forum

[Proposal] Increase AR/OD cap from 10 to 11

posted
Total Posts
54
Topic Starter
yaspo
Reasoning:
As a result of pushing player skill level, we've been seeing maps like this and this going for ranked with extreme bpms (300 and up).

Maps like this are hindered by the following in-game limitations:
  1. Approach Rate (AR) being capped at 10. Due to this cap, the amount of objects on screen becomes very high and make the maps difficult to process - even when that is not intended by the map's design
  2. Overall Difficulty (OD) being capped at 10. At extreme bpms the window for notelock becomes smaller and smaller, quickly becoming too punishing and making it hard to get consistent attempts
We've also been able to observe players dealing with AR and OD above 10 at high degrees of competence for a long time now, so it feels like there's no need to preserve these limitations.

Discussion points:
To not jump the gun, we should first discuss if increasing the AR/OD cap from 10 to 11 is beneficial at all and if there are counter-arguments against it.
Later we can worry about relevant RC changes.

note: yes we've been confirmed with ppy that implementation should be quick work
AJT
copypaste from github post:

honestly this is needed even in stable if that's even possible. maps are quickly becoming too much for AR10 to support their densities and being ranked anyways which is bad from a future-proofing standpoint

examples: beatmapsets/1263883#osu/2627024 https://osu.ppy.sh/beatmapsets/1201974#osu/2586420 beatmapsets/1390177#osu/2870806

additional thoughts. this is straight up NECESSARY for such maps. If needed you can force max AR/OD to be 10 in guidelines, that way you'd only be able to exceed it if the map truly needs it (which should be quite obvious). I don't really see any substantial counterarguments that would mean that this is a bad thing to add - people putting AR11 on random graveyarded maps that don't need it would probably actually be more of a plus than a minus since there would probably be people that want to practice it without McOsu/osutrainer
Mordred
settings up to 11 would be cool but at the same time I don't want more excuses to rank low effort 12* bait
Ryuusei Aika
as a 12* mapper myself i can agree with yaspo & ajt since this would benefit the readability & accability of anything 8*+ with 270+ bpm

i also don't think increasing the ar/od cap would be "not beneficial" since we have more room to choose not less
UberFazz
just wanna emphasize how important this is as someone who's playtested these types of maps a lot

the experience gets 1000x better (especially with ht, which is how most people will be playing the maps) cuz of density, bpm, blah blah blah we've been over this, just read yaspo's post that explains it perfectly

tl;dr:

yaspo wrote:

While 360bpm isn't listed, we can just double the number for 180bpm instead. 180bpm at AR10 has a density of 2.7, so 360bpm at AR10 has a density of 5.4. This is the same as AR8 on 215bpm or AR9 on 270 .. really dense reading territory.

Closer home for fast aim players, let’s look at DT ranges. At 270 bpm, a regular anime map + DT, AR between 11 and 10.3 means from 2.8 to 3.6 circles visible at any given time. That’s an entire 2 circles lower than this map’s density. Ouch.

You only reach similar numbers on .. 270bpm AR9 + DT or ~405bpm AR10.3, since DT only speeds up the game but doesn’t change density. On AR11 everything is lower (technically reaches 5.4 at ~540bpm).
MAKE IT HAPPEN
VINXIS
unrelated to The Discussion Starter Map i also think it would be beneficial for other maps at similar difficulty level from 8*+

IMO the only issue that comes from this is regarding balancing the AR/OD buffs at 11+DT that will definitely be used by some select mappers to bloat PP **tremendously** and shit
Mir
yeah i think this is 100% necessary as the skill ceiling for this game keeps getting higher and more and more difficult/dense maps inevitably get pushed forward

also @Mordred i get what u mean but those higher diff settings meaning more 12* bait isn't the full picture imo. the higher diff settings allow for actually good higher density maps as well

but yeah in the long run this just seems like a logical step forward
Ryuusei Aika

VINXIS wrote:

unrelated to The Discussion Starter Map i also think it would be beneficial for other maps at similar difficulty level from 8*+

