1. osu! forums
  2. osu!
  3. Development
  4. Ranking Criteria
  5. Finalized/Denied Amendments
show more
posted

ZiRoX wrote:

As others have said, I don't think the minimum 2 diff rule should be gone. As the proposal stands now, I could map a 1 min Normal and get it ranked. You could say that the proposal mentions that maps require a proper spread (and you can't have spread with only 1 element), but an explicit mention of this minimum number of diffs required would be good.
I agree. There should always be at least 2 diffs unless its a marathon. Thats what differentiates a marathon from a normal set.
posted

tatatat wrote:

ZiRoX wrote:

As others have said, I don't think the minimum 2 diff rule should be gone. As the proposal stands now, I could map a 1 min Normal and get it ranked. You could say that the proposal mentions that maps require a proper spread (and you can't have spread with only 1 element), but an explicit mention of this minimum number of diffs required would be good.
I agree. There should always be at least 2 diffs unless its a marathon. Thats what differentiates a marathon from a normal set.
i agree
posted
Yeah, make songs under 5 minutes still require minimum two difficulties and I'd support this.
Longer songs that call for a Hard (or even Normal) as highest diff instead of something higher are probably the most suitable candidates for longer Normals (or Easys) anyway. Having just one diff here wouldn't even feel like a mapset anymore in these cases.
posted
I would like to push for 4:15 as a cutoff rather than 4:30 for previously mentioned reasons:

Nao Tomori wrote:

if you people want linear just go with 3:30 - 4:15 - 5. that both avoids the "issue" of non linearity while also keeping fairly normal standards for drain time per set; for a 3 minute nhix set, that's 12 minutes and about the max amount ever needed for a set, which is perfectly fine imo... at 3:30 hix that becomes 10:30 which isn't a huge reduction and then 4:15 ix is 8:30 which is also not a big difference.
edit: ok about the 2 diff rule

I spoke with UC about it and my mind has been somewhat changed:

ZiRoX wrote:

As others have said, I don't think the minimum 2 diff rule should be gone. As the proposal stands now, I could map a 1 min Normal and get it ranked. You could say that the proposal mentions that maps require a proper spread (and you can't have spread with only 1 element), but an explicit mention of this minimum number of diffs required would be good.
This is the equivalent of ranking a 2 diff EE set, and it's been done before. Ranking a 1 min single Normal/Easy diff could even be better off because the diff underneath it could be completely unnecessary (1.6* normal, 1.4* easy for example, the easy is basically unneeded in most cases like this). This actually also solves UC's other proposal as well: https://osu.ppy.sh/community/forums/topics/726926 but that's another topic altogether.

Forcing 2 diffs may not do much for content/variety for this reason: if we forced 2 diffs minimum we could get 4:15/30 sets like II which seem pointless because there's already an insane there. The second diff may be effortlessly mapped anyways. Instead, why not focus all of the effort into one great diff instead of splitting it 20/80 for another difficulty that may not benefit the set aside from "here's another diff"?

tatatat wrote:

Thats what differentiates a marathon from a normal set.

Doormat wrote:

The idea that I could make a “marathon” for a TV Size if I only mapped a Normal difficulty completely undermines the original intent of this proposal to begin with (balancing spread requirements for longer songs).
Marathons should be defined by song length and length alone, difficulty count should be irrelevant - and kinda is - because some marathon maps have full spreads/more than one diff. Forcing 2 diffs for the sake of differentiating sets from marathons is unnecessary imho (not saying that's what your point is but I want to dismiss that idea entirely)

I think what maybe needs to happen is a rethink of what "marathon" constitutes, because it shouldn't be "1 difficulty mapset" by any means.

The "effort" argument can be seen as being "for the sake of effort" as well and may not have a benefit to the mapset at all despite being more work for the mapper in the long run. AFAIK the proposal's main aim is to gear sets towards the people that get the most out of playing them, and those people likely won't care much for filler diffs that are there just to fulfill the 2 diff minimum.

i kinda used words from UC and Hobbes2 in here too since we had a discussion about it on Discord so disclaimer some of the phrasing isn't mine
posted
you can ask gd ifyour song length is too long
posted
Single-mode mapsets must form a reasonable spread. This spread must comply with its respective mode's difficulty-specific Ranking Criteria.
My main issue is the wording here. Most people are worried that this removes the 2 diff requirement but you can't really have a spread if you only have 1 diff so it's kinda ??? In addition, this change wouldn't allowed NN or II stuff to be ranked if the two diffs are very similar in difficulty since once again having 2 incredibly similar diffs difficulty doesn't make a reasonable spread?????????

