forum

[Proposal / Discussion - osu!] Reviewing the Difficulty Guidelines for Expert(+)

posted
Total Posts
15
Topic Starter
yaspo

Current RC


Currently the difficulty Guidelines for Expert(+) look like this:

Ranking Critera wrote:

Difficulty setting guidelines
  1. Approach rate / Overall difficulty should be 8 or higher.
  2. HP drain rate should be 5 or higher.
  3. Circle size should be 7 or lower.
Looking back through the RC's github history these predate 2017 other than the CS guideline which was added here. With the recently disqualified HP0 map some of us, including me, took note of how these are only rarely enforced and mostly ignored.

Thus it'd be worth revisiting these archaic guidelines.


Proposed Changes


Here is what I propose for a baseline of discussion, I may edit this along the way to keep it up to date with where discussion is at.

Proposed guideline wrote:

  1. Overall Difficulty should be 8 or higher.
  2. HP drain rate should be 3 or higher
For anyone unfamiliar: "Guidelines" are to be seen as "Rules with exceptions" when it comes to the Ranking Criteria. In other words, above is what I'd propose to always be true unless there's a very good reason for it.

For Approach Rate we've seen pretty much the entire spectrum at this point. Approach Rate shouldn't be unrankable by itself for Expert(+), so it should be removed.
We've also had some CS8 through 10 maps come up, similarly those were accepted by the mapping community (as far as I know). So, it also makes sense to remove it in hindsight.
Overall Difficulty seems a given, anything below 8 should be too lenient at this difficulty level.


HP drain rate


HP drain rate is a more contentious topic, given by the thread on the previously mentioned map, so the value I put is a placeholder.
For the main reasoning behind this proposed change, I'd like to refer to a reasoning provided on the thread.

HP Reasoning wrote:

From a game design perspective, having an HP system at all means there is intent for proper fail and pass states/conditions.

In rhythm games specifically, getting your first "clear" on a chart or difficulty level can be fairly special, in osu's case it's what allows you to set a score to begin with.
So from that pov, unnecessarily low HP does the polar opposite and devalues the entire intent of these systems.
For HP systems I found this to be the most relevant support of setting a baseline at all. Discuss whether you support setting a baseline and if so, what value it should be.
Reminder that as a guideline, this could still be broken based on very specific map designs after proper discussion.

---------

Since this is technically about multiple topics, please don't just +1 without mentioning what you're referring to.
Ideally keep your posts to the point.
The map linked is an example but not the topic of discussion thank you very much.
Stompy_
Im fine with the OD part and stretchable AR

but HP 3 seems a bit too low for a difficulty called Expert because HP along with other parts of mapping also provides a level of difficulty.

Maybe it would be nicer to have a ''gimmick difficulties'' or ''conceptual difficulties'' part in rc instead?
fubuki
@stompy
In my opinion, seperate rc for gimmick or conceptual maps is overcomplicating. These maps have reasons good enough to break guidelines. Additionally, the terms "gimmick" and "conceptual" are broad and could be interpreted differently from person to person.

I also agree that the OD and AR change is a positive change. Also I agree with the CS part because I feel like this doesn't need concrete reasoning from the mapper. This is preference of the mapper and the nominators pushing such sets should consider if the map actually works with the CS its given.
Topic Starter
yaspo
If the purpose to add such categories is to have a higher baseline but allow generally lower values, then it can potentially be the exception to the rule -- there is room to clarify this for Expert specifically if desired.

Adding an entire category implies there's more to be said about "gimmick difficulties" or "conceptual difficulties", which doesn't feel like it's the case?

