forum

Enforce a minimum beatmapset total playtime

posted
Total Posts
50
show more
vergil chair
I appreciate this finally got some attention, now it's time to clarify a bit more what were my original concerns I brought up in mapset you've mentioned with potential suggestions on what to change.

Your proposal could be working well but it tends to be useless in certain cases. If I'm about to make a 30 second mapset (whether it be original length or cut), in order to make it rankable with proposed changes all I need to do is just add some more diffs and all is good. Meanwhile this seems like perfect solution, this ruins the idea of making small sets for rather calmer songs, or songs which aren't really that various in terms of creating more interesting diffs. Would you find more interesting small mapset for calm song or bigger set potential for exact same song?

This leads to next thing I want to mention. Your proposal dismisses original concern I brought up here regarding drasticall song cut of a featured artist. While originally I used mappers' guild rules regarding official songs cut (which is somewhat acceptable if cut contains reasonable length, can be seen of FAQ tab on their website) which doesn't really involve mapper to obey them if outside of the guild, I feel that specific rule should be live on any song leading towards ranking section where major cutoff has place, as it is disrespectful towards artist and massively goes down on variety certain song provides, especially when this is a featured artist song. Which brings me to the proposal point I want to write down here (I'm not that good in wording my thoughts in foreign language but you should get my point):

  1. Avoid drastically shortening song's MP3. This is to ensure quality of a Ranked section with variety of songs. Meanwhile minor edits are acceptable, using major cut-offs of content created by artists should be discouraged.


While I have mixed feelings about point you brought up about mapsets which doesn't really add much into ranked section (having barely over a minute of total drain), in some cases there's nothing mapper can do about it, that's how the original song was designed so this may result in content bloat if were about to make at least 6 diffs for not that intense song. That's why this should be a bit more centered about song cuts overall instead of trying to go with the way you proposed.

Minor cuts like 5150 where you usually get rid of one verse & chorus section in rather long song (which is pretty popular nowadays) still can have place because their usage doesn't fully intend to get rid of all song content artist had created. While personally I don't fully support that, this seems to be a reasonable cut. Major cuts like Uta, with 30 seconds of length out of original 9 minutes (extreme case), or longlasting in most played Make a Move, with barely exceeding 50 seconds out of original 2:50 length (questionable case), should be discouraged.

Despite the thing I came up with quoting rules from mappers guild in original discussion post, I strongly believe the way cuts are being hold should be more strict, no matter if it's official featured artist song or not.

tl;dr major cuts bad, minor cuts good, total drain time proposal questionable
dennischan
I agree with the concerns that Rolniczy has bought up. After communicating with Rolniczy, we have decided to suggest a different rule to limit excessive cuts which do not contribute much to the ranked section.

Rule writeup -
"A cut song must not be shorter than 1 minute, and it must additionally not be less than 50% of the original song. This is to prevent excessive cuts which do not provide enough content to the ranked section.

Official TV sizes are exempted from this rule as they are cuts produced by the artist themselves.


Rationale
This rule can prevent excessive cuts better than the proposed rule and it also allows interesting short songs to still reach ranked unimpeded. It also prevents maps with excessively cut songs from circumventing the proposed regulation by just making more difficulties.

The main objection to this would be that it bans TV size. As we all know, TV size anime maps are often among the most played maps and banning them would cause many players disappointment and anger. Therefore, to alleviate this problem, we suggested to exempt official TV size cuts to allow mappers to continue making these interesting and engaging maps.
DeletedUser_10414170
Don't exempt just TV sizes but any official cuts by the artist themselves. A good example would be songs from rhythm games. https://www.youtube.com/watch?v=Q9vkrmbYiDo is technically a cut of https://www.youtube.com/watch?v=koY4VSycPa4 but that's because it's for another rhythm game.
dennischan
Yea the actual wording can be changed, this is just a rough draft for people to refer to.

