Guest
home
beatmaps
rankings
community
store
help
forum
  1. Forums
  2. osu!
  3. Development
  4. Feature Requests

Change the way "map sets" works.

posted
Total Posts
36
This is a feature request. Feature requests can be voted up by supporters.
Current Priority: +24
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic

The Proposal


  1. Destroy the idea of the existence of mapsets. The idea of a "set" will still exist, but instead of treating it in a per "mapset" basis, do it in a per song basis.
Main reason to justify this proposal:
  1. The concept of 'mapset' as we understand now is incoherent with the way that current community treats it.
  2. Mapsets gives the idea that the structure of all the diffs within the set are the same, what I mean with this is, for example, a hard and an insane should have the same kind of style, patterns, structure, etc, and the difference between the two is that the insane uses a more complex version of the patterns used in hard and so on.
  3. What stated in the point above is not always true, specially if we consider the existence of guest difficulties.
  4. Current trends of mapsets is that they are very likely to include at least a guest difficulty, up to a point where only one submitted difficulty is created by the submitter. Ex: http://puu.sh/p3ky5/c764b22cac.png
  5. This has brought some problems within the mapsets as all difficulties are treated equally when they should not.
  6. There are some features requests that attempts to fix some of the issues caused by current concept of mapset ( t/451293 ; t/115773) But they are only "patch" fixes, meaning that they don't address the cause of the problem.

How should it work


First off, this is just a conceptualization just to give you all a clearer idea of how it should work. It could be completely different if it were to be implemented in order to better adjust with ranking criteria.
  1. First time a song is going for qualification, special attention should be given to the song's metadata, timing, and all the other aspects related strictly to the song. This will open the "slots" to submit the difficulties.
  2. After a song's basic info has been approved, difficulties could be submitted and annexed to it. Each difficulty should follow its own qualification process.
  3. In order to avoid approving too many "songs" that will end up having no difficulties later on, it should be done in conjunction with at least two additional difficulties going for qualification as well. These two difficulties should be at least an E+N or N+H to guarantee that the song will provide easy difficulties for new players.
  4. The basic ranking procedure still holds, as in WIP -> Pending -> Qualified -> Ranked, just that they will be done separately.
  5. Storyboards could also be submitted separately. Although this might bring some problems with diff specific storyboards...
  6. Please do note the following, mappers that still wishes to submit an infinite amount of their own difficulties to the set, they should still be able to do so.
  7. In order to add an Extra difficulty to the song, an Insane difficulty must already be qualified. This means that in order to add Extra diff, then mappers will have to work collaboratively to unlock this option! Just like a normal video game!
  8. In order to avoid an excessive amount of Insanes and harder within the set, the ranking rules could say something like: The amount of Insanes or harder cannot be more than the amount of Easy + Normal + Hard.
  9. Many of the points stated here are just examples, the actual implementation should be discussed within the QAT.

Disadvantages and Acknowledgements

  1. First one would be the very obvious one which involves the drastic change of the process of map qualification. The game has been using the same procedure since its beginning and everyone is used to it.
  2. osz2 is coming, and from what I understand of it, means that osu! is being developed in the opposite direction of what this proposal aims to. Due to this, the developers will probably not even care to consider the implementation of this.
  3. Will require a drastic recoding of osu!, as many things won't be compatible with this proposal. Ex: hitsounds, storyboards, etc Some ideas on how to deal with them
  4. Might even further promote speedranking and lack of easy difficulties within a song set.
  5. I have no doubt that many things won't work if this were to be implemented, but I am sure that all of these issues can be overcome by the developers. Just look at osu! the way it is now! Osu was not perfect by the moment it went live and it is so much more different than how it used to be!

Benefits and Advantages

  1. Difficulty's mapper can be properly attributed.
  2. After a song's qualification, it will incentive more mappers to submit individual diffs to the annex to the song. One of the main constraints seen with many mappers that does not finish their mapsets it's because they have to finish a whole mapset in one go in order to seek qualification which is time consuming.
  3. No more need to download X amount of the same song! As all the difficulties will be treated under the same song set. This would also mean that your song folder should weight a little bit less!
  4. More fairness when asking for m4m! Also, I expect an easier modding process!
  5. This will also be great news to new mappers as they can only submit 1 diff on an existing qualified song. They won't have to worry either by trivial things to them such as timing, spread, etc
  6. This new system will adapt to current mapping needs!
  7. This will also ensure that all diffs has the same metadata and timings
  8. Eliminate the need of calling a diff as XXXX's diff, which is just retarded
  9. Properly rate maps by difficulty instead of by mapset.
  10. I used to have thought of more benefits, but forgot them. I'll add them as I remember them, if you guys can also think of any, just post and I'll add!

Additional features to complement this

  1. Ability to give multiple author per difficulty, this is aimed for collabs
  2. An option to auto add mapper's name as a SB for the collabs too. What I mean is to add an pre existing format for this, and if the mapper wishes to customize it, then the mapper could always storyboard it.
  3. Give more song's metadata option, in order to be able to specify more what version of the song we are dealing with. So no more adding Tv Size, etc in song's title
  4. While doing such drastic changes to osu!, implement scorev2 system as well. Give a proper new chapter to osu!