Good point on considering exceptions though

Let's not discuss this further if possible, it feels wildly tangential
Shurelia
+1 with stompy saying 3 is too low

I'd say make it at least 4 for the minimum
lewski
fully agree with the rationale behind the OD/AR/CS changes

as for HP, while I don't like being booted out of maps, I also can't disagree with the idea that a fail state does have value; passing with an abysmal score is always gonna feel cheap. people like me should just use NF and/or petition peppy to port taiko HP to other modes

it'd be interesting to see what kind of effect different HP values have on the actual drain in different kinds of maps (preferably a lot of them if feasible); it'd be nice if the number that ends up in the RC was based on actual data, although tbh I'd also be content with just the opinions of people who are knowledgeable about the mechanic

(I feel like ideally the guideline would be concerned with what the drain's actually like in gameplay but the mechanic seems way too arcane for that)

fwiw I do at least agree that the value most likely needs to be lower than the current 5 since it's apparently common for dense maps to be brutal at HP5 which isn't ideal
ColdTooth
+1 on the proposal, and agreeing with Stompy/Shurelia.

I personally like using moderate to moderately high HP on my maps, 5 is pretty much the most normal HP rate I've seen from maps from the hard difficulty to whatever is the top difficulty. Anything higher than 6 or 7 is maps with a higher sense of failure upon missing a couple of times in a short period, which adds excitement to when you finally pass or even FC it. I agree that AR is subjective to whatever the mapper's style/song is, since they are just simply reading-based.

Happy to see changes like this for the future.
Okoayu
these guidelines were originally called "difficulty setting recommendations", they dont even work well as guidelines as there's no distinct reasoning other than skill issue for you to break them on insane and up.

imo just remove them and state that HP drain must not trivialize maps would be cleaner at this point in time
could also state that OD should not cause notelocks cuz the mechanic sucks

i think that'd be better than trying to work with these guidelines under the assumption they are to be followed like guidelines

could also be applied for all other difficulty levels or maybe rename them to recommendations for ENH because that's what these are - if you significantly break them on ENH you'll get a mob of angry people either way
Topic Starter
yaspo
think I would like some more opinions on "difficulty setting recommendations", it sounds cool since most of the diff-specific rc is codified recommendations really

though, I'm a bit wary of the pattern Guidelines sometimes go through:
person breaks x guideline for no real reason -> gets called out with proper reasoning -> it's just a guideline it doesn't matter -> "what's the point of even having it written down then"
ie. making recommendations in our criteria for ranking seems like it'd backfire

as for the HP value suggestions:
it seems that so far people want higher and not lower, from there maybe a line in the direction of
HP drain rate should be 5 or higher, it can be lower if the map is otherwise too punishing

at which point I should do my due diligence and quote Mismagius' thoughts
Topic Starter
yaspo

Mismagius wrote:

The ranking criteria and the overall mapping point of view for these features should always be seen from a contextualized point of view that takes into account what is the current player consensus on the subject, especially when we are dealing with 6*+ maps.

Think of it as choosing HP and OD for an example map that includes spaced streams. As much as mapper's preference is a thing, there's also the viability of such a choice. You cannot choose OD lower than 9 for such a map because of the way notelock will pretty much force the player to fail if the timing windows are so low as to overlap, and in the same vein, a higher HP drain on a map that includes many hitobjects in a row that could easily all be missed due to a game mechanic will feel very frustrating.

There are also, from my experience, many players who will flock to high difficulty maps with a low HP drain because it gives them a sense of accomplishment, as fake as it may be, from passing such maps ..

.. my two cents are that HP should be reconsidered more as a game mechanic than just in the scope of ranking criteria. Of course, while we wait for Lazer replacing stable in the mapping scene, we'll have to adapt to how things are .. History has shown that this attracts more players than repels them - look at maps such as Airman and Fort's Highscore, for example. This is very much a non-issue to most players other than the ones who are specifically annoyed by deliberate choices from mappers looking to walk astray from the standards in ranked maps.
which suggests there absolutely is room for a value that is generally seen as low for hard maps

pieces taken from this discussion post
Okoayu
yea my point was more that difficulty settings largely depend on the map, oftentimes explain themselves and the arguments to be made for specific values in OD / HP are usually very weak.