IMO the only issue that comes from this is regarding balancing the AR/OD buffs at 11+DT that will definitely be used by some select mappers to bloat PP **tremendously** and shit
i have came up with the "pp bloat" issue too but i guess since bng has officially cancelled pp maps this theoretically should not be a big problem
apollodw
as mentioned in the op, being able to play AR10+ and OD10+ has become a common skill outside even top top players for *years* now (easy example is ar10.3). we're kinda missing a demographic of specific maps in that sense, since that's a skillset only accessible with mods

as for higher star maps then yea, upping the cap on those settings is a MUST lol

that being said, is implementation really *that* easy? what effect would HR have on an AR10.5 map? should HR remove the hard cap on AR10 OD10? if not, then what do we do about legacy scores? i know i'd personally opt for a secondary HR mod that lacks any caps (accessible just like clicking on DT twice to get NC?) but suddenly, that doesn't sound like "quick work" to me
VINXIS

Ryuusei Aika wrote:

VINXIS wrote:

unrelated to The Discussion Starter Map i also think it would be beneficial for other maps at similar difficulty level from 8*+

IMO the only issue that comes from this is regarding balancing the AR/OD buffs at 11+DT that will definitely be used by some select mappers to bloat PP **tremendously** and shit
i have came up with the "pp bloat" issue too but i guess since bng has officially cancelled pp maps this theoretically should not be a big problem
theoretically

either way tho i agree that its a good idea to go forward with it
Nyxa
100% agree, it allows for mappers to have a bit more freedom in the songs they map and how they choose to map them.

The only counterargument I see is potential abuse, but those maps should just be gated. I know it's not that simple sometimes, but we shouldn't prevent the addition of a good feature just because people might abuse it. Or by that logic you'd just remove pp altogether.
VINXIS

mrowswares wrote:

that being said, is implementation really *that* easy? what effect would HR have on an AR10.5 map? should HR remove the hard cap on AR10 OD10? if not, then what do we do about legacy scores? i know i'd personally opt for a secondary HR mod that lacks any caps (accessible just like clicking on DT twice to get NC?) but suddenly, that doesn't sound like "quick work" to me
Could just use replay/osu file format versions to keep previous scores in tact i think so that way the game knows the difference of when AR/OD 11 was added
Belladonna
.
Shmiklak
agree with the fact that those settings limitations should be increased as all mentioned high sr sets would definitely benefit from it
zPeanut
how would this affect HR though? does HR now go up to 11 too?
Omgforz
edited opinion on this change:
yeah go for it, tbh. i think there's only a net gain here.


probably talking for the majority of players here by saying that if the hard cap is changed to 11, HR shouldn't multiply OD and AR by 1.4 as it is doing right now

any ~AR8 and higher map just becomes really stupid to play with HR and you basically can't play HR without learning AR11 which already is a niche skill set. which means almost all 4+ star maps become insanely inaccessible for the majority of the player base

so there's 3 choices we prob have
1) hard cap anything with HR at 10 (so ar10.5 base becomes ar10.. might not be ideal and counterintuitive tbh)
2) have a 2 part function (ar<=10 and od<=10 stay as current HR, any value above 10 is unchanged. ar10.5 stays ar10.5 with hr for example)
3) have a 2 part function (ar<=10 and od<=10 stay as current HR, any value above 10 gets scaled to 11 because it's multiplied by 1.4)

i'd personally find the 2nd solution the most interesting especially for tourneys, while the 3rd prob makes the most sense for now
apollodw

Omgforz wrote:

2) have a 2 part function (ar<=10 and od<=10 stay as current HR, any value above 10 is unchanged. ar10.5 stays ar10.5 with hr for example)
apologies if the HR discussion has derailed the post but i think this is ideal IMO
Topic Starter
yaspo
I think being concerned for the implementation of difficulty-increasing mods and brainstorming a bit is fine. People new to the thread should definitely have their focus set on answering "is the increase of this cap a good idea", though. One answer at a time.

One lamer alternative could be to lock mods when AR/OD is above 10 on stable and leave Lazer™ as a testing ground for actual implementations.