Ok, this is all for now. I know that there are still many things that have to be worked on, but this is just the idea. We could also use this thread to further discuss about this.
Last edited by Luvdic , edited 12 times in total.
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker
Old text
Oh uh wow. I was going to suggest something similiar during ztrot's ranking criteria drama but didn't due to how controversial it would be. I wrote the following up in google docs and locked it up, but I guess now is any better time to post it:

Have the ranking processes based per diff instead of per beatmap set and have a weighting system that promotes diversity in maps' difficulty and song length

Make a system that promotes diversity in the beatmap pool. A system that can put weight on the ranking requirements if there is too much of something and make it easy to rank if there is a lack of something. The system should weigh by difficulty and length per song. If a certain song has around 3 instances of insane difficulties then this system should make it harder to rank another insane difficulty. To make it harder to rank, the system will raise the amount of stars needed to be nominated by BN. Oh look at that, stars can actually have value in this system!

Difficulty is not the only things this system should look at, length as well. If there are 3 instances of insane difficulties of a TV size song and someone wants to map a full size version the song, then the it should be a bit easier to get it ranked. Or vice-versa if there are many full-size version and a few TV sized versions.

What will this system do? Well first the most controversial thing it will do, is shoot the idea of beatmap set’s uniform difficulty spread to rock bottom. Players will be able to map and rank mapsets with just one easy difficulty, causing more instances of of single-diff mapsets. It is controversial because we have accepted a system where there is a beatmap set with multiple difficulties.

Let me ask you this: Does it make a freakn difference if the normal difficulty was made by one mapper and the hard difficulty is made by another mapper, causing the maps to be part of different mapsets if you can just filter/group the maps to what you want? Yea it’s unorthodox, but it makes a more flexible system.

Ok, so what if the same mapper decides to later on map another diff of the same song after one diff got ranked? While unranked and qualified, it will be part of its own beatmap set. When a new diff is ranked to the ranked beatmap set that matches the metadata. This will allow mappers to focus and rank one difficulty at a time.

Con: It might take more time to rank a full spread if planned since mappers are likely to focus at one difficulty at a time. This would also require a kudosu star requirement per diff instead of the current 12 for the whole beatmap set.

Pro: Let’s face it, you probably care more about some normal diff than an insane diff or vice-versa. Some difficulties don’t receive the needed attention because everyone is hyped for that one diff that everyone is modding. Moreover, most modders have specific requirements such as needing to be a certain diff to be accepted for modding. It would allow the difficulty receive more thorough checks by the mappers and by players that support that one difficulty. The whole beatmap set doesn’t need to suffer because of that one diff that doesn’t receive enough mods.

The mapper should not be limited to one diff map at a time. By all means let the mapper do multiple diffs at a time if so desired, but this would allow a more flexible mapping plan and would guarantee a bigger probability something is ranked before the mapper gives up.

On the osu! client side of things, I’d like to point to this request since it will be even more relevant if this system is implemented. When a new diff is added to an existing beatmap set players should know that they can update the beatmap set to get the new diff.

Problem:

I hate this idea

Solution:

State your reason for hating the idea in full description to attempt to convince us that it's a bad idea. Make sure your argument was not yet countered by any of the solutions to the problems which follow.


Problem:

Don't fix something which has nothing wrong with it.

Solution:

There are things wrong with the current mapset system.
  1. The amount of time it takes to rank something.
  2. The work needed to rank something
  3. The system is inconsistent with how the community needs and behaves
  4. The system forces players to do something with no exceptions or with bad alternatives
  5. Current trends of mapsets is that they are very likely to include at least a guest difficulty, up to a point where only one submitted difficulty is created by the submitter. Ex: http://puu.sh/p3ky5/c764b22cac.png
  6. etc.


Problem:

We would just be expecting spreads like E-N-H-I-I-I-I-XXXXXXXXXX

Solution:

Have a weighting system that promotes diversity in maps' difficulty, song length, and song type. For example, if a certain song has around 3 instances of insane difficulties then this system should make it harder to rank another insane difficulty. To make it harder to rank, the system will raise the amount of stars needed to be nominated by BN. This weight would apply mapset wide as well as globally, where it would take only 1 X diff to raise the requirement by a lot while adding an X diff on a global scale makes it harder a bit. So weighted = mapset+global.


Problem:

I'm just afraid people will bother with lower difficulties even less, which is such a shame

Solution:

Manually shift more weight on harder diffs and less weight on easier diffs. This way it would require exponentially more effort to rank harder maps. I think the manually set weights should reflect the odds of a diff being ranked. For example:

  • Easy: 0.1
    Normal: 1.0
    Hard: 5.0
    Extreme: 3.5
Then this manual set weights would be applied to the automatic weight system to determine how many stars is needed to allow the map to be qualified


Problem:

We don't need speedranks
More time spent into it means more quality

Solution:

There is a difference between submitting a map and ranking it the same day with almost no useful mods, and submitting a map and ranking it one week later but with a couple of pages of mods (because the mapper was able to gather mods quickly).
Each difficulty would still go through the whole qualification process the same way it currently is. The only difference is that you can start qualifying difficulties as they start to pass the quality requirements.


Problem:

Ranking by difficulty can make a mapset have only one difficulty, if one difficulty gets qualified and the rest gets abandoned. This completely goes against the ranking criteria.

Solution:

Ranking criteria will need to change.

Problem:

Expect a lot of single diff mapsets

Solution:

With a good search engine and filters and non-crappy osu!direct, I don't see a problem. Players download the diffs they want and dont need to worry diffs they don't want taking up their hardrive space. Also faster download times if map is without vid, sb, and/or skin... not terribly faster, but at least it's an optimization of some kind. Ofc this doesn't help in the long run if there multiple single diffs with their own resources.

Problem:

This will undermine easy diffs

Solution:

So forcing people to download diffs they don't want is the correct thing to do? If a player wants to play easy, all they need to do is download it.