Something like the actual percentage allowed to be cut and the time required can also be changed according to vote or community input
shardex
I guess this proposal should exclude songs compilations for more clarification (unless someone has a word for this)
Daletto
I think Ephemeral made a good suggestion from my point of view.
celerih
I actually p much disagree with this, feels like it's trying to block one specific set without really considering the impact on other stuff. If an artist makes a calm short song I don't see why you should be forced to create a ton of diffs for it, especially when these calmer songs don't necessarily have much to offer rhythm and creativity wise. Content bloat is a serious problem and this proposal wouldn't really help with that

Something much better would be to rule out cutting songs past a certain % or just disallow cutting stuff so it's shorter than 1min. It's already been made apparent that trying to rule out cutting stuff all together isn't gonna happen but we can at least try to avoid the really really bad abuse cases. Plus this proposal wouldn't even cover other really bad cuts like https://osu.ppy.sh/s/765778 which imo are just as bad as the stuff like uta intro ver
yaspo
Don't really see the issue here from a content perspective. Not a lot of content so it doesn't get a lot of attention, fairly logical? Nobody loses out here.
From an effort perspective, UC already said it, it groups those who just want to increase their ranked map count with those who want to actually do something, and that's harmful for the latter.
I find the suggestions to avoid drastic cuts like the one linked a better solution. In my opinion, the ideal approach would be to increase the minimum drain time to 1 minute unless the song is an original work and not a cut, remix or cover. That more or less covers your drain time thing, 3 1 minute diffs = 3 minute, and also ensures the map isn't over before it starts.
Noffy
A cut can not be over 50% of the full song if the resulting length is less than 1:45. This does not apply to official cuts or recreations of official cuts.

What this prevents : cutting songs to extremely short lengths
what this allows: cutting songs to typical rhythm game lengths (~2m), and any amount of cutting so long as a minimum of the song is kept in tact.
two minute songs can then be cut to one minute still, but five minute songs can not as it is doubtful the structural integrity or appeal of the song would be kept.
This way would avoid being too harsh, as it is not uncommon to cut four minute and a half songs to tv size-like length, which is still capable of maintaining song structure (intro-verse-chorus-outro). BUT this would disallow excessive cuts of longer songs such as cutting a five minute song to a 30 second chorus section.
celerih
I could not agree more with what Noffy is proposing, this seems like the perfect way to handle cuts while still leaving the option to make cuts of similar lengths to be like other rhythm game stuff
Nao Tomori
1:45 is not tv size. I make most tv size from full mp3 to get a better ending usually, this would stop me from that. Putting number limits on these kinds of things is dumb anyway, for example pan de peace has an opening song 30 seconds long, so if someone tried to rank a set using a cut from the full version which sounds more natural, that would still be disallowed. But using a worse sounding official release is allowed.
This also wouldn't be a "recreation" because those cuts don't use the TV size label.

The rule should put judgement on to BNs rather than trying to quantify it, that way these things can be discussed rather than having some rule pointed at and misunderstood constantly like will happen if this goes into effect.
vergil chair
If you edit mp3 to imitate official tv cut on terms of its design, then there shouldnt be a problem with it because you're adjusting already existing cut but in your own way. Little differentiation is fine as long as there isn't major difference. Edits like https://osu.ppy.sh/beatmapsets/968570/discussion/-/generalAll#/999039 will be discouraged for example
Mao
I agree with many people in here in that I think the approach this proposal takes is not solving the underlying issue at all. Let's look into this bit by bit.

First of all enforcing a rule like this will promote bloating sets with filler difficulties. Can't rank you set because overall drain time is too short? Let's just add a Beginner and Advanced difficulty (which is not a lot of effort and something mappers will most likely do) and you are fine to go. We have kind of tried to avoid bloated spreads until now so this goes kind of against our current principles.

Second, in relation to my previous point, this will mostly punish mappers that care about having a well balanced spread on 30s songs they also might care about. If we take a map like the one that was linked in the OP for example, "low effort" things like this could still happen, just with 5 instead of 2 Easy difficulties. What you would actually achieve with a rule like this is that songs that have an actual target audience and therefore people who want to play them (e.g. short cartoon openings) that might not call for an Expert difficulty would either need a bloated spread or were not rankable at all, despite them adding reasonable content to the ranked section.

