Hello everyone, it has come to our attention that one of the rules regarding spread in the osu!mania RC is clouded with ambiguity and we would like to reword it so everyone regardless of mode can understand it. The rule will still be kept intact with regards of its meaning but this will contribute it to be better understood as a whole.

Current


New proposal


As an extra note please keep in mind that as quoted in the o!mRC:


An example of why this rule is too ambiguous would be this.

Before:

1. NHIX - ok
2. HIXX - not ok (there must be a normal according to ENHIX convention)

After:

1. NHIX - ok
2. HIXX (all diffs have "similar difficulty gaps typical of those between lower diffs such as E and N" ) - ok

Currently ongoing PR for this rule

We want to know what you think about this and it’s clarity. This is with the intention to allow all X spreads, this rewording should reflect that. Feel free to discuss this and give feedback whether this change is okay, it can be improved or further suggestions.

The new proposal wording will allow such as XXXX diff spread so I disagree with this proposal

RiP46 wrote:

The new proposal wording will allow such as XXXX diff spread so I disagree with this proposal

Rip, common interpretation of this rule has always been to allow such spreads (see beatmapsets/1700983#mania/3491269, beatmapsets/1928408#mania/3982474, beatmapsets/1829908#mania/3755775, etc. to name a few). This is a rewording to clarify this intention for those who are not in the mania know-how.

for teevee-size/cut ver. if this will apply, does that mean spread now requires ENHI? :psy:

The rewording definitely reduces ambiguity in how the spread rules could be interpreted. This better aligns how spread rules have worked in mania since the last relevant RC proposal.

+1

Damaree wrote:

for teevee-size/cut ver. if this will apply, does that mean spread now requires ENHI? :psy:

No, lowest required difficulty is still always going to be a Normal for TV sizes (1:30 set) since it uses the first allowance, "...lower than 2:30, the lowest difficulty of each included keymode cannot be harder than a Normal"

Damaree wrote:

for teevee-size/cut ver. if this will apply, does that mean spread now requires ENHI? :psy:

Nope, it will still be the same as it is. The only thing this proposal tries to achieve is reduce confusion between other game modes and make it more clear as to what we are trying to do.

just to clarify a previous post, the acceptable spread would be X X+1 X+2 X+3 right? not actually XXXX when all those diffs are similar difficulty of course.

I casually pitched in the idea of "scaling from the highest difficulty and making a spread with a miminum number of difficulties" during RC rework convos in the BN server way back and went with the assumption that an XXXX for sub 2:30 would be ok granted that proper difficulty scaling is observed as something like this can easily happen in 7k.

Fair to say, this eliminates ambiguity on that regard so its good.

May I know what scenarios caused the need for this rewording?

EDIT: @mint yeah thats the goal of that specific RC clause in the first place, for the benefit of some mappers who would want to push insanely hard 7k charts that happen to be below 2:30 in length.

Damaree wrote:

for teevee-size/cut ver. if this will apply, does that mean spread now requires ENHI? :psy:

no

-mint- wrote:

just to clarify a previous post, the acceptable spread would be X X+1 X+2 X+3 right? not actually XXXX when all those diffs are similar difficulty of course.

Yes, I want to clarify that we are NOT disallowing the use or ranking of X spreads. Just making it more clear for everyone. As long as you refer difficulty gaps by the numbers then yes it is still acceptable.

Hands down to this rewording. Better explanation, no fuss, straight to the point.

-mint- wrote:

just to clarify a previous post, the acceptable spread would be X X+1 X+2 X+3 right? not actually XXXX when all those diffs are similar difficulty of course.

Yes. A spread with only Xs of same difficulty cannot be ranked. However the scenario you given can be ranked since anything above I (such as Xs) should have proper progression. This should be checked by players and BNs whether these Xs scales up properly or not.

lenpai wrote:

May I know what scenarios caused the need for this rewording?