Problem:

New mappers will don't get notice because ¨Who is him lol¨

Solution:

Under my per-song, per-mapper mapset model, guest diffs will be an important tool. If a popular mapper decides to rank just one diff, then a non-name mapper can pitch in to provide a guest diff.


Problem:

how you will handle same hitsound sets, storyboard per diff, video per diff, background per diff?

Solution:

No clear solution yet. Will need to be talked about :c
Last edited by abraker , edited 2 times in total.
Kibbleru
osu! Alumni
5,846 posts
Joined August 2013
Kibbleru
on paper this seems like a great idea. however, if something like this is implemented. we would just be expecting spreads like
E-N-H-I-I-I-I-XXXXXXXXXX

lol see the issue?

also think of the filesize each set would have lol especially on the popular songs like miiro
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Kibbleru wrote:

on paper this seems like a great idea. however, if something like this is implemented. we would just be expecting spreads like
E-N-H-I-I-I-I-XXXXXXXXXX

lol see the issue?

abraker wrote:

Make a system that promotes diversity in the beatmap pool. A system that can put weight on the ranking requirements if there is too much of something and make it easy to rank if there is a lack of something. The system should weigh by difficulty and length per song. If a certain song has around 3 instances of insane difficulties then this system should make it harder to rank another insane difficulty. To make it harder to rank, the system will raise the amount of stars needed to be nominated by BN. Oh look at that, stars can actually have value in this system!
Bara-
11,978 posts
Joined April 2013
Bara-
I'm just afraid people will bother with lower difficulties even less, which is such a shame
Mapping Easies/Normals is imo even more fun than mapping extras
Sylvette
332 posts
Joined July 2015
Sylvette
I'm going to be honest about this, I hate this idea. Don't fix something which has nothing wrong with it.
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Cirno-baka9 wrote:

I'm going to be honest about this, I hate this idea. Don't fix something which has nothing wrong with it.
Well one of the things wrong currently is the amount of time it takes to rank something. Ofc dedicated mappers don't see an issue with the current system because they have enough time to do this stuff, but any common player would dedicate a bit of time before giving up. If you would be able to focus on just one diff at a time, maps should be able to get ranked faster and mappers don't have to spend as much time mapping a full spread. We need to encourage people to map, not give them requirements that would take a month to complete. No average player has that sort of dedication, not even me, who still needs to find time to fix up patterns in half of the maps in the mapset.
Bara-
11,978 posts
Joined April 2013
Bara-
We don't need speedranks
More time spent into it means more quality
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Bara- wrote:

We don't need speedranks
More time spent into it means more quality
The time spent on entire mapset can be spent on one difficulty. That means more dedication towards that one difficulty. Also speedranks by your definition happens when a mapset gets ranked too fast. What would be considered speedranking for a mapset wouldn't necessarily be the same for one diff.
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic
Im glad we could get a discussion about this ;D

@abraker: glad that at least someone agrees with me. And yes, if this were implemented, many things will have to change too, like in order for this to be effective, beatmaps will have to be downloaded via client too (anyways, the current situation where it allows us to download via website just promotes song's distribution, but lets not talk about this for now)

Kibbleru wrote:

on paper this seems like a great idea. however, if something like this is implemented. we would just be expecting spreads like
E-N-H-I-I-I-I-XXXXXXXXXX

lol see the issue?

also think of the filesize each set would have lol especially on the popular songs like miiro
How is E-N-H-I-I-I-I-XXXXXXXXXX any different from current situation when we merge certain songs? You may add a few E's and N's, but most are probably forced difficulties just to comply current rules. Here's the problem, the system is forcing people

Also, the total filesize will actually shrink because youll only require ONE mp3 file for all of those difficulties (unless there are different versions of it, which then it should be submitted separately)

Bara- wrote:

I'm just afraid people will bother with lower difficulties even less, which is such a shame
Mapping Easies/Normals is imo even more fun than mapping extras
There's nothing stopping mappers that prefers to map easy and normals to submit their difficulties. In fact, I believe that this will incentive more mappers like you that prefers to map these kind of difficulties, just find a song you like, and if you see that an easier difficulty could be added, just map it and submit it and wait for qualification!

And yes, I still agree that easier diffs needs to be added, thus each song should have at least E+N or N+H before harder diffs can be approved.

Cirno-baka9 wrote:

I'm going to be honest about this, I hate this idea. Don't fix something which has nothing wrong with it.
No, this idea is to fix what is wrong with the current situation. Basically, what is wrong right now is that the system is inconsistent with how the community needs and behaves. Putting it into another perspective, the reason for which Apple has become so successful its because it was able to cater products based on what the user needs and how they behave, they did not force the users to use their products (thus the doom of many brands, specially sony with its mini cd stuff)

Main argument about this should be to explain why the current system does actually caters to the community while my idea does not. I have already given some arguments in my original post.

Bara- wrote:

We don't need speedranks
More time spent into it means more quality
Each difficulty would still go through the whole qualification process the same way it currently is. The only difference is that you can start qualifying difficulties as they start to pass the quality requirements. For example, I have made E N H I X diffs for a song, due to the more simple aspects of E and N difficulties, they won't require such a deep mod procedure for them to pass quality standards so they can be qualified first, and then the other modders can start focusing more on the harder difficulties which is where they usually require the most attention (yes, I am talking about the extra hard ones).