Reasonable spread isn't well enough defined for this rule to properly work since a lot of it is up to interpretation and you can twist this rule to be applied for both allowing single diff maps under 5:00 and forcing 2 diffs under 5:00. Since I'm not sure what the intent of this rule change was I'm not sure what a better wording would be, but my point is that as is the rule really doesn't work currently.

But then again, I think the 2 diff should stay (or at least make the wording better where you require 2 diffs)
posted

Mir wrote:

Marathons should be defined by song length and length alone, difficulty count should be irrelevant - and kinda is - because some marathon maps have full spreads/more than one diff. Forcing 2 diffs for the sake of differentiating sets from marathons is unnecessary imho (not saying that's what your point is but I want to dismiss that idea entirely)

I think what maybe needs to happen is a rethink of what "marathon" constitutes, because it shouldn't be "1 difficulty mapset" by any means.

The "effort" argument can be seen as being "for the sake of effort" as well and may not have a benefit to the mapset at all despite being more work for the mapper in the long run. AFAIK the proposal's main aim is to gear sets towards the people that get the most out of playing them, and those people likely won't care much for filler diffs that are there just to fulfill the 2 diff minimum.

i kinda used words from UC and Hobbes2 in here too since we had a discussion about it on Discord so disclaimer some of the phrasing isn't mine
I used “marathon” in quotes because that’s probably the most recognizable thing that this new proposal will get compared to. I know it technically isn’t a marathon; I just used the word for the sake of simplifying my argument.

I understand that this suggestion to the proposal was made in order to help reduce the redundancy of something like a EE or HH spread; this isn’t what I’m concerned with. I’m more concerned with the opposite effect: people becoming complacent and only mapping a single difficulty.

The implication that I could theoretically rank a one difficulty mapset as long as it meets a minimum difficulty is what worries me. I honestly can’t see the purpose of requiring a spread anymore if you’re going to use the argument that people won’t care for filler difficulties in order to fulfill the minimum difficulty requirement. With the current amount of lower level difficulty maps, you can make the argument that making filler difficulties is unnecessary anymore because most new players can just play those until they get better at the game; we might as well just get rid of spreads to begin with.

Only we don’t do that, because that’s dumb and would only lead to a huge influx of ranked maps. When this is coupled with the rule that only one mapset of a song (assuming that they’re of the same game mode) can be in Qualified at a time, this will only lead to a huge influx in maps that will be perpetually stuck waiting to get ranked. Which will only complicate things for Beatmap Nominators as we have to decide which one should be ranked first, which only leads to a growing frustration in the community...

Like I said, I think that the minimum difficulty level is a good idea conceptually; it can help get rid of redundancy while opening up more opportunities for mappers to get their maps ranked. However, I also think that there needs to be a minimum requirement of difficulties in order to prevent potential abusing of the system.

Edit: I also dislike the idea that this proposal is “geared towards people that get the most out of playing longer length songs.” Maps should be inclusive and geared towards everyone, not some exclusive club that is going to benefit the most out of the changes. Having a minimum difficulty count helps to give players more options in choosing what they want to play.
posted
a lot to unpack, bear with me

first of all, @everyone, "reasonable spread" is a glossary term in the current ranking criteria, so please go read up on that bc it makes a lot of things more clear, like how single diff sets innately feature reasonable spreads

"single diff mapsets wouldnt feel like mapsets"
people are already well accustomed to single diff mapsets because marathon sets get ranked like every day, there will be no difference. its not like the average really think "mapset = a set of maps" anyway, they just play diffs they find, no reason to bar this from passing because of a weird pseudo-technicality

"i could make just a 1 minute normal and rank it"
yeah and you can make a 1 minute EN set right now and rank that, there's basically 0 difference in effort or value. 2diff requirement doesn't keep people from making stupid sets, and it doesn't even discourage it. changing this to a single diff requirement will not encourage it any further because people don't really like to make stupid sets anyway, or else we'd see a lot more of them rn