The "pp bloat" concern is there, though I think that's something we can eventually cover with RC. In current time you'd only do AR/OD above 10 on maps that are eligible for such extremes, which means it can be covered under difficulty guidelines.
Belladonna
.
lystia
this would be a godsend honestly so many maps have their playability killed by the forced ar10 cap and they're becoming more common by the day so. yes
]hamham[
100% agree that AR11 and OD11 would be very, very beneficial
attempting to map anything at 300bpm+ while also making it readable is a challenge, and there are several songs readily used in other rhythm games (e.g. megasphere - inferno at 340bpm) which fall into that category, but are therefore close to unmappable in osu! (at least std) while also keeping the map playable, due to the limitations in place

i also do realize that implementation would have to be carefully considered, not just plopped into the difficulty sliders in "song setup" as one extra tick, due to the way difficulty-increasing (and even -decreasing, to an extent) mods function as of now
but as the saying goes, we'll cross that bridge when we get there
kerupt
The benefits are pretty clear as has been outlined, even for players outside of the few who are capable of playing these sorts of maps (Due to HT as stated earlier). I'm kind of curious as to what reason is there to not do this if implementation is that easy? If mappers abusing OD is a concern, that's what BNs are supposed to be for, yea? :P
Belladonna
.
UberFazz

flower garden wrote:

kerupt wrote:

The benefits are pretty clear as has been outlined, even for players outside of the few who are capable of playing these sorts of maps (Due to HT as stated earlier). I'm kind of curious as to what reason is there to not do this if implementation is that easy? If mappers abusing OD is a concern, that's what BNs are supposed to be for, yea? :P
this, I quite literally can't think of a single reason on why this shouldn't be added.
only concerns ive seen so far are with how hr could interact with ar over 10 + possible abuse in terms of pp

first concern was summed up here. i agree w/ others saying that the 2nd option is the best (and seems like the most obvious one)

regarding the pp issue, we'll likely have guidelines (3rd paragraph) to make sure that doesn't happen

anyway, agree that the positives strongly outweigh the negatives
Akiyama Mizuki
100% supporting this for the reasons the people above have already said
LwL
Fully support it for all the reasons mentioned provided there's a guideline to keep it at 10 or below to prevent abuse and the already ever increasing AR on maps to become even higher, really don't need random 7* 250bpm maps to suddenly be 10.3 and even less than that we need tv size farm maps with base od11 for pp to the moon on dt.
Belladonna
.
Xilver15
Personally I don't find instances to go above AR10 even on 300bpm+ maps but if many want it, why not. I've always supported changes which force players to get better. I HIGHLY stress the ability to raise OD to 11 though, without the usage of DT. Since HR suffers from OD being capped to 10 in terms of pp potential, it would help shake up the meta a bit more if implemented which I assume many would support.

UberFazz wrote:

only concerns ive seen so far are with how hr could interact with ar over 10 + possible abuse in terms of pp
I like forz's 2nd suggestion to keep ar<=10 the same, but I wouldn't mind if it goes to 11 aswell. However I do NOT think OD needs that cap, it's not as detrimental as AR when it comes to playing experience. Many HR players who can reach high 99s on some maps find it very limiting to not be able to push the OD past 10, and I agree with them too.
Pennek
yes please! competent mappers having this control allows for more varied maps and im all for more neat stuff

edit: also please link the github thread in the main post @yaspo
beeemmmooo
Is the value of 11 specifically picked to match the maximum effective AR/OD available with Double Time enabled or just for convenience or both?

Just want to point out that max effective OD available in stable client is 11.0833333..., rounded to 11.1 or 131/12 to be precise
LwL

flower garden wrote:

LwL wrote:

really don't need random 7* 250bpm maps to suddenly be 10.3
this is... a good thing, though?

edit: correction, depends on what style of map, but in most cases I think 250bpm 10.3 is fine.

I personally think even 10 is kinda high for that, depending on the map though. For some super dense tech bs 10 or mayb even higher seems perfectly fine, for some random jump map I really don't see the need. There's been a recent (by which I mean a few years really) trend that everything is high AR and I absolutely despise it.

