This is a shared proposal with Sanch-KK

Our proposal aims at smoothing the difficulty of ranking sets with high-star difficulties across the different song lengths. It would promote a variety of playable content for mid and high skilled players without negatively affecting newer players.

We are aware that a thread with roughly the same idea already exists, but we’ve decided that it would be better to make a new post, as our proposal stands as a separate set of ideas, with partially different goals and different methods of solving problems.

Proposal body:

As player skill rises over time, we find it useful to redefine the borders of Expert category.
This change is proposed with intent to define, what the lowest difficulty should be in certain cases of song's drain time.


Forces to have an actual progression within difficulties for the sets which will fall under new rules and sets the applicable range.


Adjusts spread requirements by just the right amount and smooths the difference between different categories of drain time.

Explanations and examples

Currently there is a significant problem with spread rules: Ranking high difficulty maps is currently either too straightforward and can be done without consideration for mid-skilled players on >5:00 songs, or too demanding on <5:00 songs, where a barrage of filler difficulties is required to be made.

Many years ago, when the rules we are currently working with were written 7* maps were considered an absolute endgame content, while nowadays maps of such difficulty are very commonly found in 5-digits’ top plays. At this point too many people crave for difficult ranked maps and their demands aren’t being met properly.


Please keep in mind that this proposal is aimed exclusively at osu! gamemode because we’ve fine-tuned all the details specifically for it, but would be interested to see people from other gamemodes sharing an opinion on how something similar would work in their field.

Few examples of how this will affect spreads

(by “spread” in that context we mean minimal required spread):

Possible concerns

Big thanks to Sanch-KK for building upon my initial idea and organizing our thoughts.

We are looking forward to the discussion on this. Both of us will try to comprehensively answer all questions & concerns and work on unaccounted possible edge cases. Thank you for reading all of this, hope it will help change osu! ranked section to the better!

I think you could word it better when you define a "proper spread", usually it means the difficulty progression is linear.

I think what's also important is consider to reduce the 3rd spread rule "...between 4:15 and 5:00" to "...between 4:15 and 4:55", (arguably 4:50 drain) to be mapped in a single diff, imagine trying to find 4 people to map you GD because your 4:55 song ended up in 7-8 stars

i think you should lower the spread drain times to the one mania uses

If the drain time of any difficulty is...
...lower than 2:30, each key mode and playstyle must either include a difficulty at Normal or lower.
...between 2:30 and 3:15, each key mode and playstyle must either include a difficulty at Hard or lower.
...between 3:15 and 4:00, each key mode and playstyle must either include a difficulty at Insane or lower.

If you want to consider the other drafts we have flying around and want to pick up the spread part of the RC I'd be willing to arrange that btw

Drain times are obviously subject to change, as they were arbitrary in the 1st place, would like to see more input on this anyway tbh. Will think about them (and some other numbers-related stuff) more tomorrow, we've found this thread only now, oops.

Just on the 1st glance it shows that 7*+ maps account for significantly larger share of overall playcount than they should do, judging on amount of them getting ranked, but, again, will research this properly a bit later

Star rate goes brrrrr

I believe that imposing restrictions on longer maps would negatively impact many mappers. It's unreasonable to expect mappers to create multiple difficulties for a song that is five minutes or longer. Requiring excessive mapping for a mapset to be considered rankable is not practical. Additionally, it's not reasonable to demand a change in the criteria because of the skills of new players. Those maps weren't created with new players in mind and thus shall not be butchered to please them. While I may not be the best person to make this point, I don't see a strong rationale for this aspect of your proposal.

would love to see something along the lines of this and the other very similar proposals but

1. please don't bring star rating into the RC
2. please don't make people make/find spreads for marathons

i don't like the mania spread rules because i think the restrictive part of making spreads is rarely actually the normal and hard and insane, and especially on shorter songs, the lower diffs accrue a significantly higher percentage of plays than the filler extras. i would rather see something more similar to riana's proposal go through (which i plan on doing, just busy with recent events) since that preserves the low diffs, but allows flexibility for difficult maps which otherwise would require prohibitive amounts of "actual" diffs to meet spread rules

Since i'm at it, will respond to lewski and nao aswell -




Just a quick edit for an important notice, trying to close the misunderstanding gap: 99% (estimate) of mapsets won't be affected by this proposal at all. 0.99% will experience huge relief from filler experts, Normals, and sometimes Hards, which will barely affect amount of low diffs getting ranked overall, as explained earlier.

And there is also 0.01% that will experience discomfort for the prize of greater good, but all it actually does is just brings them on the same level as other people who map slightly shorter songs

sure I can elaborate

Starting with marathon spreads, the main issue is that the whole notion is counterproductive. The core idea behind all spread relaxation proposals is that we'll gain more by making it more feasible to rank some sets than we'll lose by giving those sets looser requirements. I fail to see how the reverse wouldn't be true for forcing top-level marathons to include an easier diff; we'd lose more by making these marathons harder to rank than we'd gain from the additional diffs.