"it defeats the point of having spreads to begin with"
again, marathons are already very prolific and are not inhibiting spreads in any way. tv sizes are super popular and always will be and will require a full spread, 3:30 H sets will not intrude upon that

"it promotes laziness"
it also promotes caring about the single difficulty you want to make instead of phoning in a forced difficulty you don't want to make. people are always going to be lazy but at worst we get 1 lazily made difficulty, which surely is better than 2

"2 diffs promotes variety/appeals to more players"
except EE/NN/HH/II are all rankable and do not promote either of those things. in fact of all the listed sets the ones we have ranked right now are EE and NN....

mir did a good job covering a lot of arguments already, was there anything else? i may have missed something..

basically by removing the 2diff requirement it makes for overall more clarity when the proposal goes through as all you need to understand is "linear spread from this diff onwards." forcing II sets in 4:30 songs is very awkward and unwarranted as it does not promote good content but instead just arbitrarily forces more content. from what i can see there is no benefit whatsoever from requiring 2 diffs

@doormat your concern about the bn rule interacting poorly with single-diff sets is valid, but that seems more of an issue with the bn rule itself, which is a very different topic. we shouldn't let bn rules interfere with improving ranking criteria.
posted
in other news, i just realized the thing i told irre was totally accidentally a lie, the thing i had that allowed for EHIX 3:30 sets got removed by mistake somewhere...
will patch that bug shortly (in draft 3 i guess)

also it seems like people are more in favour of the linear 3:30/4:15/5:00 progression? any naysayers speak up or that will be changed
posted

UndeadCapulet wrote:

"it promotes laziness"
it also promotes caring about the single difficulty you want to make instead of phoning in a forced difficulty you don't want to make. people are always going to be lazy but at worst we get 1 lazily made difficulty, which surely is better than 2
we have some other major issues if bn's are nominating lazy diffs
posted
havent read any prior discussion just want to throw my two cents.

i think every difficulty should really be involved in any map where applicable, but i think judging them on star rating like what we do now to form a perfect "spread" is a little useless. if normal comes to be 1.8 and hard comes to be 3.8, that should be allowed so long as the maps are of good quality yknow?

i also think we should be required to add breaks to (higher difficulties especially) longer maps. i hate playing through 4 minutes of a map without being able to stop. it hurts and i imagine its not good to strain ones self like that.

thats all carry on
posted
best of luck
posted

UndeadCapulet wrote:

"i could make just a 1 minute normal and rank it"
yeah and you can make a 1 minute EN set right now and rank that, there's basically 0 difference in effort or value. 2diff requirement doesn't keep people from making stupid sets, and it doesn't even discourage it. changing this to a single diff requirement will not encourage it any further because people don't really like to make stupid sets anyway, or else we'd see a lot more of them rn

"it defeats the point of having spreads to begin with"
again, marathons are already very prolific and are not inhibiting spreads in any way. tv sizes are super popular and always will be and will require a full spread, 3:30 H sets will not intrude upon that
my argument here was that if we're using the argument that filler difficulties are largely seen as obsolete, what is the point in requiring a spread? why should tv sizes be required to have a full spread when we're going to be exempting longer length maps from the same, just because they have meet a difficulty level requirement?

UndeadCapulet wrote:

"it promotes laziness"
it also promotes caring about the single difficulty you want to make instead of phoning in a forced difficulty you don't want to make. people are always going to be lazy but at worst we get 1 lazily made difficulty, which surely is better than 2
i think this is a very pessimistic way of looking at things; just because lower level difficulties may take considerably less time to make does not mean that the same level of care/effort hasn't been put into them. there are still a lot of people in this community that do put a lot of care and effort into making a balanced spread so that people of all skill levels can enjoy them. even then, if the mapset host doesn't want to make a lower-level difficulty, there are plenty of others in this community (guest difficulties) that would be more than willing to give their take on making something so that players of all skill levels can enjoy a mapset for a given song.

UndeadCapulet wrote:

basically by removing the 2diff requirement it makes for overall more clarity when the proposal goes through as all you need to understand is "linear spread from this diff onwards." forcing II sets in 4:30 songs is very awkward and unwarranted as it does not promote good content but instead just arbitrarily forces more content. from what i can see there is no benefit whatsoever from requiring 2 diffs
the benefit that i've been trying to argue is that it provides more options for players to enjoy the songs they want to play to. not everybody is going to have access to all maps, or may be selective in the type of songs they choose to download, so ensuring that there is a wider selection of difficulties for players to choose from that suit their skill level is crucial, in my eyes.

