AMV Content Rating System

[Phade]

Ask your boss at the con.

My boss at the con is ultimately the audience. A content rating provided by the audience for the audience would be extremely helpful for me to help the audience.

[Xophilarus]

Others overall have expressed my personal thoughts on this more eloquently than I can at the moment but there is one concern you have been dismissive of that I feel you really cannot brush off...

Given that the feedback is a number, it is impossible to know or infer the reasons why you picked that exact number. Did you pick the number because of nudity, language, tension, interactions, drugs, alcohol, demons, farts, blasphemy, gambling, violence, or because you don't like the color green? We, the audience and video creator, cannot know this from a number, absolutely impossible. Unless you use words publicly, you and only you will know the reason why you picked that exact number, forever.

Ultimately, the overall score is determined by a large group of people. If an individual gives an AMV a high score for [any reason] but everyone else thinks the score should be generally lower, the outlier value will have essentially no effect.

I'm going to tell you right now, with a system like this, as Keiichi pointed out, videos that feature a member or members of a marginalized group (LGBTQ+, disabled people, people of color, etc) will get higher ratings. I promise. It will be able to at best enable microaggressions and at worst actively suppress and put down videos that features these subjects/people. Even if review bombing is accounted for, unfortunately, a large amount of people will unconsciously still attribute these themes as being "less wholesome" or "more inappropriate" even if there is nothing remotely "objectionable" about the video except the mere presence of said group. Again, even if it isn't malicious, a lot of people will still have a tendency to rate these videos higher. You said it is "impossible to know or infer the reasons why you picked that exact number"—trust me, it absolutely will be clear as to why in these instances. Recently, behaviors and actions that put down these groups has become more and more normalized and enabled in society. Please don't add a system that helps bring that here too. I also implore you to keep this kind of thing in mind for future implementations of new features for the ORG. I want to be excited about the redesign. Not dread it.

[KeiichiFace]

My boss at the con is ultimately the audience. A content rating provided by the audience for the audience would be extremely helpful for me to help the audience.

gang you know who i meant

[laceproductions]

A numerical rating system as proposed so far and as expected to be used, even assuming we get enough creators/viewers to actually take the time to do the ratings, and assuming that they will adhere to the guidelines outlined, WILL skew the results in favor of the cultural views of the group/region that uses the site the most. Are we wanting to be fully internationally inclusive? Are we going to expect a skew to a more US/Canadian/UK/English speaking population? In the end, whomever has the most representation in the pool of those rating will skew those ratings in favor of the cultural leanings of that population.

The conversation on difference between violence, nudity, sexual content, and language use so far really hits on the nuances of the accepted "wholesomeness" of any piece of content. Take for instance the general stance on nudity and violence between Japan and a lot of Europe vs the US. In the US nudity and sexual content of any amount is often viewed as being extremely taboo while ultra violence can and often does get seen as benign and something often considered safe for younger audiences. In Japan and much of Europe the opposite is true, nudity is often seen in a similar light as the way the US views violence but violence is extremely taboo. This is just on general region accepted, not even going into the nuances of religious beliefs and the like.

I don't even want to start in on topics or editors that falling to any minority group, LGBTQIA+, POC, ect where political fervor and emotional charging creates large waves of debates regarding 'wholesomeness' of such topics or even anything created by members of those groups.

All that to say that when programming an algorythm to recommend or list content based on this idea of 'wholesomeness' (which has been stated multiple times is an extremely nebulous term that everyone will have differing opinions on) makes for an easy shortcut to list videos with one simple lookup, tagging of content (which I believe is already something that is planned to happen) would make for a much more accurate depiction of the video's content and allow the viewer to sort out content they and their region views as 'not wholesome' much more easily and accurately.

The number itself will end up being useless to anyone outside of the majority group of reviewers, or the member or group of members that ultimately decides what each number rating ends up being listed as in the guidelines.

[irriadin]

There are reasons why most recommendation systems (see Youtube, Netflix, etc) have boiled things down to maybe 2 choices: People don't want to spend time thinking about things, and people also try to compensate for other people's views.

