hi



the proposal below aims to clarify what is acceptable for Normal difficulties used as the lowest difficulties of osu! standard mapsets.

the proposal:

addition to the proposal:

(if amended, this addition would be its own wiki article linked to from the ranking criteria)

*refers to Normal difficulty guidelines on https://osu.ppy.sh/help/wiki/Ranking_Criteria/osu!

if it passes, this proposal would be stuck in the ranking criteria like so. this means rhythm-related guidelines are based around 180bpm songs and the glossary defines some complex words, as per the rest of the standard ranking criteria.

if any part of this proposal is concerning, please provide feedback in the thread. this draft will be up for community input for one week -- until February 3rd.

Bpm kinda plays a huge role in this... Yes for 90 vs 180 bpm it should be doubled or halved, but there's unclear ground on 220 bpm, 140 bpm or 1/3 snap it's kind of a gray area.

Imo should be more similar to the ctb RC where milliseconds is used instead of snaps. It draws a clear line on what is and isn't acceptable, and leaves less up to discuss on what is ok for what bpm.
https://osu.ppy.sh/help/wiki/Ranking_Cr ... tch#-salad

That aside, spreads on sets with double bpm don't really follow spread rules in the same way at all. Would be interested to see how the RC handles 260+ bpm.

try to avoid stuff like specific sv settings, even tho it states that it's assumed the song is 180bpm, people end up rly confused about it, going on to even make statements like that sv is too high for 90bpm. since there's no way to judge what sv setting should be avoided for non180bpm songs, its better to leave it a bit vaguer, with something like "avoid relatively high sv settings" or smth idk

everything else seems ok, as long as its clear these are still guidelines and not rules

tbh some of these should be determined on a case-by-case basis, similar to Sinnoh’s concerns.

they would be added as guidelines, which indeed means that they will be determined on case-by-case basis...

i honestly think that most of these will be broken at a very frequent basis (especially points 2 and 3) that it's mostly a futile effort to use these specific rhythm restrictions as a "middle ground"..

most players will be able to play current normals in less than an hour of playtime anyway so this guideline proposal wouldnt really be helping players either

given that these guidelines only apply to a narrow, even if most common, tempo range, this entire endeavor seems fruitless to me. the RC quite clearly needs to be *and is* adaptable; it's up to the BNs to determine on a case-by-case level if such difficulties are suitable for beginner players.. so where and why does this quantification come in? for music with odd rhythms theses guidelines will be broken, and some of them even encourage rhythm choices that don't follow the music in favor of vague "beginner-friendly" ideas.

the RC is becoming far too bloated with vague rules about what is and isn't good in mapping, or what is and isn't suitable for the ranked section. its discouraging innovation, creativity, and new-ness in the ranked section. these rules implicate that there is a correct and very narrow way you may map rhythms and difficulty in the lowest difficulty of a beatmap when in reality there are many approaches you may have. a very high ratio of sliders-to-circles will make maps easier, and so will sparse rhythm selections with a higher ratio of circles to sliders. for maps with odd rhythms (i.e.: songs in "1/3", songs with predominantly 3/4 rhythms, songs with literal constant 1/2 [see pitch fucker], songs in x/8 time signatures, songs far outside the common bpm range, etc.) these rules *should* be broken.

all of these guidelines have cases where it would be in the best interest of the map to break them, and will have to be treated on a case-by-case interest given the inherently subjective role the BNs play, so why are these rules necessary if they will likely never be pointed to? there is no way to quantify perfectly what a "suitable" normal difficulty is, especially when it comes to rhythm choice. difficulty values and/or slider velocity choices are far more sensible as guidelines and have far more 'objective' value.

