Okoratu wrote:

i know im saying i'd prefer that over the possibility that there's even more stuff that is inaccessible to new players trying to get into the game or the average player

Ah fair enough. I read that wrong then.

I personally always found the 5 min limit for marathon kind of low, and would really prefer it to be 6 min, with 2 diff requirements for 5 min and 3 diff requirement for 4 min or something, with proper spread rules along those lines, but I doubt that's ever possible anymore nowadays after the requirement was lowered.

Okay, after reading this thread, I've changed my stance.


First, can we get some actual numbers? I want to see how many people

1. Play <2 minute normal and hard diffs
2. Play 4 minute+ normal and hard diffs
   
   And also see how many:
3. Ranked 1:30-3:00 minute songs there are annually
4. Ranked 3 minute + sets there are annually

Even without the numbers, I have to agree with CDFA, where a good compromise would be diff specific drain times. Either something like "if your song's length is above 3:30, your Hard diff must be mapped up to at least 3:30 of drain time, while your Normal diff must be mapped up to 2:30 of drain time."

I'm not sure what to make the specific wording to prevent abuse while also preventing people from being able to map songs with a lot of unmappable sections where you'd put a break, but the idea is that lower diffs would have to be mapped to a lesser percentage of the mp3 than higher diffs.

This makes sense to me, as now there would be content for the newer players, because they will have a map to play in their skill range, the mappers would be much happier mapping a shorter map for low diffs they probably don't care for much in the first place (hopefully with lower drain requirements they'd put quality > quantity but that's wishful thinking), while also opening up the ranked section to more 4 minute songs.

(Also I agree with CXu, five minute songs should have a minimum 2 diff requirement if your diff is above an Insane)

Completely 100% in agreement with UndeadCapulet on this one.

Instead of writing some paragraphs about why I think it’s a good idea with pros/cons listed for both sides, I’ll just sum it up and say that I’d rather see there be content for some players (HIX or IX spreads in this case) than no content for everyone.

since the majority of the player base plays diffs below insane i dont think ie spreads should ever be a thing. therefore i really like icekalts proposal

I think in a way this benefits more the mapper, modder and the BN than the player in a way that it makes it easier for everyone just to have a smaller spread and less things to check. The true thing that i believe its affecting this is what Capulet said, i don't think someone wants to check a 4:30 min, 4 diff spread map, unless you're really dedicated to it or you know that person (you'll be modding around 18 mins of drain if you mod each difficulty).

Even though i do believe if any of this would be applied to modern day mapping and modding, osu! will experience a crash in what would be intake of new players, and i'm not only talking about standard or mania, every game mode will feel that and hard; yes maybe its true that mapping low difficulties can be tedious and boring but it opens a wide range in what would be player intake, plus why wouldn't you want to give the same chance to someone "new" rather than an already decent player?

Low difficulties around each game mode provides that "chance" I'm talking about, maybe to play their favorite song while not choosing something they cant play, improving, trying out new skins and none the less, having fun.

TL;DR

I do think this will benefit everyone involved in the creation and ranking process of a map but will damage any low player for sure, maybe something else can be implemented to make this beneficial for the modder and mapper as well for the player.

pishifat wrote:

if i were running osu and had i an interest in expanding its audience, i'd think this kind of change isnt very good. out of ranked maps selection, there would be a smaller fraction of content available for lowest level of players, alienating the potential future of hte game or something along those lines. the idea that new players are already alienated with 5+min songs, so being lenient with other lengths should be fine sounds more like a reason to remove approval from game dev perspective (which i'm not interested in doing).

But back in the days maps even shorter than 5min were approved (like The Big Black). Didn't really hurt the game's future now did it? More like the opposite considering how popular and iconic it became. Imagine having to map a full spread for that song. lol

Anyway, I do think that doing something about this is a good idea, rather than leaving it as it is with .mp3 extensions becoming ever more popular.

What might indeed seem as a problem is that as a new player, I would hate it if I couldn't play, say, my favourite pop or metal song because it's like 3:30 in length and has no easy/normal diffs available for it.