I think the issue we should actually be discussing instead are "lazy" cuts of songs that have long versions available or other edits of songs that are purposefully made to abuse the minimum drain time rule. Lanturn and I made an example some time ago of what could actually be done with the current ruleset (a map that's essentially just a stream) and things like this or minimum time cuts would not be stopped by a rule like this.

Also in case the rule should actually pass, 2:45 is ridiculous as it would not allow for actual E/N/H/I (which I consider full spreads) on 30s songs so the absolute maximum I'd go with a rule like this would be 2 minutes sharp.

Also strong +1 to Noffy
Mordred
would rather add a rule / guideline like "you must not cut more than 70% of the original audio file", filler diffs aren't gonna solve anything

basically agree with noffy
pimp
there was a rule preventing songs longer than a minute to be cut to less than a minute, i don't know when or why it was removed...
should be easier to implement something similar instead of that total playtime rule.
Full Tablet

Mordred wrote:

would rather add a rule / guideline like "you must not cut more than 70% of the original audio file", filler diffs aren't gonna solve anything

basically agree with noffy

That wouldn't work as a rule, take for example パインツリー - サニーミルクの紅霧異変 (Circle: UNDEAD CORPORATION) https://www.nicovideo.jp/watch/nm19311262 which is 55 seconds of a piano intro, followed by 123 seconds of silence, and then 114 seconds of death metal. If you can only could cut 30% of the song, you would be left with too much silence in the audio file to make the beatmap rankable, making the song effectively unrankable.

The map that brought this matter into discussion is also an example where this wouldn't work https://osu.ppy.sh/beatmapsets/968570#fruits/2026676, the song has a calm piano intro, followed by metal. Someone wouldn't be able to map the intro of the song alone, even though artistically it could stand as a song on its own.
Vulkin
I think adding a BPM as a base for short songs would be better if youre that into this

Personally I dont like it cuz it would pretty much kill most R3 Music Box maps, even if theyre not that interesting
Nifty
Too many people think maps that can only support 30 notes as a top difficult can be considered interesting. I'm pretty sure a larger consensus can be reached when saying a 30 second map can be interesting if it can support a full spread, as Eph has stated.

I've also never came across any 30 second mapset that is anything to behold as anything more than low effort, SS-farming ranked fodder, so I would love this rule addition.

Note: A full spread meaning something like this, https://osu.ppy.sh/beatmapsets/961335#taiko/2012566, not like this, https://osu.ppy.sh/beatmapsets/958476#taiko/2006814.

I would not be against banning all r3 mapsets from being ranked in the future either.

Vulkin wrote:

Personally I dont like it cuz it would pretty much kill most R3 Music Box maps, even if theyre not that interesting


That's the point.

shardex wrote:

I guess this proposal should exclude songs compilations for more clarification (unless someone has a word for this)


What 30-second song compilations are you listening to?
Vuelo Eluko
As an extremely inconsistent player with shaky aim that is lucky to hit a 300 combo, I mostly agree with this change. However, I'd rather see more of a star rating or pp penalty to very short maps because dumb gimmick maps like Vivid are good for the memes. A map does not necessarily lack the merit or worthiness of being pushed through the ranking process simply because it is objectively short on real-time play value. You can have a quality experience that doesn't last very long, also I don't think osu! is optimizing a business model around keeping people playing long maps like some kind of subscription based mmo.

Anyway that's just my two cents. If a 20 second long map isn't worth the effort I would expect BN's to simply gloss over it. At the very least, they are easier to vet and assure quality by virtue of being shorter.
Stefan
The fact 'low effort' is used as arguement to disallow small and short mapsets like we have right now is hilarious at its finest.

Five minute difficulties are done in half an hour in certain cases, mapsets with +8 difficulties where the uploader has done two at max are a thing, and you rather mock about short mapsets having a short playtime?
Capu

Ephemeral wrote:

Establishing a minimum 'effort' cutoff via total beatmapset playtime addresses this issue, and ensures that every beatmapset that hits Ranked is at least offering players a reasonable variety of challenges to overcome.