OD can be justified with the style of map you're making, AR can be justified with the style of map you're making and the only thing that HP should be doing is ensuring that the map isn't absolutely trivial

We walked back the recommendation part originally because we thought it'd open the door to a lot of stupid stuff, but honestly for ENH, these guidelines usually work and if you want to break them people expect you to provide a ton of reasoning

for anything insane and up it really depends on the map you're making so that's why i suggested to generalize that HP shouldn't trivialize maps (i.e. check this on your own) depending on how a map is set up HP10 is usable, depending on how the map is set up hp4 will feel punishing, the way it currently works is not very intuitive

writing HP of all stats down to some value IMO isn't the right approach if we want to clarify what we want here i think it's better to go with a general guideline as the HP setting applies to all difficulties albeit in different ways

/2c
Dabbe_01
As much as I commend the proposal for trying to adjust Expert(+) difficulty settings guidelines to better fit the current mapping scene, I feel it's still lacking enforcement of putting HP drain on maps very prone to being abused for medal hunting (as presumably intended based on what caused this discussion to arise) as the mapper could simply provide an overengineered paragraph of little more substance than "I want it to be that way". While I understand mapper's intent should be considered and opinions on whether medals should contribute to rankability are split, at least a middle ground may be found regarding cases like the map in question

I'm likely in the minority here, however I'd go as far as adding a rule, not a guideline, to Expert(+) diffs to limit abuse potential for obtaining medals from them in the future (already ranked maps that may allow to get them more easily cannot be changed anymore, obviously). Like in the initial state of the proposal, it's just a vague suggestion that can be edited after discussion:

HP drain rate must be 2 or higher

Rebuttal to a few responses this may receive:
  1. "What about CS/AR/OD/HP0 or otherwise gimmick maps?"
    The only two instances of all settings set to 0 I recall are Gourmet Race and I'm Humanity, neither of which exceed Extra difficulty and this gimmick would be basically unplayable on Expert(+) difficulties after a single misplay due to notelock/recovery issues. Furthermore, I don't believe other gimmicks require handholding the player with such low HP values
  2. "What about the peak difficulty maps?"
    None of the currently ranked 11*+ are set below HP2 (the lowest of them is Xeroa at HP2.5) and all of them have at least 1 NM pass (disregarding different input methods for now). Most of these were set around/exactly at 3, with outliers close to this difficulty range such as Circus Gallop at 6, granted its star rating comes mostly from speed cap removal that happened in 2021 but the point still stands
  3. "What if the mapper intends their map to be trivial to pass because xyz?"
    It doesn't work in my opinion (and a few others in the thread) due to aforementioned pass medals and the HP system existing for a reason, as described in the initial post - to meet the conditions of passing the map and potentially reward a player with gratification for it respectively to their skill level, or even a medal for achieving new SR milestones. If I were to decide, 8*+ medals ought to have a hard lower limit for this reason, on top of common sense regarding more gimmicky high SR maps such as Oshama Scramble and further increasing HP on them if they turn out trivial to pass (like this did with HP3 in qualified)
Mizunashi Akari
just going to point out that difficulty settings guidelines below extra are actually useful for "teaching" the player the right way to play the game.

Obviously I dont care about anything > extra but low HP on a hard most likely can encourage bad playing habits.

also I'm just gonna say that the medal discussion was entirely tertiary and is already easily abused anyway. turning this guideline into a rule just seems like complete overkill especially given how black box HP is as a system.

and can we get some actual Players in here and not just Mappers because it would be nice to get some input from the people these changes indirectly affect.
-Mo-

Okoratu wrote:

these guidelines were originally called "difficulty setting recommendations", they dont even work well as guidelines as there's no distinct reasoning other than skill issue for you to break them on insane and up.
I do wonder if we should keep the RC as a resource for new mappers to learn what are the recommendations for mapping, or have it as a criteria solely to state the barebones criteria needed for quality/playability. On one hand the RC shouldn't be too restrictive or too elaborate, but there should probably be some sort of resource for new mappers to learn what the recommended settings are and why.

The lower diff guidelines have a bunch of this stuff already, like mentioning time-distance equality and slider velocity, but even those could be consistently broken under very special circumstances.

I realise this is not super related to this proposal though.
Okoayu
will address in diffspecific draft for rc cleanup
Please sign in to reply.

New reply