Sinnoh wrote:

clearly scale by bpm. double bpm stuff sucks

last time this was brought up, rc guys were opposed to it. it does make things easier to follow though so it could be worth discussing again, especially since the guys below your post also think it makes guidelines unclear

dsco wrote:

.

valid concerns here. BNs should determine if a diff is suitable as the lowest diff case-by-case, i agree, though that kind of falls apart when there's nothing keeping BNs generally on the same page. the rc dictates what nominators can/can't rank, so if the rc says diffs with constant 1/2 rhythm, high spacing, and fast slider velocity are ok for rank, then there's a problem with the ranking criteria (which as you correctly said is because of how vague a lot of its content is). like for example, your hard diff for pitch fucker is technically okay as the lowest difficulty now that the 2 star rule is gone, while it's not beginner friendly considering it's uhh... a hard diff.

this proposal basically says "don't use rhythm that's too dense, spacing that's too high, or slider velocity that's too high for a conventional beginner-friendly normal" with values comparable to most normals as reference. it's not about good/bad mapping, but rather mapping that's more intense/straining than what newer players can handle. the hope for this is to not be as vague as the current normal diff guidelines, which only go to the extent of saying "use occasional 1/2 rhythms on normals" with no reference to spacing or sv, therefore allowing just about anything on a lowest diff.

inevitably, there are gonna be situations that don't align with the guideline examples (like the 3/4 based songs you mentioned), though not specifying anything about beginner-friendly lowest diffs because of those exceptions results in more vagueness. this could be somewhat better if scaling bpm were taken into account more clearly, though it definitely wouldn't be perfect

that last point probably applies to the other concern about the rc restricting too much about how to map normals now, but i really don't think this proposal is going to affect that (however other parts of the rc are definitely guilty). if a map is designed unconventionally and breaks some of these guidelines while being beginner-friendly, it could pass. if it's deemed too extreme, an easy could be added. if it's conventional, it most likely should be following the guidelines because it might be too difficult, or again, there should be a lower diff

hopefully these explanations make it clearer as to why this proposal is more than vague rc bloat. will be getting thoughts from other ubkrc people too.

Mishima Yurara wrote:

rhythm guidelines broken frequently, especially 2 and 3

most players will be able to play current normals in less than an hour of playtime anyway so this guideline proposal wouldnt really be helping players either

im not really sure about the first thing. the rhythm guidelines are based on like, the maximum rhythm density most ranked normal difficulties use, so if there are normal diffs that break these frequently with 5+ consecutive 1/2 sliders or 3+ measure 1/2 chains, i'd be interested in seeing them

as for the second thing, i also agree, but the new rule permits normal difficulties way harder than most current normals, which is where the problem is. most current normals fit within the proposed guidelines

It feels like some of these responses are forgetting that these are supposed to just be guidelines. They're not supposed to restrict creativity, and they are expected to be broken under special circumstances. These guidelines are supposed to give a general rule of thumb to what you should aim for in a lowest-diff Normal, rather than drawing hard lines on what is and isn't allowed.

Yes there may or may not be a time when spamming 1/2 sliders over two measures is suitable. But for most cases, it probably isn't, and especially so for the lowest diff in a set.

The current RC and guidelines don't really outline what is suitable for the lowest difficulty, so we needed some general middle ground for people to agree on for the regular cases, lest people start making things up and we start judging on vague unwritten rules.

Also note that even if these guidelines aren't too helpful to experienced mappers attempting to push boundaries, they will be most useful to newbie mappers who aren't sure what the expectations of mapping are.

-

Judging based on millisecond timing would probably make things more clear cut, but I feel like it would transfer less well between tempos in standard mode. It feels better to base measurements on the number of beats, and scale for double or half BPM styled songs/maps. If you wanted to use maths and hard numbers to justify why 1/1 is too small of a gap at 270 BPM then you could, but it feels less clunky and more natural to just use beats for the guidelines.

shouldn't there be something about a limit to passive 1/2 too? or does it kinda tie in with the active 1/2 rhythm? feel like if it does become part of the guideline it should be about 3-4 (4's kinda pushing it tho lol)

this is basically saying "here are the bare minimums use common sense from there" because that's pretty much what a guideline is. these rules aren't restricting anything beyond the bare minimum of what the lowest diff normal should be, and from there its the BN's judgement as to whether the diff is appropriate. so it seems fine to me