To clear up ambiguity. Current wording is too hard to understand.
The new proposal clears up everything, though both are technically the same.

lenpai wrote:

May I know what scenarios caused the need for this rewording?

Other game modes saw this rule and as they have difficulty levels defined as ENHIX, thought that an XXX spread being allowed was confusing since they’re virtually the same in that written concept.

Hence why this proposal to clear doubts between game modes and make it more coherent throughout.

Ryu Sei wrote:

lenpai wrote:

May I know what scenarios caused the need for this rewording?

To clear up ambiguity. Current wording is too hard to understand.
The new proposal clears up everything, though both are technically the same.

so preventive? gotcha
thought some heated discussions came out of interpreting this

EDIT: thanks for the explaning the background for this change @antalf

I think the additional diffs should have a fairly noticable decrease in difficulty from the top diff. I dont mind if the spread is technically XXXX, but it still should be obviously decreasing. For example, 7.1*>6.5*>5.8*>5.1*.. just throwing random numbers out there.

But with this wording, it seems like a set that is XXXX could have pretty sporadic variations in difficulty throughout each diff with not much care of a linear gaps between difficulties. Like 4 diffs that are all around 6.5* for example, where the "lowest" diff could end up having harder sections than the top diff. And that type of spread feels like a problem to me. I imagine i wont have many people agree with me but i thought i should give my opinion anyway.

edit: ah it seems that this was kind of cleared up already. but since a few people are initially confused about the same thing, there may still be a problem in wording

That falls into unjustified difficulty spike. If it's meant to be 'lowest' of Expert difficulty, then the map should be subjectively 'easiest' across all sections.

Crisper wrote:

I think the additional diffs should have a fairly noticable decrease in difficulty from the top diff. I dont mind if the spread is technically XXXX, but it still should be obviously decreasing. For example, 7.1*>6.5*>5.8*>5.1*.. just throwing random numbers out there.

But with this wording, it seems like a set that is XXXX could have pretty sporadic variations in difficulty throughout each diff with not much care of a linear gaps between difficulties. Like 4 diffs that are all around 6.5* for example, where the "lowest" diff could end up having harder sections than the top diff. And that type of spread feels like a problem to me. I imagine i wont have many people agree with me but i thought i should give my opinion anyway.

edit: ah it seems that this was kind of cleared up already. but since a few people are initially confused about the same thing, there may still be a problem in wording

I think that it will be good to mention that we have something set in place for this that is the last sentence of that spread ruling regarding of how difficulty gaps are expressed. I’ll add it just in case for further clarification.

Antalf wrote:

I think that it will be good to mention that we have something set in place for this that is the last sentence of that spread ruling regarding of how difficulty gaps are expressed. I’ll add it just in case for further clarification.

Cool as long as thats addressed then all looks good to me!

Crisper wrote:

Cool as long as thats addressed then all looks good to me!

This is as expressed in the ranking criteria.

+1,

The idea is spot on, and I do believe that it might help clear off some questions and confusion (make it more coherent). However, if this were the case, I do believe some altercation of what the definition of a "proper spread" should be made.

Currently, the idea of a spread is quote:

On difficulties Insane and harder, a proper spread is defined by relative difficulty gaps similar to those formally defined between lower difficulties.

I do believe this to be a little too vague, given that although Insane / Expert difficulties do not have many spread guidelines, I do believe for this definition to run over ALL difficulties not just those Insane or higher.


I believe also rephrasing some definition of "spread" might create a coherent judgement of this RC Proposal as well. As the current definition might not progress very well and create some ambiguity.


I propose if we change this, to change the definition of spread from:

BEFORE

On difficulties Insane and harder, a proper spread is defined by relative difficulty gaps similar to those formally defined between lower difficulties.

AFTER

On all beatmaps, a proper spread is defined by the relative increase of progression in each difficulty from the lowest difficulty to the highest difficulty.

OR

On all beatmaps, a proper spread is defined by the relative increase of progression in the set's difficulty from the lower difficulties to the higher difficulties.