In my honest opinion, this doesn't provide a good argumentation for given "issue". If the goal is to ensure, that a player has the option of overcoming a variety of challenges regarding the whole set, single diff marathons shouldn't be a thing as well. They're not offering content for the whole playerbase as well. Having a 5* 5minute marathon is most likely not enjoyed for 50% of the playerbase, since they do not provide challenges for all skill levels either, yet they are a valid thing.

I'd be more concerned about using super short songs in order to "farm" a number of ranked maps. Therefore it'd make more sense, to introduce some sort of "short songs ranked per (add reasonable time value here)" for every mapper.

Example:
I map a 30 second "set" with 2 or 3 diffs, which gets ranked on the 1st of July. With my previously mentioned idea in mind, my next small set with super low drain time would have to wait at least 3 months, in order to get ranked, which would be the 1st of October.

This would not force a mapper to make a bigger set, while decreasing the farmability for the ranked map counter. Many songs that feature a low drain time, just do not offer the kind of "flow" in order to really make a clean spread throughout the difficulties. This would much rather lead to 6 diffs being nearly the same, which would not form a good spread and drastically decrease the quality of the set in its whole. Other than that, some songs are just super good, yet extremely short. Why force the mapper to make 6 diffs of that same song, without the opportunity to even create a good spread in order to get the map ranked?

Furthermore, there are people that just enjoy very short maps. For whatever reason. Bringing the previously mentioned marathon argument up again, some people do not like long maps, yet they're totally fine to be ranked. (and once again, they're not even playable for a big amount of players, when the map reaches a certain difficulty level.)
Short maps are playable for everyone. No one is forced to play these maps, yet it is additional content. If the quality of given sets is good enough, they should be able to get ranked, aside from the low drain time.
Topic Starter
Ephemeral
Limiting cuts introduces another problem: cuts are not intrinsically bad. In my view, there is no issue with someone making even very short versions of a song in order to map particular sections. Yin and a bunch of other beatmaps have done this successfully over the years with no issues. Granted, there have also been a number of sets that have done it poorly as well.

It is a greater affront to mapping freedom to arbitrarily limit the length someone can cut a song for than it is to assert minimum levels of total playtime for a given set. That being said, they are also entirely different issues!

What I am principally concerned with is sets with largely insignificant total playtime being passed through the ranking system and emerging on the other side offering almost nothing to players who choose to challenge them. I don't particularly care if people choose to map R3 or other short cuts of music so long as there is adequate content for players to challenge at the end of it all.

That is the crux of this proposal, not whether cuts are okay or not.

The arguments that this may cause possible spread bloat in an effort to circumvent the requirement are valid and need some consideration. Bloat is not ideal and having a 30s set with 6 beginner difficulties does not sound particularly fun to anyone, but neither is having a 30s set with two beginner difficulties. Some limit needs to be applied for these kind of cases, but I am not really sure that attacking cuts is the best or even an appropriate way to do it.

With this in mind, does anyone have any ideas on how to tackle the issue, or is community consensus more aligning towards restricting the usage of cut songs? Remember, freedom is key here.
Capu
As a direct answer to your (Ephemeral) question at the end of your latest post regarding a fix for the issue:

I could potentially see a rule working, which implies a strict no-GD policy for songs, that are (arguably) 45 seconds or less of drain time with at least 3 difficulties.
This would not harm the freedom of song choice in any way, but would force the mapper to really map it by himself, in order to achieve a certain amount of self-mapped time.
There is of course plenty of detail to this, that could be added, but I feel like this is an idea that could work well with some more specifications.

Edit:
More ideas for this would be something like this

If the drain time of the song is less than 45 seconds, the set is not allowed to contain a guest difficulty, except for if the host has mapped 4 difficulties by himself. For every guest difficulty added to this, the host needs to map 1 additional difficulty by himself.