However, as was said before, most new players don't like playing very long maps right off the bat. If you think about it, the most popular genres in this community are things like anime and electronic music, not pop and metal really, they're more niche.

If there is a pop or metal song they'd really like to play but can't quite reach because the lowest diff is a hard, it should motivate them to improve their skills in the game enough to play it. If the only genres of music they listen to are niche in this game, and they're completely unable to tolerate playing easier maps of genres they don't really listen to, I don't think they'd last as a community member for very long to begin with.

With that in mind, and as much as it feels weird to let go of these spread rules I've gotten used to over the years, something needs to be done, and as long as nobody comes up with a better alternative, I'm supporting this one.

1. if the drain time is <4:00 your set's lowest diff must be normal or lower
2. if the drain time is 4:00-5:00 your set's lowest diff must be hard or lower
3. >5:00 can be anything

I'm thinking more like this. since H is way more accessible than N. IX only is just too small imo. Since we are having problems with 4:30 maps, we don't need to do anything with anything below 4:00

agree with ancul, insane as the lowest diff sounds to me like too much for something non marathon, but otherwise it's a neat idea

AncuL wrote:

1. if the drain time is <4:00 your set's lowest diff must be normal or lower
2. if the drain time is 4:00-5:00 your set's lowest diff must be hard or lower
3. >5:00 can be anything

I'm thinking more like this. since H is way more accessible than N. IX only is just too small imo. Since we are having problems with 4:30 maps, we don't need to do anything with anything below 4:00

yeah this seems the most reasonable at least from the suggestions

AncuL wrote:

1. if the drain time is <4:00 your set's lowest diff must be normal or lower
2. if the drain time is 4:00-5:00 your set's lowest diff must be hard or lower
3. >5:00 can be anything

I'm thinking more like this. since H is way more accessible than N. IX only is just too small imo. Since we are having problems with 4:30 maps, we don't need to do anything with anything below 4:00

Hmm, I think a change like this would at least motivate people to map longer songs more. It probably won't make .mp3 extensions that much less popular of a choice when it comes to stuff like 4:50 length songs, though. But even with an IX spread requirement, why wouldn't you just extend the .mp3 instead if you've already done that before? A spread requirement change like this is less impactful on both the newbie playerbase and mapper perspectives, but also generally safer.

I definitely agree with the idea of scaling spread requirements based on song length. Right now I think it is pretty silly how a few seconds determine whether or not you need to map 5 difficulties or just one.

Personally I think that this will increase the amount of content we have due to the lack of longer songs as of now, despite decreasing the amount of content for newbies available in the long run.

However, I think if we went with this, a compromise with people that think this would reduce content too much would be having an insane difficulty required for drain time over 5 minutes, and reserving single difficulty mapsets for songs above 10 minutes, as has been previously proposed.

Overall, reducing the effort gap between approval length and spreads is a good idea, hopefully we can solidify the details in this thread.

the argument that spreads shouldnt start at insane is kinda odd the more i think about it because marathon maps dont need a spread at all. also the appeal for longer maps rises with player skill (at least thats how it was when i learned the game). also what you guys have to think about is that right now 2 diff spreads are already a thing if the hardest diff is hard or normal. if with ucs proposal a 7* for example was made you would still need 3-4 diffs for a reasonable spread with insane as the lowest diff, so the hardest <5min 2 diffs spreads would be like a high 5* and an insane diff witch does sound that unreasonable to me.

i would still love to see some number about number of plays on beatmaps for each diff and song length so the "majority of players" argument can properly be
talked about

im bad at excel but draw your own conclusions


taken from t/631455

i dont agree with getting rid of low diffs at any lenght but if i had to choose id go with what AncuL proposed

i agree with the scaling length thing. this eliminates the issue of normals not being mapped at all, while also lowering the amount of necessary effort spent on diffs that will be 1/4 as popular as the higher ones. this effort is a huge factor in why people don't map long songs.