In a proposal whose main goal seems to be to get more difficult maps ranked, this tradeoff makes no sense to me. The proposal itself mentions that the kinds of top-level maps it's targeting are "extremely scarce" and that most mappers will still be mapping easier difficulties anyway. The stated benefit of making certain songs more accessible is tiny in comparison.

Additionally, the justification that ranking marathons is "too straightforward" isn't something I can relate to in the slightest. It may feel too easy compared to shorter top-level maps, but that's only the case because those are currently too hard to rank. We should just tackle the actual issue instead of making longer maps harder to rank. Don't stub your toe on purpose just to forget your headache.

Finally, I want to point out that the higher the final spread cutoff is, the worse it'll feel to be on the wrong side of it. Near the suggested 10 minutes, you get as much additional required drain time as you would with a 4:59 song that needs a single expert between the insane and top diff. In this case, though, you've already put twice as much time and effort into the top diff, and the additional work is harder to divide between GDers to boot. Something like 7 minutes would be more reasonable, but the issue (as well as the more important ones mentioned above) would still be present.

edit: oh yeah for what it's worth I'd be completely fine with 4:00-5:00 requiring an expert if the mania cutoffs happened to be brought to std since it wouldn't be stricter than what we have at the moment


As for star rating, it can't really measure anything but the most mechanics-focused maps in its current state, so I don't think it's a satisfactory way to define a difficulty level. You're right in that people tend to use it as a reference at the insane/expert boundary, but in my experience, that's all it is: a reference.

The thing with insanes and experts, though, is that we've been distinguishing between them for well over a decade at this point. Star rating isn't the only metric, there are lots of others as well; cross screen jumps and spaced streams are even in the RC. Again, this is especially important when categorising maps whose challenge comes from more than just mechanics.

Meanwhile, it hasn't been necessary to distinguish between expert and expert+ so far, so, as far as I know, there's no existing consensus for what the difference actually is. The diff icon just changes colour at an arbitrary SR cutoff and that's it. It'd be great to have some actual guidance beyond "look at the number" if you want to make the distinction relevant to the rules.

Okoratu wrote:

i think you should lower the spread drain times to the one mania uses

If the drain time of any difficulty is...
...lower than 2:30, each key mode and playstyle must either include a difficulty at Normal or lower.
...between 2:30 and 3:15, each key mode and playstyle must either include a difficulty at Hard or lower.
...between 3:15 and 4:00, each key mode and playstyle must either include a difficulty at Insane or lower.

If you want to consider the other drafts we have flying around and want to pick up the spread part of the RC I'd be willing to arrange that btw

zekk thinks this is awesome

To clarify: I think including NHI when practical is more important. I believe those diffs are much more important than filler extras. So I oppose the proposals that can reduce the incidence of those low diffs. I support proposals that alleviate the burden of multiple filler high diffs.

I think including NHI is important because the point of the game is that players can play a much wider variety of songs than other rhythm games. Part of that includes new songs. The argument that there are already enough low diffs does not apply to new songs that would require low diffs, which is what I want to prevent drying up if possible.

The argument that there are already enough low diffs

it feels like you are totally on the different page with what we are suggesting tbh, cuz this argument was never made in the 1st place. Everyone acknowledges the fact that low diffs are important for the game but fail to understand that the amount of sets with extremely hard top diffs is extremely scarce, allowing people to drop some low diffs won't affect their incidence even NEARLY as much as previous easing of the spread rules (<5 minutes - full spread down to N was required) did.

ALSO this proposal aims at reducing the burden of the filler extras aswell. Explained above probably twice at this point.

Is it that hard to just... read, guys? I honestly just want people to imagine possible scenarios of different spreads under proposed rules and see that most of the time it would still require NIH, or it would still be the most logical way to map the spread. Allowance to drop NH, and in the case of lower songs I would only be sensible for maps with 8-9* top diffs, at which point, if they are mapped in a fashion that implies their rankability, song is probably not too supportive of the comprehensive and actually enjoyable Normal aswell. Anyway, even if it is, it would be an extremely rare case, in which giving this small encouragement to a mapper and helping him to promote high star content forward would outweight the loss of say NH.

The main issue I see with enforcing high 7*+ solo sets to be above 10 minutes is the density of said map causing a score overflow / high enough object count to cause Score V2 to be enabled.
Yes, this can be alleviated with lazer, but we are focusing on Stable play for now due to how leaderboards are enforced.

Take for example: Hunger Strike and The Sun The Moon The Star. Both over 10 minutes and are above 7*, however both difficulties are near the 7,000 object hard-coded "limit" before Score V2 mod is enforced. (Right now, Hunger Strike is impossible to SS with Hidden as it currently goes over the score limit and overflows just before the ending).

Right now, the proposal feels overly complex and is intended to force high sr mappers to map "gunathons" way more, causing way more effort on both the mapper, modders, BN and players all together (also loosing more motivation from those users altogether).