Hopefully we can come up with a solution for this :3

[ Decku ] wrote:

I do believe this to be a little too vague, given that although Insane / Expert difficulties do not have many spread guidelines, I do believe for this definition to run over ALL difficulties not just those Insane or higher.


I believe also rephrasing some definition of "spread" might create a coherent judgement of this RC Proposal as well. As the current definition might not progress very well and create some ambiguity.

Will wait for more feedback regarding this. Should be okay but I will like to see what others think first.

So based on the wording and how this should work, i think when there are mutiple top difficulties for example, pretty close difficulty range but different styles, they should be counted as only one in the 'proper spread'

i don't know how others think about it, but i do feel it can be questioned when using terms like 'roper spread containing at least 4 difficulties'. The clear difference between XXXX and XX+1X+2X+3 has been mentioned above would not cause too much confusion imo, however when there are similar difficulties in spread, like NHHI or HIXX, you will have to define this as an 'inproper spread' or 'proper spread containing only 3 diffs' based on current wording

Personally feel it's just not that intuitive from logic, for this one is what i think going a bit worse than the original wording using the term 'difficulty level', just my thought tho

Overall +1

gzdongsheng wrote:

So based on the wording and how this should work, i think when there are mutiple top difficulties for example, pretty close difficulty range but different styles, they should be counted as only one in the 'proper spread'

i don't know how others think about it, but i do feel it can be questioned when using terms like 'roper spread containing at least 4 difficulties'. The clear difference between XXXX and XX+1X+2X+3 has been mentioned above would not cause too much confusion imo, however when there are similar difficulties in spread, like NHHI or HIXX, you will have to define this as an 'inproper spread' or 'proper spread containing only 3 diffs' based on current wording

Personally feel it's just not that intuitive from logic, for this one is what i think going a bit worse than the original wording using the term 'difficulty level'

But overall +1

I understand your concern and that's why we want as much feedback as possible. I do have to quote though the that we do have something explaining what a "proper spread" should be. But once again, let's see if this is an ongoing confusion and if it needs some work as well.

Literally just goes from exclusive-1 to inclusive with no real rule change so pop off I guess

how are you measuring difficulty levels?

I heard some people use SR which seems weird and would allow for potentially very stupid configurations as the first reply pointed out

Also, removing the need for their difficulty levels to scale meaningfully (or at least explicitly) in some way you could make a bunch of Insane level difficulties that get marginally harder and call it proper so long as there's at least 4 difficulties total on any song?

That seems fucked? I dunno?

For clearer expression and reduced ambiguity, overall it's pretty good from my side.

Okoratu wrote:

how are you measuring difficulty levels?

I heard some people use SR which seems weird and would allow for potentially very stupid configurations as the first reply pointed out

Mania does not use SR for measuring difficulty levels if a difficulty is not an insane it should not be named an insane in the first place. So imo that shouldn't be a problem we already have a bunch of like super difficult 4*s ranked

Tailsdk wrote:

Okoratu wrote:

....

Mania does not use SR for measuring difficulty levels if a difficulty is not an insane it should not be named an insane in the first place. So imo that shouldn't be a problem we already have a bunch of like super difficult 4*s ranked

good, i was momentarily concerned for y'all as I heard that floating around on some place i could see 👍

To clarify, in osu!mania a difficulty is seen from how does the map plays and the pattern looks. The complexity of osu!mania patterns makes SR is disregarded in determining a difficulty. There is always a case where expected difficulty to star rating perception is different than what the game system calculates.