i've noticed that almost all the people against this are the minority who are completely fine with churning out massive sets by themselves or finding gds if needed; i think that you guys don't realize how hard that is for most "normal" mappers who do not have 3-4 years of mapping experience and integration into the community necessary to find bns and gds for such long things. that breed of "normal" mapper makes up the vast majority of the mapping community too, so i really believe it's safe to say that the increase in 3:30+ songs will be quite large if the barriers to making the spreads for them are lowered. this will counteract the fact that quite literally 75%+ of ranked maps are under 1:30 nowadays.

im not exactly against this i just want this to be less lenient than what uc proposed
i barely map so i dont think this rule will affect me but im still taking in consideration casual players who like the song but arent skilled enough to play harder stuff. i experienced taht when i was new

It's only a minimum requirements for a map to get rank. If you really that carz about making a fullspread for 4+mins then there's no ine stopping you. The minimum requirements need to be reworked and the reason are already well stated by many user up there. However, I'm jot against fullspread either, it's harder to find modder for it than to map it after all.

Still on topic, ban mp3 manipulation to loophole duration requirements because it's literally ridiculous

Saturnalize wrote:

It's only a minimum requirements for a map to get rank. If you really that carz about making a fullspread for 4+mins then there's no ine stopping you. The minimum requirements need to be reworked and the reason are already well stated by many user up there. However, I'm jot against fullspread either, it's harder to find modder for it than to map it after all.

Still on topic, ban mp3 manipulation to loophole duration requirements because it's literally ridiculous

i tried to do that it didnt get received well

CXu wrote:

I'd say add a minimum amount of diffs required for the spread as well in some way, so say minimum 3 diffs for maps <4:00, and minimum 2 diffs for maps <5:00. (Not sure how this would affect those Easy+Normal sets only though).

If you're mapping a 4min ballad you can get away with only a hard since Insane/Extra might not fit well with the song, and Easy/Normal aren't required.
This should make it so that easier difficulties for newer players are still created whenever slower songs that would fit lower diffs better are mapped as well.

I second this option so much. It's by far more flexible and better than making thing easier by default.

Okoratu wrote:

Saturnalize wrote:

Still on topic, ban mp3 manipulation to loophole duration requirements because it's literally ridiculous

i tried to do that it didnt get received well

still sad people prefer to put less time on their creations than making it available for a fairly larger audience :/

Just like probably everyone else I've thought about this for forever, and it definitely seems necessary to have less of a hard cutoff at 5 minutes where the workload goes from potentially huge to low.

Though I share the concerns about making longer songs too inaccessible for newer players, so my personal idea of this has always been "Maps >4:00 and <5:00 need to have at least two difficulties, one of which must be a Hard or below". If we want to expand this to 3 minute maps, same thing with three diffs and maybe additionally set the requirement of a somewhat even spread but allowing it to be with only 3 diffs (to avoid N-H-7*), though there doesn't seem to be that much of a shortage of maps in that range so I'm not sure it necessarily needs changing.

A rule like this would also make banning mp3 extensions much more reasonable fwiw, since adding 50 seconds of R3 music box will no longer mean you need to map/get GDs for 20 minutes of drain time less. I've always been in favor of that under the condition of a more gradual spread requirement based on length.

just my thoughts, i like this idea.

there has been many times when I simply decided to not map a song because it was 4 minutes and 55 seconds long and i dislike the idea of editing the song.

as for new players, i think they would get really bored of playing a 4 minute beatmap, i know i would. so i see no problems here. tv size beatmapsets are regularly made so i doubt they will ran out of beatmaps to play.

Okoratu wrote:

Saturnalize wrote:

Still on topic, ban mp3 manipulation to loophole duration requirements because it's literally ridiculous

i tried to do that it didnt get received well

I would like to know the final conclusion about this. So you know, I can avoid attempting an Artist / R3 Music Box-map : )
At the same time any map comes after the mp3 extension ban won't be qualified.