One question, how can newbie play a map if easy don't have? 1,99 is really hard for newbie

closed for revision thanks everyone

we back

pretty much everyone who commented complained about a lack of clarity when a song's bpm isn't exactly 180. because of that, the guidelines were hard to interpret for other BPMs. this concern makes a lot of sense because uh... nobody's ever explicitly shown how ranking criteria BPM scaling stuff works -- especially around middling BPMs like 135 and 270 lol

as a solution, we created a short reference sheet explaining how we think about BPM scaling. like the rest of the proposal, its content is based on the upper limit of how mappers currently create their Normal difficulties.

here it is: https://gist.github.com/pishifat/14a624e79b70d4286e43777b9d2f0d24

if y'all find this okay, it would be linked on the ranking criteria as an external wiki article like this

with scaling info publicly accessible, it should be more obvious that a diff like this is acceptable, avoiding confusion from bns and qats regarding what would get dq'd and what would be rankable

(edited the op to include this as well)

reads nice enough, so long as people of power dont get too nitpicky about 121bpm having a stray 1/4 i think it's good

i think in general this is pretty good, since there are guidelines and can bend a decent amount. however, the line about objects overlapping seems a bit odd. since you acknowledged bpm scaling, i would honestly just remove that guideline and make an even more generic one, saying “ds should be more or less consistent and similar to the map’s sv.” i think this would be better for specifically lowest diff normals, without making bpm scaling a weird issue with overlaps and sv

Just saying but the lowest star rating should be set back, be it 2.00* or 1.99* (even tough it should be a bit lower than that) but please, now the mappers won't even try to make it easy enough as they only have to throw random 1/1 sliders with absurdely high spacing and they are going to set AR3-4 because it isn't as dense as Insane, and don't even try to tell me it is not happening as I am stating things in the most realistic way. As long as the star rating is not always the most reliable source of how the maps work we should have more factors than just a few words saying what a true easy/normal difficulty is. We used to make difficulties below 3* in the old star rating system, then that was 1.99* and even then the boundries have always been extended because the easy mappers and modders would say "okay it is acceptable" but it is not going to be the case anymore as the whole community is only going to keep going harder and harder and we will see normals being 3* in around a year time. I am being serious and grave at the very moment, no one is going to bother asking if maps are easy enough, mappers will say "hey, my map is high bpm so it is okay that I have 10 notes in a row as it makes perfect sense, it is countable and it fits the rhythm!", who cares that we have to simplify the map so it is both rhythmical AND beginner-friendly. Also, as we have some people working on the ranking criteria like in the United Knights stuff (I can't remember the whole name of the group, pardon me for that), how about we actually get people who are good at making and judging easier difficulties? Of course they will argue on what is acceptable and what is not, just like Okoratu and Irreversible may argue on what is appropriate for the Insane difficulties, it's the same analogy.

Moreover, that is sad but I do not believe that there are more people like ezek or Bakari that make more than 15% of the BNG and QAT groups counted together who understand how the lowest difficulty should look and play like, and some very particular QAT members know it very well from me first-hand. Just tell me how many mappers in the BNG and QAT we have like, let's say, MkGuh, me or the ones mentioned above who actually care to provide playable maps for the beginners and are in the QAT or in the BNG. Honestly, most of the people there do not care that much about the easies and normals as much as they care about the highest difficulties because "making the lowest difficulties is a pain and no one cares about them" just as said by pishifat. So, how about we change that?