In case of determining if two difficulties are considered on the same rating or not, it's always a good idea to discuss with osu!mania community (mappers, players) and BNs.

relaying my thoughts from the bn server here, skip to bolded terms for tldr:

i think this change is good for standardizing it a bit more with the other modes rc mostly because the confusion mainly came from them as i understood (as well as including the top difficulty into spread decisions). however, this leaves a lot of ambiguity about lower difficulties, and i feel the exclusion of the term "difficulty levels" is the main contributor to that.

the addition of "level" back when the rule was originally proposed was meant to add level of relativity to it. "level" was meant to encompass all sorts of difficulties that fit within that spectrum, such as u (or defined as x+) x spreads for songs 3:15 - 4:00, and h+ix spreads for 2:30 - 3:15 etc. focusing on the latter example, at the time there were many difficulties that simply did not make sense as a "hard" or a "normal." osu mania's rc is written in the sccope of 180 bpm, but judging guideline breaks beyond that was very difficult because skillsets in mania do not scale linearly (such as stamina/enduranced based maps, jacks, and faster single streams). and so this allowed maps that fell within the "n+" or "h+" to be indicated appropriately and still be apart of a rankable spread. essentially loosening up spread requirements.

the removal of this term technically still allows for xx+x++(etc.) spreads to be rankable (in the 3:15-4:00 range). as the term "proper spread" is included which is latter defined by:

osu!mania Ranking Criteria wrote:

On difficulties Insane and harder, a proper spread is defined by relative difficulty gaps similar to those formally defined between lower difficulties.

however, this doesn't apply to these lower difficulties. and it doesn't make sense to apply them to the lower difficulties since they define this rule in the first place. i think this is what decku attempts to do with their post (community/forums/topics/1746229?n=20) but i feel this is still too vague. i think it's important that we still use these lower difficulties as the basis for what a proper spread looks like in higher difficulties. this is inherently due to how lower difficulties will result in a noticeable gap in difficulty.

i think the solution here would be to include the term "difficulty level" and list out within the rc what exactly we consider to be "difficulty levels" similarly to wiki/en/Beatmap/Difficulty, including the "n+" "h+" and "i+" difficulties. as well as denoting what the "+" means in a separate clause. here is my proposal (using one as an example):


and within the glossary under difficulty names:

i'm sure the wording could be improved on the last one, and some changes would need to happen on the difficulty naming page to fit a proper reference but I think this is an okay start for now. do keep in mind I think the definition for "+" along with the existing guideline mentioned earlierr is necessary to allow spreads that may fall into the "x x+ x+2" predicament as technically "x+2" can exist now.

Guden, while I do agree we could define those things you then enter shenanigans territory (would a HIX set be a spread? Why is N+ not H? What is an N+ and not a H?). I would move that to a different proposal as it is probably worth debating but since this proposal is kindof a clarification of how spreads work for mania I would focus on that wording first.

I understand the intention is to allow XXXX spreads but i think that in itself can become quite problematic so I disagree with this proposal

SilentWuffer wrote:

I understand the intention is to allow XXXX spreads but i think that in itself can become quite problematic so I disagree with this proposal

they have been rankable for some time and proven to be not problematic
see: community/forums/posts/9060742

my main concern is that this will cause a lack of content for newer players because mappers tend to prefer making higher difficulty maps

Well the question is that do you feel there's the lack of newer players content in the current ranked section?

If the answer is no, that means the current proposal won't be a problem either.

This proposal aim is only to reword what's already been allowed in the ranked section for 1.5 years. If the current ranked section have decent amount of lower level difficulty coming out on constant basis, this proposal won't change anything on that at all.

SilentWuffer wrote:

I understand the intention is to allow XXXX spreads but i think that in itself can become quite problematic so I disagree with this proposal

As Maxus stated, X only spreads have been allowed for quite some time now. If you wish to change this, that will need to be an entirely separate proposal. This proposal exists because people from other gamemodes had some confusion about how "difficulty levels" should imply that X only spreads cannot exist despite that being the case in practice.

Welp, here goes another update.

After further discussions with different game mode users, the approval of the community including everyone not only staff, this was decided to be pushed since it was the wording that gave the most clarity in regards to the use of X spreads and what we are doing in mania so far. You can keep up the discussion of the following PR for this change in:

https://github.com/ppy/osu-wiki/pull/9078