A phenomena that has happened to certain beatmaps, is that if they even possess that "odd" difficulty that no one agrees with, it will also drag all the other diffs which are perfectly fine to graveyard, making all that effort go to waste. Treating them separately will allow those good difficulties to be qualified, and kill that odd difficulty (or postpone it until the mapper finally decides to make the appropriate changes)
ferret irl
212 posts
Joined July 2015
ferret irl
The way mapsets already work is good. No need to change it. Like what Cirno said,
Don't fix something which has nothing wrong with it.
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Chewy-san wrote:

The way mapsets already work is good. No need to change it. Like what Cirno said,
Don't fix something which has nothing wrong with it.
You are failing to see what is wrong with it. I and Xanandra both stated what is wrong already. Try to prove that the things we stated are not flaws within how mapsets currently work.
Zexous
713 posts
Joined July 2012
Zexous

Xanandra wrote:

Kibbleru wrote:

on paper this seems like a great idea. however, if something like this is implemented. we would just be expecting spreads like
E-N-H-I-I-I-I-XXXXXXXXXX

lol see the issue?

also think of the filesize each set would have lol especially on the popular songs like miiro
How is E-N-H-I-I-I-I-XXXXXXXXXX any different from current situation when we merge certain songs? You may add a few E's and N's, but most are probably forced difficulties just to comply current rules. Here's the problem, the system is forcing people

Also, the total filesize will actually shrink because youll only require ONE mp3 file for all of those difficulties (unless there are different versions of it, which then it should be submitted separately)
Consider this: mapping lowdiffs is not popular in the meta right now. The ONLY reason the majority of the time that lowdiffs are mapped is PURELY because they are required. If they are not required........no lowdiffs. Sure, you will get people who map lowdiffs because they enjoy them and don't have to map highdiffs anymore...but those people are VERY few. They're forced difficulties, true, but they're at least there.

And the filesize, consider this: hitsounds. You have two alternatives in your idea: either download one full set with every diff, or download the song and then download each diff individually. There are problems with both.
For the first: you have to download every diff, and with all their hitsounds, because who mandates what the hitsounds of the set are? No one, that makes no sense, so each mapper would use their own hitsounds, but that means the map has to have EVERY hitsound EVERY mapper used. This even includes duplicates, since there's no way for BSS to know. There's another issue here too: file naming. Mapper A has normal-hitclap. Mapper B has normal-hitclap. What then? You could separate into subfolders, I guess. But the filesize issue remains. There's also the issue of storyboards, that I see you acknowledged. Storyboards are MASSIVE typically. How do you accommodate each of them? Have them be downloaded individually? But then that defeats the entire point of your system and works better on individual map system anyway.
For the second: if your system runs on a "download base song and then individually add each diff" system, this defeats the purpose of your system because unknown mappers will be ignored even more. Further, the hitsounds/storyboard issue remains, so you'd have to download the hitsounds/storyboard individually with the diff....which defeats the purpose as well.

Either way your system works, it would be an absolute mess to try and sort through what could well be 100+ difficulties on a song. On the current system, it's much easier to find a map (for example, "the miiro by wkyik" is found easily, rather than searching miiro and looking through all the difficulties for wkyik's).

Another point to consider: this would decrease playing of lowdiffs. Many decently skilled players sometimes play the lowdiffs because they're there, they come with the package. If they're no longer a package deal, why bother to download them?

Xanandra wrote:

Bara- wrote:

I'm just afraid people will bother with lower difficulties even less, which is such a shame
Mapping Easies/Normals is imo even more fun than mapping extras
There's nothing stopping mappers that prefers to map easy and normals to submit their difficulties. In fact, I believe that this will incentive more mappers like you that prefers to map these kind of difficulties, just find a song you like, and if you see that an easier difficulty could be added, just map it and submit it and wait for qualification!

And yes, I still agree that easier diffs needs to be added, thus each song should have at least E+N or N+H before harder diffs can be approved.
This system means: "map whatever you want, it's ok", but this implies: "no matter what you map, all difficulties will eventually be covered". So no mapper would bother to map a lowdiff just because they see none is there already, because they believe the lowdiff will be there eventually. This narrows the pool of lowdiffs down to mappers who actually enjoy mapping lowdiffs, which I doubt I have to clarify is abysmally tiny and dreadfully insufficient for the mapping community as a whole.

Xanandra wrote:

Bara- wrote:

We don't need speedranks
More time spent into it means more quality
Each difficulty would still go through the whole qualification process the same way it currently is. The only difference is that you can start qualifying difficulties as they start to pass the quality requirements. For example, I have made E N H I X diffs for a song, due to the more simple aspects of E and N difficulties, they won't require such a deep mod procedure for them to pass quality standards so they can be qualified first, and then the other modders can start focusing more on the harder difficulties which is where they usually require the most attention (yes, I am talking about the extra hard ones).

A phenomena that has happened to certain beatmaps, is that if they even possess that "odd" difficulty that no one agrees with, it will also drag all the other diffs which are perfectly fine to graveyard, making all that effort go to waste. Treating them separately will allow those good difficulties to be qualified, and kill that odd difficulty (or postpone it until the mapper finally decides to make the appropriate changes)
The issue with this: is there a cut off, or is the mapset perpetually growing? For example, let's say the map is E N H I. All diffs are modded, and now it's ranked. But under your system, all diffs go to the same mapset if it's the same song. So let's say someone adds an X. Now you have ranked E N H I and unranked X. Do you push the X individually for rank? Under your system, yes. But people who downloaded it after the initial rank only have the initial E N H I. Now let's suppose it's a popular map and after the initial E N H I, there's an additional H H I I X X X. Let's say these do not happen at the same time. How many times do you expect people to redownload if they want all these maps? And since everyone takes different amounts of time to push through the ranking process, it would be very reasonable to expect separate ranking timeframes for each of those diffs. So do you expect people to redownload the map 7 times? Or even more? To redownload the map every time a new diff is ranked? And since your system bundles all the maps on one set, that means redownloading the whole massive mess every time, which is of course super wasteful.



