this entire thread is built on assumptions that ppl are using 7/10 as their average. and not taking into account that maybe ppl just arent watching much bad anime
Feb 8, 2016 10:23 AM

Offline
Jun 2014
3667
You have my full support. That's really one of the best suggestions I've ever seen here.

Edit: @galimx (or anyone else), can you help me with a screenshot of the Top 20 please (I can't scroll the page till down there for some reason)?
AgafinFeb 8, 2016 10:30 AM
Feb 8, 2016 2:39 PM

Offline
Nov 2010
656
@Agafin - you asked for OP website, right?

Feb 8, 2016 4:44 PM
Offline
Mar 2014
2421
Delenai said:
No, on the MAL scoreboard 5 is called "average", that is obviously bullshit. For all your information the REAL average on MAL is 7,23 atm. Fun fact that there's hardly anyone who rate anime like the scoreboard say, but then there are ppl like him who do it and try to "force" others to do it too.

I'm not forcing anyone to rate a certain way. The scores will just be processed differently than they are rated.

Delenai said:
@FunnyGames I didn't say it's good, even that should be fishy that the link he provided there's not a single score, only rankings (I guess).

Times like this is when that signature come into my mind that a mod (if I remember right he is a mod) uses: "When you are losing an argument post sites with doubtful reliability" or something like that

If you were to go through the trouble of reading the second and first posts, you would know that I am not the creator of the site. If you really care this much, look through its PHP. It's not difficult to do so, nor is it heavily guarded. The data isn't fabricated in any way.

Also, I'm a girl. Please stop referring to me as "he".

@moodie/@Phann/@galimx
I don't care about how you guys rate or why you do it. This is not a 'How do you rate your anime?' thread. I am aware that it doesn't benefit everyone, but it is fairer than the current system. Please stay on topic.
vegetablespiritFeb 8, 2016 4:51 PM
This topic has been locked and is no longer available for discussion.
Pages (5) [1] 2 3 » ... Last »

More topics from this board

» Reply button should show text that will be quoted.

Euthymia_Gerv - Apr 15

8 by Alexioos95 »»
Yesterday, 10:22 AM

» @ sign spam/attack

kuroneko99 - Apr 16

3 by Jhiday »»
Apr 17, 11:11 AM

» Preview for "About Me" section

NubFix - 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 it

Thruddingsfly - Feb 15

39 by 3miL »»
Apr 14, 11:07 AM
It’s time to ditch the text file.
Keep track of your anime easily by creating your own list.
Sign Up Login