To make my post relevant to the thread:
I myself also support CXu's idea regarding spread requirements that are dependant by mp3 length (duration) and speed (bpm). It is strongly advised for the set to be 1 level apart and not a duplicating level tier, though (for example, if an mp3 reached 3 diffs minimum, it should not be ENI, ENX, EHI, EHX, NHX, and NIX. EXX, EEX, and any kind of stupid thing like that is also not advised) to ensure the curve of the map doesn't go wild (also EXX is stupid and probably one of those X diff is GD)

Also:

Stefan wrote:

Okoratu wrote:

Saturnalize wrote:

Still on topic, ban mp3 manipulation to loophole duration requirements because it's literally ridiculous

i tried to do that it didnt get received well

still sad people prefer to put less time on their creations than making it available for a fairly larger audience :/

Doesn't matter if the mapper is good therefore the map is good because I'm the bootlicker and want to make my god feel proud and accepted uwu

CXu wrote:

I'd say add a minimum amount of diffs required for the spread as well in some way, so say minimum 3 diffs for maps <4:00, and minimum 2 diffs for maps <5:00. (Not sure how this would affect those Easy+Normal sets only though).

If you're mapping a 4min ballad you can get away with only a hard since Insane/Extra might not fit well with the song, and Easy/Normal aren't required.
This should make it so that easier difficulties for newer players are still created whenever slower songs that would fit lower diffs better are mapped as well.

i don't remember correctly whether there's a rule regarding difficulty count, but i'm sure that the only length appropriate for a one-diff map is 5 minutes and above. 5 minutes of length and below still needs to make at least 2 diffs. therefore it isn't valid and the mapper still needs to map another diff (whether it's normal, another hard, or insane) for the map to be able to be rankable

AncuL wrote:

CXu wrote:

I'd say add a minimum amount of diffs required for the spread as well in some way, so say minimum 3 diffs for maps <4:00, and minimum 2 diffs for maps <5:00. (Not sure how this would affect those Easy+Normal sets only though).

If you're mapping a 4min ballad you can get away with only a hard since Insane/Extra might not fit well with the song, and Easy/Normal aren't required.
This should make it so that easier difficulties for newer players are still created whenever slower songs that would fit lower diffs better are mapped as well.

i don't remember correctly whether there's a rule regarding difficulty count, but i'm sure that the only length appropriate for a one-diff map is 5 minutes and above. 5 minutes of length and below still needs to make at least 2 diffs. therefore it isn't valid and the mapper still needs to map another diff (whether it's normal, another hard, or insane) for the map to be able to be rankable

I know. I'm not talking about the current state of things, I'm saying that it would be good if the difficulty requirement for those lengths were lowered. Currently, if you have a 4:50min song and you have a really difficult map at like 7* or something, you would be required to map at least NHIX, most likely another X and possibly an E too, because people see the spread as not good. What I'd like is basically that something like, say, HX, IX, NX or something along those lines would be okay instead.

I guess it's not exactly the minimum amount of difficulties according to length that's the main proposal I'm trying to propose now that I think about it, but rather than the spread requirement gets progressively laxer until it hits 5min of length (or 6 in my opinion but yeah xd), where no spread is required anymore.

As for what laxer spread requirements are, it could a more relaxed easiest-diff requirement (such as it being a hard), but still with appropriate spread required (so ENH, HIX are okay, IXX, EHX are not), or it could be dependent on amount of difficulties and consistent spread, but not necessarily that the gaps are small/appropriate (so EHX, NIX are okay, but HIX or IIX are not), or something else entirely (or both).

I kinda disagree about huge diff gaps like NX or HX being okay. If you can map both N and X for a song, there's almost no way you cannot map anything in-between.
But i think you are saying that you want a spread based on star rating instead of difficulty name? Mind you that SR gap issue is already very subjective (with a certain value being the exact limitation), so I wouldn't agree to make it laxer with length as it isn't that necessary, this proposal is already enough to make mapping 4-minute maps rankable easier, and a good spread that isn't skipping a difficulty is good for certain audience who think 3.9* hard is too easy but 5.3* extra is too hard

Then again for mappers this change is cool, but players is not, you want to introduce people to the game and having less content for them here https://osu.ppy.sh/p/beatmaplist , since new players usually search for maps at the first page (atleast that's what I did, back in the days).