Krfawy wrote:

Just saying but the lowest star rating should be set back, be it 2.00* or 1.99*

the sr thing was mainly removed because spread doesnt depend on sr but difficulty elements which is entirely different, putting it back would cause the same problem

Krfawy wrote:

how about we actually get people who are good at making and judging easier difficulties?

mapping and modding a pretty dissociated in terms of skill just like mapping and playing, you dont need to be a good mapper to be a good modder and vice versa. the qat and ranking criteria team thing are proficient at judging ( that is why they are in that position in the first place ), so they do not necessarily need to be good at making low diffs to judge what is acceptable or not

regarding the proposal, i have trouble understanding the third point. does it mean 1/2 sliders should not have more than 4 reverses in a row, or that the guideline applies regardless of whether or not the sliders have reverse arrows ? i think wording could be a bit more precise so as to avoid any potential confusion, especially for people using the rc as reference when modding beginners who are not necessarily familiar with basic low diff design

Krfawy wrote:

how about we actually get people who are good at making and judging easier difficulties?

"how about we actually get people who cannot/doesn't want to do this and neither wants to take this job"

This is how it sounds like and not how it works. The community preferably plays Normal and Insane difficulties, nothing you could change.

Hi! As someone who's makes a lot of stuff for newer players, Id like to add my two cents.
I think as long as it does a good job of undermapping higher difficulty patterns (or setting players up for what is present in higher difficulties) while not undermining the song itself, it should be fine. I dont know if this has been discussed or not, but I don't think star rating should be included in criteria for this just because I'm a firm believer that star rating isn't the best at judging difficulty. Especially since star rating is extra sensitive for lower difficulties in my experience. Just a bit of a notice though, I haven't read the thread because navigating is on mobile but I just thought I'd throw another stance into it to help!

Veridian wrote:

Hi! As someone who's makes a lot of stuff for newer players, Id like to add my two cents.
I think as long as it does a good job of undermapping higher difficulty patterns (or setting players up for what is present in higher difficulties) while not undermining the song itself, it should be fine. I dont know if this has been discussed or not, but I don't think star rating should be included in criteria for this just because I'm a firm believer that star rating isn't the best at judging difficulty. Especially since star rating is extra sensitive for lower difficulties in my experience. Just a bit of a notice though, I haven't read the thread because navigating is on mobile but I just thought I'd throw another stance into it to help!

This proposal is a consequence of the removal of star rating from the spread rc because not all normals are equally easy to play , so if the lowest difficulty of a spread is a normal then it should not be on the higher end of the difficulty spectrum ( hence those guidelines enforcing the idea of a relatively easy normal should it be the lowest diff of a spread )

Mh.
Experience has shown me that star rating may underrate some maps but never overrate a map. The physical difficulty is the thing that star rating captures almost perfectly.
Personally I think that a 2.0* border for the lowest difficulty does make sense for the following reasons:

Physical difficulty is the first thing people struggle with because none is trained to move the mouse in such ways when they start out. While some people are more talented than others and go through Easy and Normal diffs in the blink of an eye, many people are not and these are the ones that suffer. I for once played almost 300 osu! standard difficulties below 2.0*, at least 2/3 of that when I was an actual beginner. This just goes to show that these maps are not irrelevant and that people may spend more time on them than "2 Easy diffs and then never look back" (I would have looked back at the statistics that Ephemeral posted at some point but I couldn't find it for shit and I tried).

The lowest difficulty of a set should not contain any reading difficulties that cause a notable inaccuracy of star rating. While it may be true for higher difficulties, no lowest difficulty of a set should ever contain patterns that require advanced reading and have an increased difficulty that is not reflected by the star rating.
Assuming this, star rating can be regarded as accurate with a very low margin of error within this range of difficulty. As a result a border cannot be regarded as arbitrary and setting a maximum star rating does make sense.

Physical difficulty on lower difficulties is most influenced by density. Star rating functions as a built-in judge to how the density relates with spacing leading to the respective physical difficulty (and rightfully being very sensitive to this as observed by Veridian). A fully-fledged Normal difficulty is easily twice as dense as a respective Easy difficulty, which is quite the obstacle.
Personally I have yet to see a 2.2* Normal difficulty that I would consider as suitable for a beginner.

tl;dr: All these things make me think that star rating is a better judge of the map's difficulty and as such suitability for beginners than a bunch of guidelines that may and will be broken.