EDIT: Also the more general effect of this would just be gating these maps from a massive amount of players because ability to read that high AR isn't exactly common outside of like 3 digits (whack-a-moleing some jumpspam map does not qualify as reading).
Belladonna
.
Houtaro
agree! banger idea, 100% support
LwL

flower garden wrote:

i don't think you have much experience playing high bpm then, because 10 is certainly not enough for a lottt of maps that are ranked at high bpm, high ar is also not a bad thing either, it's fun and rewards players for faster reading.
I've been playing high bpm maps for like 5 years and have a decent amount of 250-300bpm maps in my tops what

if you look at currently ranked high bpm 7* you will find that many of them are not even ar10 becaues it really is not needed.

And I can say the same thing about low ar, it's fun and rewards players for being capable of reading dense maps.
beeemmmooo

LwL wrote:

flower garden wrote:

i don't think you have much experience playing high bpm then, because 10 is certainly not enough for a lottt of maps that are ranked at high bpm, high ar is also not a bad thing either, it's fun and rewards players for faster reading.
I've been playing high bpm maps for like 5 years and have a decent amount of 250-300bpm maps in my tops what

if you look at currently ranked high bpm 7* you will find that many of them are not even ar10 becaues it really is not needed.

And I can say the same thing about low ar, it's fun and rewards players for being capable of reading dense maps.
LwL much as I personally would like people to learn how to read density mapping more I don't think 7* is really the range this change is intended for
Belladonna
.
Abyssal
I think currently is the best time to make this happen. There's always been a need for it at certain points in time, but never more than now.
omgforz' "solution" to the HR issue is really solid & from "pp abusability" standpoint it could easily be countered with ranking criteria changes (e.g. subjecting maps with higher AR/OD to more scrutiny in the ranking process and only allowing it when there's a *clear* need for it, rather than getting hit with the "subjective. closed.")

i think overall there's more to gain than to lose from this change.
LwL

beeemmmooo wrote:

LwL much as I personally would like people to learn how to read density mapping more I don't think 7* is really the range this change is intended for
Yes that's exactly why I said what I said

Change is good just make sure preemptively that it doesn't end up changing prevalent difficulty settings on maps that are perfectly fine as is
R3m
Honestly agree with the majority being said here.

With the rise of more maps that would benefit a lot from a higher AR/OD value (mainly better readability for high object density maps), it seems like the correct time to push a change like this in my opinion.

That and as mentioned by previous posts, AR/OD 10+ has become quite a common skillset within the higher ranks, so it is definitely not out of reach for players either.
Purplegaze
Something really important that I haven't seen pointed out yet:

Setting OD to 11 will not help with notelock anywhere near as much as they do on DT, if AR and OD use the same formula from 10-11 as they do for 0-10.

Notelock occurs because of the super large 50 window of the previous note, which effectively decreases much more on DT than it would with a nomod OD increase since it's just dividing by a constant.

The formula for the duration of the 50 window before and after a note is:
floor(199 - (10x)) + 0.5

50 window at OD9 nomod: +/- 109.5 ms (~ 274 bpm 1/2)
50 window at OD10 nomod: +/- 99.5 ms (~ 302 bpm 1/2)
50 window at OD11 nomod: +/- 89.5 ms (~ 335 bpm 1/2)

Compare with the notelock windows at common ODs with DT:
50 window at OD8+DT (~9.7): +/- 79.7 ms (~ 377 bpm 1/2)
50 window at OD9+DT (~10.3): +/- 73 ms (~ 411 bpm 1/2)
50 window at OD10+DT (~11): +/- 66.3 ms (~ 452 bpm 1/2)

As it currently stands, playing high bpm DT maps is much, much better of a gameplay experience because no matter how high you turn up nomod OD to, notelock is way way less forgiving on nomod. Above 335BPM, even on OD11, literally any early hit on a jump section will result in notelock. Faster maps, such as Psychopath Justice, which is 366bpm, will still be unplayable even at the maximum setting, unless the player purposely hits everything late, adding even more control over cheesing the game mechanics on top of the initial difficulty of being able to play 12* jumps.

I agree that increasing AR cap will help immensely with how easy maps like this would be to read, but it's important to note that they will still be incredibly frustrating for anyone at their skill level to play if notelock is not reworked.