If the drain time of the song is more than 45 seconds but less than 60 seconds, the Set is allowed one guest difficulty and the host has to map 3 difficulties by himself. For every guest difficulty added to this, the host needs to map 1 additional difficulty by himself.

An exception could apply for both of these, if the whole set is reaching 7 difficulties (for songs with less than 45 seconds drain time) and 6 difficulties (for songs with more than 45 seconds but less than 60 seconds of drain time) more guest difficulties are allowed, if the host has mapped the previously mentioned amount of difficulties by himself.
Topic Starter
Ephemeral
For transparency's sake, the 165 second cutoff in the proposal would have prevented the following sets from being ranked (all are from the start of 2019):



Judging from playcount as a measure of engagement, this alone sort of turns my initial proposal on its face, as people are clearly engaging with these lower total length sets as is. Hmm.
Mao
Yeah, those sets have much more of an audience than say one diff 4 minute Insanes.
dennischan
yea it sounds like the Noffy proposal is flawed according to the clear statistics from eph's post. However, I can't think of a way forward since well both proposal have their downsides.

Eph's proposal will force 30s maps to get bloated content to fit this rule, while Noffy's proposal bans maps that the community enjoys and engages with. Maybe we should get more discussion before moving forwards, or maybe even finding a way to survey the wider public on the matter.
Pachiru
I support Ephemeral's idea.

osu! is a rhythm game, which is made to bring challenge to the players and it would be more fitting to have 1:30/2:00 songs with Easy/Normal/Hard/Insane/Extra sets rather than an 30 seconds R3 Music Box (taking this as an example since it's the first thing that came up in my mind) with two Easy difficulties that are mostly 3 4/1 sliders along the whole map.

I'm not saying that osu should be an another popular rhythm game copycat though, but it should at least fits what it's made for.

Maybe reducing the time to 2:30 instead, to avoid harming some sets.
Noffy

dennischan wrote:

yea it sounds like the Noffy proposal is flawed according to the clear statistics from eph's post. However, I can't think of a way forward since well both proposal have their downsides.

Eph's proposal will force 30s maps to get bloated content to fit this rule, while Noffy's proposal bans maps that the community enjoys and engages with. Maybe we should get more discussion before moving forwards, or maybe even finding a way to survey the wider public on the matter.
My proposal would restrict but not kill short cuts, as short cuts of short songs are still allowed to an extent, and official cuts always allowed. The statistics ephemeral posted are for maps that wouldn't be ranked with his own proposal
Nao Tomori
that proposal makes no sense though. basing it off of arbitrary numbers will just have an adverse effect overall by restricting several types of normal cuts. applying a massive blanket to target 1 (one) extreme case is illogical.

i have a proposal too, based on what has been commonly accepted and not accepted. this leaves it up to BNs to decide what is suitable or not and removes confusion while allowing more freedom for cuts that are ostensibly not being targeted.

Guiding Examples for Acceptable Song Cuts

Acceptable:
1. Cuts which replicate official releases of songs with minor or no changes. Examples: https://osu.ppy.sh/beatmapsets/970410#osu/2031002 , https://osu.ppy.sh/beatmapsets/875083#osu/1828823
These cuts are basically cuts I made because the ending of the TV (or game) size was terrible, so I made my own version.

2. Cuts which remove some of the song but maintain significant parts. Examples: https://osu.ppy.sh/beatmapsets/782955#taiko/1659378 , https://osu.ppy.sh/beatmapsets/745312#osu/1571309 , https://osu.ppy.sh/beatmapsets/665232#osu/1407924
All of these cuts remove some parts of the song but keep most of it intact.

3. Cuts which create TV size/game size equivalents of full songs. Examples: https://osu.ppy.sh/beatmapsets/909739#osu/1898484 , https://osu.ppy.sh/beatmapsets/775932
The idea behind these was that it would be how the song would be cut if it were used as an anime or eroge opening. They have the same structure as a normal TV size.