UndeadCapulet wrote:

@doormat your concern about the bn rule interacting poorly with single-diff sets is valid, but that seems more of an issue with the bn rule itself, which is a very different topic. we shouldn't let bn rules interfere with improving ranking criteria.
i actually think that these two go hand-in-hand. if potential ranking criteria changes have the potential to affect the rate at which maps will be ranked, then it can become a potential problem for the community: nominators will be swamped with an even bigger workload, and mappers will grow more frustrated at having to wait for their maps to enter the Qualified section if it so happens that somebody else qualified a smaller-sized mapset before theirs. getting rid of the "same song restriction" is also open to a whole bunch of new issues as well, as i'm pretty sure nobody wants to see the same song in qualified a bajillion times. these issues should be addressed and solved together, not separately.
posted
point 1: difficulties in a spread are not filler difficulties, they are difficulties in a spread. this proposal is only upping the difficulty of the starting point for a spread for longer songs to better cater to mappers' needs and to better suit spreads to players' needs. it is not saying that spreads are not valuable. forcing 2 diffs in a mapset is not forcing the contribution of a spread, it is just forcing a filler difficulty. once again, we have had single-diff sets in the form of marathons forever, and they have not changed the value or perception of spreads in any way.

point 2: nice to hear people can make gds or other diffs if they want, but this is not a reason to force it.

point 3: this sounds more like you are against the entire proposal itself instead of just being against single-diff sets. forcing a second difficulty in a set does not give more options for players, as shown by how the rule is currently forcing EE/EN/NN sets that offer basically no variety. alongside the proposal, forcing 2 diff sets would at best only sometimes result in a spread-based difficulty, but would often just add a filler difficulty of the same difficulty level. there is no reason to force this.

point 4: ok then "solve it together" by changing/addressing the bn rule, not by inhibiting a change to the rc.
posted
i like the 3:30/4:15/5:00 idea better too

Doormat wrote:

my argument here was that if we're using the argument that filler difficulties are largely seen as obsolete, what is the point in requiring a spread? why should tv sizes be required to have a full spread when we're going to be exempting longer length maps from the same, just because they have meet a difficulty level requirement?
That's a stupid question. Filler difficulties and spreads are not interchangeable. We require spreads so that newer players can play the songs, too. Filler difficulties, in this context, are exactly the same (or perhaps higher) difficulty and complexity, and thus contribute nothing to the spread except the minimum required effort on mappers', modders', and nominators' parts.

Doormat wrote:

i think this is a very pessimistic way of looking at things; just because lower level difficulties may take considerably less time to make does not mean that the same level of care/effort hasn't been put into them. there are still a lot of people in this community that do put a lot of care and effort into making a balanced spread so that people of all skill levels can enjoy them. even then, if the mapset host doesn't want to make a lower-level difficulty, there are plenty of others in this community (guest difficulties) that would be more than willing to give their take on making something so that players of all skill levels can enjoy a mapset for a given song.
And if they put a lot of care and effort into making a balanced spread, they're free to do that. Let's just not force the ones who won't put much effort into it to do the same.

Doormat wrote:

the benefit that i've been trying to argue is that it provides more options for players to enjoy the songs they want to play to. not everybody is going to have access to all maps, or may be selective in the type of songs they choose to download, so ensuring that there is a wider selection of difficulties for players to choose from that suit their skill level is crucial, in my eyes.
And not everybody is going to be happy and willing to make silly filler difficulties that, in terms of spread, are unnecessary at best. If you want to ensure that there is a wider selection of difficulties, then you are in a great position to do so through your own maps and through the maps you nominate. Let's encourage what you see as positive here instead of outright disallowing what you don't for reasons that do not appear to go beyond niche personal preference.