I would also like to see the following guideline to carry over from the Easy-guidelines as spinners and getting your cursor back into place are just the most confusing thing ever for beginners and most notably, this is the one physical difficulty not covered properly by star rating so it should have a place in that consideration:

Spinner recovery time should be at least 4 beats. This is to ensure adequate time to click a hit object following a spinner.

Discussion about the re-incorporation of star rating into the RC is irrelevant to the topic at hand and should have been brought up when the changes to the mapset header were brought up months ago.

The objects overlapping guideline is alright as is, since it basically currently means "the most dense timeline gap allowed should overlap on the playfield". The BPM scaling guide covers when a more dense snap is viable to be introduced as well as a less dense one, but I don't think it'd be constructive to change it to say that since then we'd have to define another guideline for what the most dense allowed snap is. Reasoning for violating that guideline would be BPM dependent and should be judged by BNs with logic similar to: "what is the most dense timeline gap? is it overlapping? if yes, good - if no, not good".

Guidelines for the note densities specified in the guidelines and the external guide to scaling normal difficulties are to be enforced like rules anyways, but the reason we cannot make them a rule is due to variences in BPMs and tolerance (e.g. density for something like a 120 bpm song could easily be as such for 115-125). Enforcement will be handled similarly to how spinner recovery is handled, so I don't think there should be concerns about guidelines being consistently broken.

The current clarifications for Normal difficulties are also clearly indicated as additions to the existing criteria, as such stuff like spinner recovery for normal difficulties is already included.

Let's try to keep discussion in this thread for what is currently proposed.

It is as a fact the star rating is not always truthful, but as Endaris stated and that I agree - it is true that the star rating can underrate how a difficulty actually is and we all know how Kaede by Hollow Wings works and that it is a great example of how much the system can be flawed aka sliders are harder to be followed than you think. Still, now as the star rating thingy is no longer existent, we are of course supposed to follow the rules said yet no one really cares how hard the difficulty comes out because no one has to really test it. Also, note that of course, as some maps are faster due to the high bpm being high bpm we are going to have a little bit more demanding normals but because of that the differences between the lowest difficulty (Normal that actually plays like a Hard even though it is full of sliders but the high bpm is obviously not helping) and the next difficulty (something that is supposed to be called Advanced but it actually plays like a regular Insane due to some small jumps and triples + once again the pace of the song) people are not going to make another difficulty in order to even the difference between the difficulties, they are just going to make the lowest difficulty harder than it already is because making it even harder isn't against the rules and who cares about the normals and easies and, indeedly, now the SR issue doesn't exist as it was only enforcing people to obey the numbers.

I mean, hello, does it really sound like a scenario taken out of nowhere and out of the very context? Again, at least people have attempted to create difficulties that were more bearable for the newer players and now people are going to put more and more and even more 1/2s as the rules don't really say "try to make your difficulty as easy to go as possible and try to not put too many hardcore patterns", the rules say something similar to "don't put too many 1/2 notes in a row" so now it basically means that in a 180bpm song I can make a 1/2 slider-circle-slider, a pause, repeat the section, and again, and again and again and sometimes use three circles in a row constantly mixed with slider-circle-slider with most of the notes being circles only. And no one will even bother making more 1/1 or longer breaks if there won't be a huge backlash from the community and as the time passes people will make even harder maps because why not, it's not that hard. And now I have a feeling that people saying that the star rating is not the most valuable quantity are more likely to think of the rule being irrelevant as they don't think of what the easiest difficulty should look like but what the spread between two difficulties is supposed to be between a Normal and a Hard yet the Normal still being a regular Normal and not the easiest in the set. I still haven't seen anyone giving me any other answer for my issue than "this is how it works" and I would like to get a more comprehensible statement.

And as long as I can understand the fact that I should have brought the star rating thingy earlier it is not too late to talk about it now as we are clarifying what the standard Normal difficulty means, so it's more of "it's high time I did something" instead of "invalid, stop your mere mumble Krfawy". It is very important to mention the fact that the star rating used to have some values, back in the beginnings of 2015 people would still even recommend to map easies and normals that were closer to 1.50* values and not the ones that are closer to 2.50* In these days we are facing the latter scenario. Again, the SR is not the most objective factor yet it does have a value and we can agree that years ago difficulties were at least suppoed to be as easy as possible even if that meant that they would simplify rhythms in an extreme way... Isn't it how easier ones should work?