Case by Case:
1. Cuts which are shorter than equivalent or theoretical official releases, or cuts shorter than 1 minute. Examples: https://osu.ppy.sh/beatmapsets/649373 , https://osu.ppy.sh/beatmapsets/765778#osu/1627148 , https://osu.ppy.sh/beatmapsets/730303#osu/1541293
These cuts tend to remove a lot of the structure of the song, so they should be reviewed and discussed before being deemed explicitly acceptable. (technically speaking Harumachi replicates the official release but it's retarded)

2. Cuts which remove parts of the song not conducive to mapping.
Don't really have examples of maps for these but theoretically if someone cut out the first bit of https://osu.ppy.sh/beatmapsets/617687#osu/1302335 you wouldn't miss much of the interesting bits in the song. Alternatively, if there was a minute long break in a song, cutting it would be fine.

Not Acceptable:
1. Cuts which remove significant parts of the song. Examples: https://osu.ppy.sh/beatmapsets/823960 , Absolute Soul (a cut which ended right before the chorus of the TV size started).
The essence of the song is lost when significant parts are removed. Basically the reason for this proposal in the first place.
Kibbleru
Yeah, i'm not really for the idea of setting a drain time threshold for ranking, as there does exist songs out there that are 30 seconds or 1 min that are actually good (the one that comes to mind is the ori and the blind forest intro).

I'm more against people cutting shit into "ringtone size" to abuse the ranking criteria.
Irreversible
We have different problems with maps currently, let's focus on those instead of such superficial criteria that, as you say yourself, won't have an impact on the mapping scene.
Stefan
It's just funny that the proposal exists because a song has been cut off, and isn't naturally short. Also dito @ Kibbleru
Krfawy
How about we abolish the requirements for the length of the song and just let people map extras alone no matter what drain time it is as long as the drain time equals at least 2 minutes and 45 seconds? That would be more productive than forcing people to such weird standards as written in the proposal.

On a serious note, forcing at least 2:45 drain time for the whole set is as brilliant as the idea of forbidding mapsets with similar level difficulties proposed a few years ago. If the issue is them songs being cut too much, we can always have a guideline or a rule forbidding doing that instead.
dennischan

Noffy wrote:

dennischan wrote:

yea it sounds like the Noffy proposal is flawed according to the clear statistics from eph's post. However, I can't think of a way forward since well both proposal have their downsides.

Eph's proposal will force 30s maps to get bloated content to fit this rule, while Noffy's proposal bans maps that the community enjoys and engages with. Maybe we should get more discussion before moving forwards, or maybe even finding a way to survey the wider public on the matter.


My proposal would restrict but not kill short cuts, as short cuts of short songs are still allowed to an extent, and official cuts always allowed. The statistics ephemeral posted are for maps that wouldn't be ranked with his own proposal



Sorry for misunderstanding the statistics that eph released. In that case, I think that Noffy’s proposal would be the most appropriate since it would limit low effort cut songs while not killing short cuts of short songs.

The only thing I would think of to improve the Noffy proposal would be to reduce the minimum cut off time to be 1:30 since many rhythm game songs are also approximately 1:30. I feel like 1:30 is already sufficient length to produce an interesting map that keeps the audience interest throughout. That is why I believe rephrasing the time limit to 1:30 would be better.
Drum-Hitnormal
what about TV size that's actually really short?
https://osu.ppy.sh/beatmapsets/981859#osu/2054965 adding useless notes thats not mapped to any sound just to make it 30 sec is ridicilous
UndeadCapulet
seems like the idea that's caught on more in this thread is an entirely separate proposal about cut length. thats cool and all but should be rerouted to a separate thread prob to be iterated on more effectively

so uh here's that thread i guess https://osu.ppy.sh/forum/t/917229/start=0
pishifat
given:
- the flat length cutoff can be circumvented with unnecessary additional diffs (and thats lame)
- people are more concerned about stupidly cut songs than short maps
- short maps are still engaging according to https://osu.ppy.sh/community/forums/posts/7101191
- thread is inactive

gonna archive this. if there's a better way to implement the idea, outlining it in a new proposal would work better than continuing this hard-to-follow thread
Please sign in to reply.

New reply