Didn't the old approval thing work for years and was only changed cause it was redundant cause of pp leaderboards?

I dont see what the issue is with the 'new player experience' when it worked before

kwk wrote:

Didn't the old approval thing work for years and was only changed cause it was redundant cause of pp leaderboards?

Correct me if I'm wrong but wasn't it eventually changed to just be what we have now? (just that now it's both the same category because pp). It was "super hard/unusual or long maps" for a while and then just "long maps", though I'm not familiar with the exact reasoning behind the change.

That being said I don't see much of an issue with even an HX spread, if you have a 3.5* and a 6* diff then there should be something for most people (you can nomod/hr/dt the 3.5* and once that is too easy you should be able to play the 6*), only exception being very new players, and I doubt there'd be much of a shortage of maps for them to play anyway, Full TV size sets still don't require more work than a 4 minute 2 diff spread, and something like N/I would probably also happen as far as 4 minute maps go.

New players only stay at the "Normal" stage for a short while I'm pretty sure, once you get the basics of the game down you can at least somewhat play most hards and frankly they're way more fun (at least they were for me) since they aren't undermapped to shit. So I agree it's not much of an issue if we have a few less normals if the tradeoff is having more ranked maps just under 5 minutes.

AncuL wrote:

I kinda disagree about huge diff gaps like NX or HX being okay. If you can map both N and X for a song, there's almost no way you cannot map anything in-between.
But i think you are saying that you want a spread based on star rating instead of difficulty name? Mind you that SR gap issue is already very subjective (with a certain value being the exact limitation), so I wouldn't agree to make it laxer with length as it isn't that necessary, this proposal is already enough to make mapping 4-minute maps rankable easier, and a good spread that isn't skipping a difficulty is good for certain audience who think 3.9* hard is too easy but 5.3* extra is too hard

No, I just said that I want the requirements of a spread to be laxer, regardless of if that's based on the gap between diffs, or what the lowest diff required is (which is the proposal in OP). The first "alienates" players inbetween (so H/I players, if the spread is NX), while the other "alienates" new players (if IX), but we already do this at a hard cap at 5min. It's not so much about if someone can or can not map a proper spread, but if enough people are willing to do so that it's worth keeping the current system over changing it to make more people willing to map songs of these lengths. Most 5min+ songs can have a full spread, if people are willing to put the time into it, but as we saw before the approval limit was lowered as well as the trend now, they're not really willing to spend that much time, and we're seeing a similar thing happening for maps around 4:30-4:59min.

But just lowering the approval limit itself isn't going to fix the problem, as you'll still get a huge difference in amount of work needed between a 1 second difference between, say, 4:29 and 4:30, if we just lowered approval cut-off.

What I want is that instead of having a cut-off for marathon length, we have a more gradual laxing of the spread requirements. The marathon length cut-off today is essentially just a "okay if your song is longer than this you don't require a spread".

As for why I mention the gap thing; if the minimum diff required is I or lower, as OP suggested, then if someone has an IXX map for example, and wanted to add an easy diff, they would also be required to add an N and H, which would detract from people adding Easy diffs if we follow regular spread requirements.

from the stats we have, we can see that Hard and Insane are easily the most played we have in spreads.
so to me, if we were to ever create a system where we can allow less maps via duration, i'd say as long as the mapset contains a Hard or Insane (or both), then the spread above or below that should not matter.

this way people won;t alienate the majority of players, and also not have to put as much effort into the sets to do a full spread especially for SRs above 6*

for example, a 3.14* Hard, a 4.8* Insane, and a 7.5* star Expert for a song in the 4 minute range would be valid

As requested from kwk, here are some (barebones) graphs. The delay was because my old scraping script broke between last year and now, since M a r v o l l o had a map with a tab character in the tags and SOTARKS had a map with a newline in the tags (I kid you not). Of course in my .tsv file the delimiters are tabs and newlines.



My interpretation is that the majority of Easy, Normal, and Hard plays are not on long (4+ minute) marathon (5+ minute) maps. Feel free to draw your own conclusions or request other graphs.