Doormat wrote:

i actually think that these two go hand-in-hand. if potential ranking criteria changes have the potential to affect the rate at which maps will be ranked, then it can become a potential problem for the community: nominators will be swamped with an even bigger workload, and mappers will grow more frustrated at having to wait for their maps to enter the Qualified section if it so happens that somebody else qualified a smaller-sized mapset before theirs. getting rid of the "same song restriction" is also open to a whole bunch of new issues as well, as i'm pretty sure nobody wants to see the same song in qualified a bajillion times. these issues should be addressed and solved together, not separately.
Are you implying that you want to force mappers to map at least two difficulties no matter what because it limits the speed at which they can produce sets, effectively trying to kneecap them?
posted

Mun wrote:

Are you implying that you want to force mappers to map at least two difficulties no matter what because it limits the speed at which they can produce sets, effectively trying to kneecap them?
That is not at all what I'm saying. I'm saying that we're going to need to come to a compromise in regards to the current proposal; we shouldn't be trying to brute force the proposal in its current state when there are proponents on both sides arguing for and against difficulty count requirements. I was giving my points to (hopefully) explain why I think we need difficulty count requirements.

I do understand the grounds for which the removal of a difficulty count was proposed, and I do agree that we shouldn't punish mappers for wanting to be lazy. What I don't agree with is the notion that spreads should be optional for longer length songs, based on the current proposal: there needs to be a better progression line since playing a shorter 1:30 Insane is going to be much different from playing a 4:30 Insane.
posted
Spreads aren't optional for longer songs, the rule just changes what is actually defined as an acceptable spread. This rule says, IF you have a 4:30 insane, that IS a spread now.

If you have a 4:30 extra, that is not a spread, and it would require an insane to become a spread.

I don't see an issue with this. Are you saying just because it's not 5 minutes you would like to force 4:30 insane diffs to also have a hard diff? But then what's the point of the this entire proposal.

edit -

The entire point of the proposal is to set new minimum difficulties for certain time slots. 3:30, the minimum diff is a hard. Anything above and below is optional. At 4:30 (or 4:15, whatever), the minimum diff is an Insane. Anything above or below is optional. If you add diffs above this minimum, they must form a reasonable spread.

Having 5 minutes be an arbitrary line where you say anything below this needs 2 diffs defeats the point of the proposal to me.
posted
The proposal implicitely still requires a minimum of 2 diffs, as you can't have a spread with a single element (Oko confirmed this). I see no reason to keep this implicitely stated, as it can potentially lead to ambiguity and differences in interpretation, which is why I'm suggesting adding the rule back.

As for Mir concern about forcing EE or EN spreads where one E diff wouldn't add much (which is a pretty edge case, to be honest), I'd personally be open to moving said rule to guidelines. It should still be followed by an ample majority of mapsets, but would allow for ranking a single diff on songs that are so calm/slow that you can't add/remove objects to make a significant difference, up to BNs/QATs and ultimately community's discretion when qualified.

@Hobbes2: I think the original goal of this proposal was that, if you have a 4:30 Insane diff, you aren't forced to map 2 additional diffs (a Normal AND a Hard), but only one (a Hard or an Extra).
posted
@Hobbes2: I think the original goal of this proposal was that, if you have a 4:30 Insane diff, you aren't forced to map 2 additional diffs (a Normal AND a Hard), but only one (a Hard).
I interpreted it based on what I see when reading the proposal, not what it was 'meant to be' or whatever. If it's the case that this is what the community overwhelmingly wants, as in forcing 2 diffs below the 5 minute point, then sure, whatever.

EDIT - I feel like there's some assumption here that the 2 diff requirement is some kind of inherent requirement but it's really not. It's a remnant of when we had to have 2 diffs to even move a map to pending from a technical forum perspective. It is possible for it to not exist.
posted

Hobbes2 wrote:

@Hobbes2: I think the original goal of this proposal was that, if you have a 4:30 Insane diff, you aren't forced to map 2 additional diffs (a Normal AND a Hard), but only one (a Hard).
I interpreted it based on what I see when reading the proposal, not what it was 'meant to be' or whatever. If it's the case that this is what the community overwhelmingly wants, as in forcing 2 diffs below the 5 minute point, then sure, whatever.

EDIT - I feel like there's some assumption here that the 2 diff requirement is some kind of inherent requirement but it's really not. It's a remnant of when we had to have 2 diffs to even move a map to pending from a technical forum perspective. It is possible for it to not exist.
I think it stems from the assumption that the word “spread” implies more than one. If we’re saying “only one difficulty is acceptable as a spread,” why aren’t we just removing the spread requirement altogether? I think the pushback against the proposal also comes from the idea that we’re basically just removing marathons and saying “anything goes” as long as it meets the bare minimum requirement.
show more
Please sign in to reply.