I don't know how hard it would be to convince peppy to change stable notelock to work the way it does in lazer, given that it's a much more difficult change to implement than AR is, but if you managed to convince him to increase the AR/OD cap, maybe there's a chance.
7ambda
Can we get confirmation on whether this is more of a bandaid fix or permanent solution? Purplegaze has a valid point, but yaspo mentioned any remedies to notelocking should have quick implementation. If the answer was as simple as porting Lazer notelock to Stable, I doubt this thread would've been posted for discussion.

Yes this could be done to mitigate Stable notelock in any sort of way, but perhaps not consider it (OD11) as resolving the issue.
Topic Starter
yaspo
This is not a "fix notelock" thread. What specifically has an easy implementation is changing the AR and OD caps. Everything else is a question-mark unfortunately, mods included.
Lazer has genuine improvements on notelock, so once it's released we can put that issue behind us.

Regardless, what purpleglaze says seems valid and a very useful point of reference. In terms of pushing skillsets raising the OD cap to 11 still seems good (see Xilver's reply for example).

In any case, everyone seems agree that increasing these caps is a good thing. Beyond that we can now discuss if any changes need to be made for the Ranking Criteria.

The following are things that should be covered in my opinion, feel free to add your own suggestions
- AR10+ shouldn't be allowed until above a certain BPM (~290-300), to not AR-gate players until the map necessitates it
- OD10+ nomod shouldn't be allowed until certain "Extreme" levels of difficulty, to fit general skill progression and prevent excessive OD inflation

Implementation/change to mods is probably tentative, will communicate with ppy
]hamham[
rapid-fire first thoughts

***

@yaspo's suggestions

not sure how much good tying ar 10+ to bpm would do, as that could simply be circumvented via bpm doubling, so to do that would require banning doing so from the rc, which is probably next to an impossible task
it also raises the question of variable bpm maps (say, deltamax) and whether they should be allowed to use ar10+ if a given section reaches 290bpm+, although the rest (potentially a vast majority of the map) may not

i'd personally tie it more to note density, like "if there are more than x hitobjects within y milliseconds of each other, ar10+ is fine, otherwise it's not"
although i recognize that's difficult to measure manually, especially for longer maps, so this would probably require implementation into aimod

i'd probably tie od10+ to the current cutoff for the "expert+" difficulty level (i.e. 6.5*+), for sake of consistency with currently existing definitions
although this might be a bit low of a limit

***

i would also like to point out another mod that could be considered a spanner in the works: namely, "easy"
even though the notion that using said mod on any map with a difficulty of "insane" or higher makes the map more difficult rather than less is so commonly known and accepted, it's practically baked into the game at this point, the difference between playing an ar/od 10 map with +ez and an ar/od 11 map with +ez would be ludicrous

at present, ez makes ar/od 10 ar/od 5 (equates to "approach time 450ms -> 1200 ms", 2.666x increase), and would make ar/od 11 ar/od 5.5 (using current formula, "approach time 300ms -> 1125 ms" if i calculated correctly, a 3.75x increase)
to put it another way, using +ez, the difference of approach time between ar10 and ar11, a reduction by 33.333%, would become a reduction by 6.25%, which, especially considering how long the approach time is at ~ar5 compared to those measured at ar10/11, would render the change close to unnoticable using the mod and leave the maps completely unreadable, save for a few niche players
this is to say nothing about the od change, which would be a headache and a half

so yeah, if we're looking at mods, i think it's probably worth looking at +ez

***

(obv feel free to point out if anything i said is stupid and dumb/unreasonable, i'm pretty new to this whole "talking about ranking criteria" and even "discussing mapping in a public space" thing)
Topic Starter
yaspo
with the maps applicable making this amount of fuss over EZ seems really silly

as you said there needs to be a prerequisite note density similar to 300BPM+ 1/2, at that density AR5 to AR6 ranges between 12 to 10.6 objects on screen at a time. Basically AR0 territory (I think) on top of something that's insanely hard
I'm not sure how the introduction of higher ARs makes this more "completely unreadable" than it already is.
It's also not like EZ suddenly stops functioning, for smaller differences in time windows it just does smaller changes - shocker I know.