Exactly. Users are accustomed to uncomplicated ratings systems because of its frictionless nature. They don't need to learn an entirely new system, it's inherently obvious to them (like / dislike vs the proposed rating system here). Unfortunately, for better or for worse, a-m-v.org will be judged by the UX standards of sites with similar social features. If the system proposed were implemented, I think that most people would not understand it and choose to not engage with it. Except for bad actors, potentially.

Safety and Trust teams at social media sites (ones that still exist...) constantly think about how features can be used and abused by users. The potential for harm is high with basically any feature that you put in the users' hands. So I think it would be prudent to really consider a few things before deploying new or drastically updated features: 1. Will this bring tangible benefits to users? 2. What is the potential for mis-use by bad actors? 3. What are some externalities or novel situations that you can game theory out? Consider Campbell's Law which stipulates that the more important a metric is in social decision making, the more likely it is to be manipulated.

One other way to look at this: users who choose not to read the manual (e.g. most people) will choose the path of least resistance. In this case, what will that look like? Maybe they'll just pick the closest approximation. This seems likely. Other more chaotic users could select at random or purposefully choose the wrong rating classification.

Let's also consider how other users can change the ratings. First off, many users will not choose to engage with the social rating feature. This might be the majority of users. Once we take them off the table, the remaining users that do use the ratings system will probably operate in good faith and rate according to the proposed metric. But there will also be users who use the system to willingly subvert the desired outcome, whether for personal gain, antipathy to the editor / song / anime, etc.

What's the solution? Something that is the simplest way of communicating to both editors and users. If we're thinking about objectionable content like violence, sex, etc. then we could have:

1. General (most everything)
2. Mature (intense violence, some sex scenes. The equivalent of an R rating)
3. Adult (Actual Hentai or equivalent content)

I think that making a distinction for kid-friendly content is a path I would advise against; there are strict child protection laws that you DON'T want to run afoul of, and I doubt a-m-v.org wants the heat that can come your way.

[irriadin]

I don't think there's a way to edit my post so I'm going to double post!

I want to add that marginalized groups are typically among the first to be targeted by bad actors so please do listen to everyone bringing this up. This is one of the vectors of harm that can be exploited by social systems.

[SQ]

For me personally, my biggest worry is about showing AMVs in public to an open audience. [...] What is relevant is the consensus of the general population when it comes to this AMV. If the general population thinks I'm worrying too much and the AMV is actually just PG (my personal numeric value is greater than the average value), this is very good to know and will give me the confidence to include it in the set. [...] However, knowing what an average general audience would think is extremely helpful.

This system will not help you with your set because the org is not your con panel's audience and US Culture is not a monolith. California will have different views on what is permissible than Georgia, and views can differ heavily even when you separate cities from rural areas.

If you choose to filter the data by region, this is also problematic because (1) it's highly invasive for a hobby site and (2) IP location is inaccurate, even before accounting for the people who are using VPNs (which, if the site stays in its current direction, will be more than normal. I had to download a VPN to participate in this thread.)

Furthermore, please consider stepping back and looking inward at why your first instinct to connect with your panel's audience was to develop a convoluted rating system and subject the thousands of users of your international website.

[seasons]

I think a rating system for content isn't a misguided or doomed idea. I believe you're approaching it in good faith. I think it's worth having a conversation over.

I'm also extremely skeptical that the average Org user, let alone the potential new users that I assume a site rebuild is hoping to attract, can be motivated to take the time necessary to understand it or honestly engage with it. Perhaps I'm even a bit fearful that it will be instantly overwhelming to such users and drive them away from the site (even if it's one in the near-ish future that's ostensibly going to otherwise be easier to use on a whole than the one we've got right now). I understand that asking users to rate the content of the AMV they just watched on a numerical scale is an extremely simple concept, but I only think it's simple in a vacuum, or when imagined in the context of the culture that existed way back when this site was first launched. The average Internet user and/or AMV fan in 2025 will likely find this simple ask to be an intrusively hostile demand on their time and mental bandwidth. You can believe that's ridiculous. Perhaps it is! Perhaps we also live in ridiculous times. I'm not saying that we need to aggressively coddle the absolute laziest users, only asking to acknowledge that the rest of the Internet really bends over backwards to do this and has absolutely set that expectation for like 90% of people in diamond-encrusted stone. TLDR: I don't see this playing out successfully.