Source code

Edit: Show up to 6 min bins and add color.

Seems reasonable to conclude that Normals shouldn't be required for maps that are 4:00 or longer.

Should hards also be required at 4:00, or have that requirement to 4:30

x86 wrote:

My interpretation is that the majority of Easy, Normal, and Hard plays are not on long (4+ minute) marathon (5+ minute) maps. Feel free to draw your own conclusions or request other graphs.

Can you make graphs with 150- excluded and play count values unified for better visual presentation?

Sinnoh wrote:

Seems reasonable to conclude that Normals shouldn't be required for maps that are 4:00 or longer.

Should hards also be required at 4:00, or have that requirement to 4:30

i mean judging from that data hards also get barely played beyond 4 minutes

@x86 it also seems insane is not played as much either. we need to compare it to the playtime of all harden and insane’s because our samplset for 4 min+ is very small comparatively. i think what i suggested holds water at least, considering the amount of plays we see on Expert+ on the graphs

Mentai wrote:

@x86 it also seems insane is not played as much either. we need to compare it to the playtime of all harden and insane’s because our samplset for 4 min+ is very small comparatively. i think what i suggested holds water at least, considering the amount of plays we see on Expert+ on the graphs

the fact that insane play counts bump up at 300seconds implies that players at that level do in fact have interest in longer maps id say.

im wondering wether this data is graphed against the number of maps there are for each lengths or if its total numbers. because that would change how to view this data significantly

sdafsf wrote:

Mentai wrote:

@x86 it also seems insane is not played as much either. we need to compare it to the playtime of all harden and insane’s because our samplset for 4 min+ is very small comparatively. i think what i suggested holds water at least, considering the amount of plays we see on Expert+ on the graphs

the fact that insane play counts bump up at 300seconds implies that players at that level do in fact have interest in longer maps id say.

im wondering wether this data is graphed against the number of maps there are for each lengths or if its total numbers. because that would change how to view this data significantly

this could maybe support the idea of having insanes on these sets even more then. i think having a hard as well would be a good idea still

Sieg wrote:

Can you make graphs with 150- excluded and play count values unified for better visual presentation?

I'm not a fan of juxtaposed bars so I've used stacked bars here.

sdafsf wrote:

im wondering wether this data is graphed against the number of maps there are for each lengths or if its total numbers. because that would change how to view this data significantly

To clarify, I am subsetting every ranked/loved standard map on every combination of hitlength (30 s) interval and difficulty, and then summing the playcount for all those maps.

x86 wrote:

.

I mean this proportion should be expected seeing the more active players tend to get better at the game just from playing, but its no real argument against the current 5 minute rule. the best argument against those kinds of rules is when the playcount by worse players against the number of maps that exist for that length of song is equivalent to zero. why? because people who are bad at the game tend to not care about length of beatmap for the most part, they just end up playing their favorite music.

I believe if you guys move forward with this there should be an incentive or safety net for these lower difficulty players otherwise there exists a possibility that lower difficulties will become arbitrarily extinct and the player base for this game will no longer grow.

Having some sort of gradual and player interactive tutorial to get players to be able to play at least hards/insanes should exist.


Also I remember when i tried to get better at the game, at some point I would player the longer easier maps to try and create a sort of stamina or real consistency lol, and without longer songs having lower diffs to them that kind of approach would no longer be possible.
Similarly saying I'm not saying that lower diffs wouldn't exist anymore, but think about how newer players get into this game, they don't spend 15 minutes looking for the perfect map for them to play. They look at the recently ranked section and just download some song they might recognize, or they search for the songs they like to see if there is something with their SR to it. So having lower difficulties end up being much more sparse just makes it to where they can't just find that one recently ranked map that has a diff they can play (given the possibility stated prior).


A lower workload for mappers will be appreciated for certain, but a game without some hand holding introduction like the one that exists with the 5 min rule now would lead to a decrease in number of players who start to play this game, which could lead to less people who play this game overall since many players can get bored and just move on to other games. At that point, who will play your insane/expert diffs?