Some final words about this system: this would encounter a massive issue of deindividuation. Much of people's drive for ranking is the personal investment in their own map because of the personal reward they'll receive. If you make every map a communal effort...you could think of it like internet communism. Except communism only works if everyone has a large personal stake in the outcome, and this is the internet...so that won't happen. With such a communal mapset, there's no personal recognition to be had by ranking the set. The only personal reward is from having your difficulty ranked. But this means that as soon as a mapper has submitted their difficulty and fixed its modding, they're done. No more need to care. If I submit an Extra, I only have to care about the extra. I don't care about getting mods, I don't care about getting BNs. There's no mapset owner, which means there's no one there to push the set. If you want to say something like "the community will push it forward"....yeahhhhh, good luck with that.

By the way, this system invalidates the only really effective method of unknown mappers getting out there, which is GDing. So yeah.
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic
@Hitsounds: these are not things to be bothered currently. This is just the conceptualization of an idea, and they can be overcome in the development process. But an early conceptualization from me, would be that when the difficulty is going to be qualified, all of its used hitsounds will be added to the song's hitsounds slots, thus is later submitted difficulties also want to reuse them, then the new difficulties could reuse them, and if new hitsounds are qualified, then they just keeps added to the slots list. Not sure if I am explaining myself well.

@Storyboards: Yes this will be an issue, but an early conceptualization of how to deal with it that I have in my mind is to treat it as a ''Difficulty'' where the player can download and annex to the song. For cases where SBer is different from the submitter, they can also be properly addressed as the author of it!

Zexous wrote:

Consider this: mapping lowdiffs is not popular in the meta right now. The ONLY reason the majority of the time that lowdiffs are mapped is PURELY because they are required. If they are not required........no lowdiffs. Sure, you will get people who map lowdiffs because they enjoy them and don't have to map highdiffs anymore...but those people are VERY few. They're forced difficulties, true, but they're at least there.

Another point to consider: this would decrease playing of lowdiffs. Many decently skilled players sometimes play the lowdiffs because they're there, they come with the package. If they're no longer a package deal, why bother to download them?
Very valid reasons, and I understand that. Just as there are pros, there are cons, no need to keep stick into one place, Maybe after some more discussion Ill add a Cons section in my original post.

Zexous wrote:

For the second: if your system runs on a "download base song and then individually add each diff" system, this defeats the purpose of your system because unknown mappers will be ignored even more. Further, the hitsounds/storyboard issue remains, so you'd have to download the hitsounds/storyboard individually with the diff....which defeats the purpose as well.
First of all, Im not exactly sure what you think what my main purpose with this idea is. As stated previously, it is to better accommodate what the current mapping community needs.

Also, you can also think at it as... if there's already an additional difficulty for it, and I love the song, why not? Let's download it and try it!.
Nevertheless, this is something that I dont believe to be able to be determined unless we actually try it out and see the outcomes.

Zexous wrote:

Either way your system works, it would be an absolute mess to try and sort through what could well be 100+ difficulties on a song. On the current system, it's much easier to find a map (for example, "the miiro by wkyik" is found easily, rather than searching miiro and looking through all the difficulties for wkyik's).
I don't see how it is any different from current situation. With this, you could say that all diffs are GD's

Zexous wrote:

This system means: "map whatever you want, it's ok", but this implies: "no matter what you map, all difficulties will eventually be covered". So no mapper would bother to map a lowdiff just because they see none is there already, because they believe the lowdiff will be there eventually. This narrows the pool of lowdiffs down to mappers who actually enjoy mapping lowdiffs, which I doubt I have to clarify is abysmally tiny and dreadfully insufficient for the mapping community as a whole.
I agree, which is why I propose that in order for harder difficulties to be qualified, qualified easier difficulties must exist first

Zexous wrote:

The issue with this: is there a cut off, or is the mapset perpetually growing? For example, let's say the map is E N H I. All diffs are modded, and now it's ranked. But under your system, all diffs go to the same mapset if it's the same song. So let's say someone adds an X. Now you have ranked E N H I and unranked X. Do you push the X individually for rank? Under your system, yes. But people who downloaded it after the initial rank only have the initial E N H I. Now let's suppose it's a popular map and after the initial E N H I, there's an additional H H I I X X X. Let's say these do not happen at the same time. How many times do you expect people to redownload if they want all these maps? And since everyone takes different amounts of time to push through the ranking process, it would be very reasonable to expect separate ranking timeframes for each of those diffs. So do you expect people to redownload the map 7 times? Or even more? To redownload the map every time a new diff is ranked? And since your system bundles all the maps on one set, that means redownloading the whole massive mess every time, which is of course super wasteful.
I understand the issue, which is why in order for this to be effective, the downloads have to be handled from the client, where you first have to download the qualified "song" first, and then you have the ability to annex the difficulties as they become available to avoid the need of redownloading them all.

And yes, this is a drastic change, and it will also even make osu! support pointless, making this feature even less unlikely to happen.