ie. this consideration is completely impractical so we're skipping it in further discussion

1. BPM changing issue is valid
2. double BPM is valid, BPM doubling however is not because wrong timing is unrankable!

For 1. the suggestion can just be a guideline, so exceptions that are close to the intent of the rule can be treated on a case by case basis. This also addresses any "289BPM doesn't count lol" cases
For 2. just defining it as "note density equivalent to <x>bpm 1/2" should work, having a relatable definition over a technical one is a bit easier to take in and evaluate

I'd like to avoid tying OD to strict star ratings, but for the generic-anime-NM1-equivalent we're probably looking at 7.5*+ or something ? I'm not sure
apollodw
i think the OD line seems entirely reasonable, 7.5*+ seems about right to me

that being said, i think for AR10+ i'd opt for a lower BPM threshold - my reasoning rests on the fact that AR10.33 is very common on AR9 ~180BPM + DT maps. that works to around 270BPM, which sounds preferable to me atleast. would there hypothetically be any reasons against moving the threshold down to 270?
Topic Starter
yaspo
Personally would say because AR, in this intent, shouldn't be used to replicate DT maps. Rather, it's necessary to maintain a manageable density. AR10 on 270bpm gets a tad denser but it's far cry from actually needing higher AR.
]hamham[

yaspo wrote:

with the maps applicable making this amount of fuss over EZ seems really silly
fair i guess, i tended to play with ez a bunch so maybe that's why i felt it was worth making fuss over

although i do feel i was misunderstood basically entirely, so i'd like to clarify my post, even though the issue mentioned is only tangentially related to the introduction of ar/od11 and the discussion is, indeed, not worth continuing here
(gonna spoilerbox it to make it more easily ignorable)

zzz

yaspo wrote:

Basically AR0 territory (I think) on top of something that's insanely hard
yea that's my problem

yaspo wrote:

I'm not sure how the introduction of higher ARs makes this more "completely unreadable" than it already is.
it doesn't
it's just that the maps applicable for ar/od11 highlight the fact that ez is flat-out unplayable past a certain difficulty level, and that increasing the ar from 10 to 11 does very little in the way of mitigating that, if it does anything at all
meaning the introduction of maps in the difficulty range where ar/od11 would be applicable to the ranked section will probably require an ez revamp as well later down the line

this would probably need its own discussion though, will make a proposal if i have time

***

yaspo wrote:

For 1. the suggestion can just be a guideline, so exceptions that are close to the intent of the rule can be treated on a case by case basis. This also addresses any "289BPM doesn't count lol" cases
yep good idea

yaspo wrote:

For 2. just defining it as "note density equivalent to <x>bpm 1/2" should work, having a relatable definition over a technical one is a bit easier to take in and evaluate
agreed

yaspo wrote:

I'd like to avoid tying OD to strict star ratings, but for the generic-anime-NM1-equivalent we're probably looking at 7.5*+ or something ? I'm not sure
i also don't believe tying od to sr is the best call but it was the first thing that came to mind lol
if it is gonna be tied to sr, 7.5*+ would make more sense to me as well
note density could be a factor here too, to make it more in line with the ar requirement (any map that's, say, 290bpm+ and utilizes 1/4 rhythms should be applicable or something to that extent)
abraker
please dont tie it to sr. that thing is not an accurate form of difficulty measurement

Also gonna make mappers make diff spikes to boost SR from 6* to ~7.5* just for the sake of being allowed to use AR 11
Topic Starter
yaspo

abraker wrote:

that thing is not an accurate form of difficulty measurement
Pretty much what I want to avoid yeah, it's just a good starting reference to get a feel for it.

One option is to expand on existing Difficulty Specific Guidelines, adding 2 steps (Extreme and ??? idk). Tying OD to note density seems a bit Meh on first glance, I'm not sure what other options are out there.
Topic Starter
yaspo
Alright so, the thread unfortunately goes on hold since "A fresh take is required" in regards to mods. Aka wait for lazer™
Was kinda banking on that we could get a band-aid implementation for mods that doesn't but it appears that won't be happening, unlucky.

See y'all next time
show more
Please sign in to reply.

New reply