The voluntary features of the Org, whether they're comparatively lengthy Opinions or even the once-unbelievably simple expectation of leaving Quick Comments, are ignored at least 99% of the times that users have an opportunity to engage with them. No, I don't have the data to back this up, just the fact that I've received only 3 Quick Comments in the last ten years and a hunch that that's probably representative of everyone else. Even the mandatory 1-5 star rating system that caps users' ability to download a certain number of videos without leaving that most bare-bones form of feedback, is largely regarded as a nuisance and often (as testified earlier in this thread) reluctantly engaged with via the least amount of authentic effort or thought possible. I struggle to imagine a future where this new ratings feature would be greeted with any degree of curiosity or diligent duty by new users, when this is how the current batch of users who are extremely familiar with a different sort of ratings features are opting to use/not use them. I want the site rebuild to succeed! I want this community to grow! I don't want the process of downloading or engaging with videos here to grow more complicated and inadvertently derail what could be a promising new era for this place and all that it represents.

Sorry I'm not more optimistic about this. I believe that any potential rating system needs to be significantly simpler than the one that's proposed, not because it's a bad idea on its own but because I don't think it can work in our current climate where attention spans and reading comprehension are in an a once-unthinkable nosedive. This is all completely aside from other users' concerns, which I don't have much to add to but believe need to be taken seriously.

[Phade]

My original goal with the proposed single-value system had one benign focus: education. I wanted to know how my perspective of an AMV compared to a general AMV-enthusiast audience when it comes to public display, and to use a calculable system to determine the comparison.

My personal use case was to answer two questions that often haunt me whenever I create sets for AMV exhibition shows: “Am I being too permissive compared to the general AMV-enthusiast public?” and “Am I being too restrictive compared to the general AMV-enthusiast public?”

To answer this question when needed, I envisioned a system where I could provide an arbitrary numeric value and compare it to an average given by the AMV community. I could then allow the AMV community use the same system to answer the same question for each individual who also wanted to know how their view compared to a group average.

However, it was then brought to my attention that certain aspects of this system, as proposed, could be used in unintended ways by bad actors.

Could we create a true representation that negates efforts from bad actors within the system, especially when it comes to marginalized groups? If this were successful, we could have the opportunity to educate the misguided, uninformed, and confused and potentially change their perceptions for the better. To me, that would make the world itself a better place.

Is the AMV community as a whole, and more specifically, is the AMV community at AnimeMusicVideos.org more likely, less likely, or equally likely to contain bad actors, are uneducated concerning marginalized groups, and are unable to provide a fair and valid consensus?

My perception (and I could easily be wrong) is that the AMV community as a whole is more educated when it comes to marginalized groups, is more likely to provide a fair and valid consensus while keeping marginalized groups in mind, and is less likely to contain bad actors compared to the general global population. I have encountered more fantastically accepting, loving, considerate, kind, and helpful people in the AMV community than anywhere else I have journeyed. I put the AMV community in a much higher regard concerning marginalized group recognition, understanding, and compassion, with the ability to act with consideration compared to everyone else in the world. However, I could be wrong.

I wanted education for my own personal growth and to encourage the education and growth of others.

Here within this conversation, I now realize that displaying a specific calculated numeric score could reveal subtle and nuanced differences that could skew perceptions unnecessarily negatively. I have become more educated.

I now realize that terms like “wholesome” can be considered in ways I did not intend. I have become more educated.

I feel that I have grown in a positive way over the last few days as a result of this recent interaction and education.

As a conclusion, the site will not display any content rating as a specific calculated number. Instead, we will continue to explore and develop a non-numerically-based content rating system to be revealed at a later time. For now, we will move on to the more popular and requested content descriptors aspect (this AMV has language, nudity, gore, etc.).

I greatly appreciate the comments everyone has provided, especially those that are direct, pointed, and well put. This conversation has enlightened me to perceptions that would not have come to mind otherwise. I hope that you continue to provide feedback, encouragement, and understanding as we move forward to making the AMV community (and hopefully the world) a better place.

With humblest respect,

Phade.