Maybe the developers are smart enough to come up with another way ;)

Zexous wrote:

Some final words about this system: this would encounter a massive issue of deindividuation. Much of people's drive for ranking is the personal investment in their own map because of the personal reward they'll receive. If you make every map a communal effort...you could think of it like internet communism. Except communism only works if everyone has a large personal stake in the outcome, and this is the internet...so that won't happen. With such a communal mapset, there's no personal recognition to be had by ranking the set. The only personal reward is from having your difficulty ranked. But this means that as soon as a mapper has submitted their difficulty and fixed its modding, they're done. No more need to care. If I submit an Extra, I only have to care about the extra. I don't care about getting mods, I don't care about getting BNs. There's no mapset owner, which means there's no one there to push the set. If you want to say something like "the community will push it forward"....yeahhhhh, good luck with that.

By the way, this system invalidates the only really effective method of unknown mappers getting out there, which is GDing. So yeah.
The main idea is to individualize even more each difficulty. It is stupid to have the Creator tag of a Guest difficulty to another user. I dont know where you got the idea of communism here. If anything, the submitter is taking the EXTRA WORK by finding mods for the GDs he added, and if it is a m4m, then the submitter will also have to take the weight of it, thus making the current process even more communal.

Thus, there is no MAPSET ownership, but there is the proper DIFFICULTY ownership

And yes, if you submit an extra, then you have to take care of that extra diff that you added by yourself, by finding by yourself your own mods for it, as well as BNs, etc. Ranking process will have to be handled individually except probably the first time where the two easiest diff have to be submitted.

EDIT: Also, I dont believe that currently there exists the idea of mapset ownership, in order for this to happen, every difficulty within the mapset has to be made by a single creator, and those are rare.
Kibbleru
osu! Alumni
5,846 posts
Joined August 2013
Kibbleru
my point is you need to force people to map lower diffs otherwise there would lack diversity in that area.
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Kibbleru wrote:

my point is you need to force people to map lower diffs otherwise there would lack diversity in that area.
Didn't I already reply to this with the idea to make a system that promotes map diversity?
Kibbleru
osu! Alumni
5,846 posts
Joined August 2013
Kibbleru

abraker wrote:

Kibbleru wrote:

my point is you need to force people to map lower diffs otherwise there would lack diversity in that area.
Didn't I already reply to this with the idea to make a system that promotes map diversity?
map diversity? impossible with the mindsets of the current mappers out there xd

insane + or nothing x d

unless ur literally willing to reward ppl for mapping low diff
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic

Kibbleru wrote:

map diversity? impossible with the mindsets of the current mappers out there xd

insane + or nothing x d

unless ur literally willing to reward ppl for mapping low diff
Yes yes, I agree with you, but this is not important at this stage.

To prevent this kind of occurrences, one could set that if the song already has three difficulties of insane or harder, at least 3 diffs of easy or normal must be already qualified before qualifying more insanes+.

Although this will also raise an issue when there are several insanes queuing up for qualification, which is gonna be qualified first? Etc. But this is not important at this moment of discussion.

Mind you that if this were to be implemented (for which I am 99.99999999999999% sure it wont), it would require recoding almost the whole osu code, thus this thread should be treated more as a mere conceptualization, these details are not important. Whats more important currently its to discuss if whether you guys agree that osu! Should be developed in that direction.

Anyways, my point is that these kind of issues have solution, just in the same way that ranking criteria and the osu! Code has evolved over time to fix these issues. Just that what I am trying to point out is that the direction taken to solve them is not the best for current community. What I really want is to abolish the idea of mapsets as we currently understand then.
Endaris
5,282 posts
Joined June 2010
Endaris
huh huh
*drops related thread*
p/2078510/
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic
Yes, very related actually, just that I am suggesting a more drastic change to really consolidate the issue.

Well, feel free to merge topics if deemed appropriate

(At least that less drastic proposal has much more acceptance than mine)
Endaris
5,282 posts
Joined June 2010
Endaris
Doesn't make sense to merge them but I think it's nice to know that request as well when looking at this one.
pkhg
2,507 posts
Joined May 2010
pkhg
kibbleru is like "stfu i dont want this to be changed cuz i dont like changes"
like me whe ppy changed hidden mod

this seems to be a nice idea. imagine adding diffs to old maps, there are a lot songs that i like but im lazy and making a entire mapset and getting it ranked takes too long
i can already see in #announce something like "a new difficulty has been added to [maplink]" xd

support~
Sylvette
332 posts
Joined July 2015
Sylvette

Xanandra wrote:

No, this idea is to fix what is wrong with the current situation. Basically, what is wrong right now is that the system is inconsistent with how the community needs and behaves. Putting it into another perspective, the reason for which Apple has become so successful its because it was able to cater products based on what the user needs and how they behave, they did not force the users to use their products (thus the doom of many brands, specially sony with its mini cd stuff)
If you pay attention to the community, rather than your own two wishes, then you'll realize that few people want this. Most people might think this idea sounds good on paper, but if you think about it, it's really a terrible idea.

Xanandra wrote:

Main argument about this should be to explain why the current system does actually caters to the community while my idea does not. I have already given some arguments in my original post.
  1. Everyone has already gotten used to this original way of ranking.
  2. Ranking by difficulty can make a mapset have only one difficulty, if one difficulty gets qualified and the rest gets abandoned. This completely goes against the ranking criteria.
  3. As most people here have said, the idea is terrible. If your idea "makes the ranking process consistent with what the community needs" then why do few people in the community like this idea?
  4. Ranking process now is simple and easy, your ideas can make it way more confusing than it should be.
Only listed a few reasons because I'm lazy, and need to leave soon.