And still, I'd love to hear an opinion from even more people who specialize in structuring and explaining the phenomenon of easier maps because we don't have too many opportunities to hear lots of them. Of course we all know they are the unicorns of osu! but they definitely do exist as we can see on this page of the thread and I would like to thank AyanokoRin, Veridian and Endaris for voicing their opinion on the matter and being the unicorns together with me.

Stefan, I do not understand the line in quotation marks...

To me it seems more reasonable to define the lowest difficulty normal in terms of concepts used in the difficulty, rather than basing it on star difficulty which, as said, can be misleading to the true difficulty. Not only because a 2.2* map could be considered easier than it seems, but also you can have a 1.8* map be harder.

The current RC doesn't outline what should be acceptable in such a case where a normal may be considered too hard. Mainly the "Avoid long chains of hit objects with 1/2 gaps." What are we considering too long of a chain in this case? This is what we're trying to establish with this proposal.

I feel the concerns addressed are that the proposed guidelines are still too difficult for a beginner player. In which case, what would you suggest should be acceptable for 1/2 rhythms for the lowest level normal?

These are fine as guidelines. the intention is to enforce the standard type of normal difficulty here, and im sure most BNs can judge by using common sense. so it should be good.

but i also don't really like how we're enforcing how a normal diff should be mapped. this makes stuff like https://osu.ppy.sh/b/869732&m=0 rather hard to deal with

for those asking for star rating to return, i think what mo and halfslashed explained should be conclusive enough for that topic. star rating isn't a consistent way to determine difficulty because it weighs certain factors more than others. if you want normal difficulties to be simpler than the criteria here, think about propositions based on difficulty elements instead (such as less 1/2 usage, lower spacing, etc)

@kibb this is for normals used as the lowest difficulty of a set. if a set has an oddly challenging normal like that, it's fine as long as there's an easy diff too

---

setting aside star rating stuff, i'm not seeing much talk about the proposal's content anymore. most people seem ok with it? i think?

i'll leave this thread up for a few more days. if anyone wants to give feedback about additions/modifications to the guidelines or the scaling guide, now's the last chance. if not, it'll be ported to the rc shortly

pishifat wrote:

1. Passively played 1/2 rhythms that do not overlap, such a sliderend followed by a circle, are acceptable.

Glad this is included because I do it a lot.

Would like to note that if this is the case for a given Normal diff it should be done consistently. A mix of overlapping and non-overlapping 1/2 on the same timing might cause misinterpretation of 1/2 with 1/1

I still do not understand whether ' reversing 1/2 slider ' refers to multiple reverse arrows or multiple sliders, and as a result i think wording could be clearer ( after all, the purpose of this draft is to put words on common sense )

It's multiple sliders - which seems pretty clear since avoid 1/2 sliders with multiple reverses has already been published as a guideline

i must have overlooked that comment oops

as trust said, it's referring to multiple sliders. the guideline "Avoid 1/2 sliders with multiple reverses" on the current normal diff ranking criteria should make that understandable

it could worded as "this includes 1/2 sliders with reverses" but idk if that's any clearer. both mean the same thing to me lol

Yeah i agree that both mean the same. My bad for not taking in account the rest of the normal guidelines !

Can we make it so that this only applies if said normal diff is above 2.0 stars?

This will be a good way to deal with lower bpm imo, since we should be able to be more flexible on low bpm

a bpm flexibility thing was just added
and the whole point of this is to get sr metrics out of rc bc they're terrible

alright looks like we're ready

bns will be getting a forum pm when this is merged to the wiki so look forward to that!!

https://github.com/ppy/osu-wiki/pull/1108 thx oko we're almost there

merged boys