Xanandra wrote:

Bara- wrote:

We don't need speedranks
More time spent into it means more quality
Each difficulty would still go through the whole qualification process the same way it currently is. The only difference is that you can start qualifying difficulties as they start to pass the quality requirements. For example, I have made E N H I X diffs for a song, due to the more simple aspects of E and N difficulties, they won't require such a deep mod procedure for them to pass quality standards so they can be qualified first, and then the other modders can start focusing more on the harder difficulties which is where they usually require the most attention (yes, I am talking about the extra hard ones).
Again, something that sounds good on paper, but is actually a bad idea. This idea does enforce speedranks, and no matter how you try to say it doesn't, it does.

Xanandra wrote:

A phenomena that has happened to certain beatmaps, is that if they even possess that "odd" difficulty that no one agrees with, it will also drag all the other diffs which are perfectly fine to graveyard, making all that effort go to waste. Treating them separately will allow those good difficulties to be qualified, and kill that odd difficulty (or postpone it until the mapper finally decides to make the appropriate changes)
Except, it's the mapper's fault for that difficulty. It doesn't matter if all the other difficulties are good.

Would argue a bit more but I need to leave.
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic
EDIT:

Was trying to edit original post, but I edited this post instead, and now the contents of this post is lost.... I fail.

If it is still recoverable, please do so mods.
Last edited by Luvdic , edited 5 times in total.
Sylvette
332 posts
Joined July 2015
Sylvette

Xanandra wrote:

t/451293 ; t/115773 -> features request would become obsolete as these issues wouldn't exist if they were treated as per difficulty. Now, I dare you to give me arguments why we should not implement these features.
Would just like to point out that the examples you gave are completely unrelated to this topic, other than the fact that they deal with seperate difficulties.

I'm not going to argue anymore, as this is completely a matter of opinion, and from my experience, arguments about opinions always end up horribly.
Seijiro
osu! Alumni
3,438 posts
Joined April 2013
Seijiro
The examples provided by the OP convinced me tho and it looks like a cool idea for how the community tends to map nowadays.

What I hate about the comments here is the misinformation: using "everyone" is something really bald, considering nobody in the world can be friends with everyone and know what they like/think, so comments like "everyone is accustomed to the current ranking process" are just lies.
Unless you can prove you asked every mapper in the community, that is.

Another misconception I read in here is "speedranking".
Idk what you call speedranking, but I can ensure you that nowadays this concept doesn't exist as in the past, maybe it even vanished (but I don't check every map so let's say it is less than before).
There is a difference between submitting a map and ranking it the same day with almost no useful mods, and submitting a map and ranking it one week later but with a couple of pages of mods (because the mapper was able to gather mods quickly).



Back to the topic: as I already said, this concept looks cool and the few problems that may exist (storyboards and hitsounds overwriting) can be fixed somehow in any case.
Nakano Itsuki
1,218 posts
Joined July 2014
Nakano Itsuki
Originally I was against this type of stuff, but after thinking for a while I see which direction you are headed.
The post seems a bit dizzy to me though, so is the system you proposed like,

1. A list of songs on the website
2. Mappers submit diffs to it, aka a single page of the song with multiple diffs being able to be submitted
3. Players choose what diffs to download

Something like that I guess?

(Dont try to convince me sliders need accuracy though, I'm extremely against scorev2 for this reason)
Sylvette
332 posts
Joined July 2015
Sylvette

StarrStyx wrote:

The post seems a bit dizzy to me though, so is the system you proposed like,

1. A list of songs on the website
2. Mappers submit diffs to it, aka a single page of the song with multiple diffs being able to be submitted
3. Players choose what diffs to download

Something like that I guess?
Nothing on the list you gave is what they are proposing. They want a system which allows qualification per difficulty, so, one difficulty can get ranked while the rest of the mapset stays unranked.
Nakano Itsuki
1,218 posts
Joined July 2014
Nakano Itsuki
Thats what i was thinking as well.

And thats what I do think is a good idea: the idea of ppl submitting individual diffs and ranking them on their own.

Doesnt seem too bad tbh, in fact I like the idea.
I doubt this would be the final form tho, since theres obviously another sort of system being cooked up by Loctav; but I'll just see how things go.

(And i still hate scorev2)
Luvdic
2,113 posts
Joined July 2009
Topic Starter
Luvdic
What I wish to see implemented is exactly this:

Destroy the idea of the existence of mapsets. The idea of a "set" will still exist, but instead of treating it in a per "mapset" basis, do it in a per song basis.

What stated by you both is just a consequence of this.

Also, made first post easier to look at and added some more stuff, but originally edited the wrong post... so now that post is gone forever...
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker
I think I will make the second post easier to read as well and have it behave like an addendum to your post by adressing some conserns players brought up, my solutions (if any), and additional features your post doesn't mention.
Natsu
8,076 posts
Joined September 2012
Natsu
To be honest I don't know what kind of mapper would want this to be implemented, since it basically kill new mappers, old mappers work, mapping quality, diversity of songs and creativity.

  1. New mappers will don't get notice because ¨Who is him lol¨, this will tend to new mappers to do ridiculous hard diffs to be notice which lower the mapping quality and their possibilities to rank a map.
  2. Old mappers will lose publicity of their diffs, because of the ridiculous amount of diffs added to the set.
  3. Song variety will die, with the actual system we get like 7 beatmaps of the same song, with what you propose we gonna have 50+ diffs of the same song being all Extra or Insane diffs.
  4. Another reason why song variety will die, is because with the current ¨beatmaps sets¨ people tend to lose motivation of mapping the same song, you will just increase it and the game will become more anime related than what actually is, we gonna have sets of 4 diffs for english songs and sets of 60 diffs for anime maps.
  5. When I map, I want my work to be shared and played for almost every player, this basically kill my purpose of mapping, because my diff, storyboard, hitsounds, videos or anything else, will be rarely notice.
  6. You also will kill Easy, Normal, Hard and Insane diffs, since no one (maybe a minority of mappers) like to map that diffs, everyone will be mapping Extras and Ultras. This is a fact E-N-H-I-XXXXXXXXXXXXXXXXXXXX
  7. This is also unfair, because the first mapper need to map a full set (no one will map Normal and Hard alone, unless low bpm, let's be realistic)
  8. Mapping quality will decrease, since everyone making s**t mapping to get attention :p, the new meta ¨wait for someone to map a set and lets add s**t diff to it
  9. As a beatmap owner I visualize what the song provide me and I make a mental set of what diffs I'll add, so for example: https://osu.ppy.sh/s/428156 I don't want insane or extra diffs, there because I believe the song don't need them. If someone else map their own set then is their problem.
  10. As a GD mapper, I provide what the mapper owner request me, because of the previous one.
  11. This one maybe sounds egoistical, but I think I'm not the only one who feels like that, I want my Own set, I don't want to share it with multiple people, I map to have my creations alone, I add custom BGs, I care about hitsounds etc, because is my creation, having it open to public diffs kill that tbh.
  12. I think the staff with the ranking criteria friends are still doing something with : t/420223&start=0 so basically your idea also kill the spreads lol, since we gonna have (SR just an example) 1,9 normal , 2,9 hard - 3,9 insane and 6 stars + stuff being spammed :D, iiirc what they want to do is kill maps with multiple insanes and extra diffs, and your idea is totally the contrary.
  13. Also, how you will handle same hitsound sets, storyboard per diff, video per diff, background per diff?,
  14. each time a diff is added you will restart the map to the first page in the ranked section? if so then we gonna have miiroo, highscore and no tittle forever on the first page?
  15. if the song have multiple bpm and the first one get ranked and we found that the map have wrong timing, then we gonna just add more diff to the wrong timed map?
  16. If a song get copyright?
  17. what if I get a permission from the artist to map his song, but the artist don't want other people to map it?
  18. Beatmap contest?
Per diff creator is a cool idea that should be added, adding diffs to the same set shouldn't happen IMO. sorry for bad english lel

oh also forgot to add that people will based their mapping on existing ones to have a ¨safe map¨, which means you gonna get a lot of similar diffs in the same set.!!
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker
Unlike Xanandra, I dont believe making a "hive set" on a per-song is a good idea either. I still think mapsets should be per-mapper AND per-song. This would potentially kill the concept of guest diff, however. I do see guest diffs helpful as a way to promote no-name mappers, so it's not a good idea to kill that. To solve the problem, I suggest some sort feature which would allow to "overlap" specific difficulties of multiple mapsets. It would work like a shortcut. The guest diff will appear in the another mapper's mapset just as a shortcut, where the actual diff would be located in the guest mapper's mapset. Here is what I mean visually:

In this example, guest mapper's Insane would reside in the Guest mapper's mapset, while being linked as a shortcut to other mapper's mapsets. Also additions to multiple mapsets would be allowed. To prevent abuse such as using a diff in too many mapsets, the number of times the guest mapper's map is used in other mapsets should be limited to a certain amount, and disallow adding guest diffs which are already ranked (doesn't mean that already an added guest diff can't be used anymore, however).


One issue I see with this is the reoccurance of the same diff in multiple mapsets. Perhaps allow the website and osu!direct to not display and not download the guest diffs if you already have a score on them. That is, only diffs belonging to the the actual mapset would be displayed and downloaded if you have scores on all the guest diffs.
Kathex
200 posts
Joined September 2014
Kathex
And if instead suggest a big change like that, just suggest the change of how you send and store in your pc and osu! servers a audio file to reduce the memory usage?

Like: When you go to map editor it offer to you a list where you can first search all mapsets, then you can download just the Audio file with a individual identification for every audio file. If you dont find the song that you want, then you can upload just the audio file with only the name and autor (maybe tags too to make easy find the song). Then when you upload the song file to osu! servers it will receive a audio file ID, then you can just take this ID and use it to adress your map. Then more mappers can download the same audio file just using the ID when they are going to make a new mapset.
It would reduce the storage size for osu! servers and could be implemented to client version by donwloading audio files in a folder separate of mapsets files, like storyboard, difficulties and hitsounds.

I dont will create the problems with actual serching system and each mapper will do their mapset separated of others.

taking advantage of the momentum, if they do a update like that, i suggest too change the limit of maps that you can upload, like: You still only can upload 3 audiofiles by month. But if you use a already uploaded audio file ID, u can submmit more mapsets like 10 or more. Cause mapsets files without the audio file, are less than 1MB, just if you using a png background, hitsounds and a heavy sb, it can reach like ~5MB ...

Im crazy? :?:
abraker
Global Moderator
8,327 posts
Joined July 2014
abraker

Kathex wrote:

...
This idea quickly crumbles when you consider osz v2.

Edit:
I dont remember the details, but it had something to do with making mp3 files something something... I dont really remem, honestly, but it did sound like it wasnt as modular as you make it be
Last edited by abraker , edited 3 times in total.
Kathex
200 posts
Joined September 2014
Kathex

abraker wrote:

Kathex wrote:

...
This idea quickly crumbles when you consider osz v2
Well i dont know what to expect about osz2
Please sign in to reply.

New reply

0
1 / 36