Guest
home
beatmaps
rankings
community
store
help
forum
  1. Forums
  2. Beatmaps
  3. Mapping Discussion
  4. Ranking Criteria
  5. Finalized/Denied Amendments

[Proposal - osu] Accounting for sliderend leniency

posted
Total Posts
25
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
Considering stuff like glacierfall is being pushed in qualified, might be a good idea to ensure visual components are still kept intact for cases like 00:33:955 (1) - where the sliderend is at the top right of the slider but the visual is at the bottom right

cuz sliderends (not reverse arrows or sliderheads!!!) have a 36 ms offset

this is also an issue for beatmaps like FALLEN 02:31:998 (1) - where u have to do this in order to 300 the slider where no player will (probably ever) play the slider like this

Maybe controversial since it would limit le choices for slider design but maybe a rule that states to ensure sliderend leniency overlapping with the visual sliderend like
Ensure -36 ms sliderend leniency offsets overlap the visual sliderend displayed for a slider
or a guideline to ensure that the path that players need to take to 300 a slider is obvious or something like
Ensure players following through a slider via visual cues can 300/fully complete a slider
 See community/forums/posts/8237803

Not sure, but I'm mostly writing this to discuss and resolve to a definite answer regarding this, personally I would encourage a rule for leniency overlapping the visual end currently but there may be issues that I am missing with that
Last edited by VINXIS , edited 9 times in total.
Mismagius
9,239 posts
Joined July 2008
Mismagius
agree + if you dont want to limit the possibilities too much, you can have some sort of visual cue to where the 36ms leniency ends (say, a small slider-tumor) so it can at least serve as a point of reference to the player
Belladonna
77 posts
Joined February 2018
Belladonna
.
Last edited by Belladonna , edited 1 time in total.
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS

Mismagius wrote:

agree + if you dont want to limit the possibilities too much, you can have some sort of visual cue to where the 36ms leniency ends (say, a small slider-tumor) so it can at least serve as a point of reference to the player
yea i think that rule would allow those
Nyanaro
185 posts
Joined March 2014
Nyanaro
Disagree, high level players, especially those adept at playing difficult sliders have developed muscle memory and skills to specifically identify the 36ms offset and play sliders accordingly.

- where u have to do this in order to 300 the slider where no player will (probably ever) play the slider like this
This is wrong, players already do this, and have been for years, even those who don't know about the 36ms offset do this instinctively.

Adding a rule limiting slider design when players are already accustomed to said design is not good imo
UndeadCapulet
1,100 posts
Joined April 2013
UndeadCapulet
really disagree with this. i dont like those sliders you posted either but that doesnt matter because people that play the difficulty levels those diffs are for know how to aim the sliders, or at least can quickly learn how to. there isnt a good argument for why the 36ms window should have to overlap the slidertail other than the idea that its important for sightreading, or some vague "it *should* be that way", but lots of basic slider arrangements dont play how they look they would, this is just the same idea but for a single slider.
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
@Nyanaro Well I don't actually know if you're right but I feel more inclined to believe otherwise/think this is a very uncommon way people play considering lower diffs condition players to follow sliders completely and I asked players who apparently either don't know how to identify it so they just go to the end or know but go to the end anyway because there isn't any visual definition to know how to actually 300 the slider and the next best thing is just the end

I'm having trouble understanding how players would even identify the actual end of those 2 slider examples though fwiw what is the process, i dont see how they can unless they guess and check each run or something which I think is way worse than being abel to intuitively just go to the end
Last edited by VINXIS , edited 2 times in total.
Nyanaro
185 posts
Joined March 2014
Nyanaro
@VINXIS The problem is this applies to all sliders, not just fast ones, and lessening incentive to learn the skill will make everyone worse at sliders. -36ms offset is a game mechanic and it should be encouraged for players to learn how to identify it instinctively

edit: I genuinely believe that both the glacierfall and FALLEN sliders are intuitive to play and sightreadable, given the skill of the player to identify the offset.
Last edited by Nyanaro , edited 2 times in total.
moonpoint
311 posts
Joined January 2017
moonpoint
Agree but put it as a guideline so that it's an expectation for the mapper to follow - I imagine kroytz would prefer the look of the arrow over the playability of it for example
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
@Nyanaro
but non-fast sliders dont have the issue where u 100 by being on the sliderend tho, its like a contradiction of cues so
Nyanaro
185 posts
Joined March 2014
Nyanaro
@VINXIS, even so, it is less optimal to go straight to the end and given margins of error a player is more likely to 100 a comparatively slow slider by going straight to the end than by identifying the offset and playing optimally.

edit: Besides, to myself and many other slider players, the cue is to identify the true end based on the length and shape of the slider by calculating the offset in real time. It happens instinctively. The cue doesn't need to be directly visual, it can be calculated
Last edited by Nyanaro , edited 3 times in total.
YokesPai
150 posts
Joined May 2015
YokesPai
as a player who has done a lot of slider cheesing

don't agree with nyanaro, i had to learn about the 36ms offset before i even started learning how to chesse sliders for 300s. even if i can now automatically aim before the sliderend it's still not 100% reliable and i'll look in editor if i mess up too many times.

like at best it's literally educated guesswork cuz ur trying to calculate 36 milliseconds (u can't even count this in ur head cuz it's such a small time) away from the slider based on how fast u know the slider goes, and what the sliderpath actually is.

it's just lenient cuz the sliderball leniency is big.

what does hold true is that a lot of players probably move away from the slider before it's actually done, because of the 36ms check. i do believe changing that would cause a lot of players having to relearn slider leniency or getting more dropped sliderends from releasing a key too early or snapping the cursor away too early
Last edited by YokesPai , edited 1 time in total.
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
Fwiw Nyanaro I need you to prove to me these things like how those sliders are sightreadable and that people generally calculate thigns like these im not taking those at face value just as it is, and im mainly getting responses of the contrary so idk

But either way, im not talking about sightreadability

tell me what the general process of following a slider is, we all know we dont actually follow sliders completely, and for cases like 1/4 sliders and 1/2 sliders that have their ends pointing away from the next object and all that are obviously cases where u dont completely follow sliders

but sliders where their leniency end are not overlapping with their visual end are not in the same vein

for the above examples i wrote in this reply, u arent breaking by reaching the sliderend at any capacity, for ones that are so fast u cant even make a straight slider horizontally u are breaking by being ON the sliderend; by accounting for the -36, u arent removing whatever instinctive calculation for -36 ms that u are talking about to 300, but u are also making it possible to 300 the sliderend by being on the sliderend

Okay I'm going to be honest idk how to explain this that well, the basis of this is mainly regarding that sliderends not overlapping the end is "breaking" moreso than a "sightread" thing, like I literally would be fine with things being added to beatmaps that would not be sightreadable as long as it's not punishing players for whatever reason

This is just discussion anyway but yea
Last edited by VINXIS , edited 5 times in total.
Nyanaro
185 posts
Joined March 2014
Nyanaro
but sliders where their leniency end are not overlapping with their visual end are not in the same vein

this is a good point, there is a distinction to be made there, however I still don't believe that barring these sliders from ranked or forcing visual cues like tumors is healthy because there are players who do understand the 36ms offset and their skills should be tested.

Then again I'm not against having a guideline for this, as visual indicators are good for intuitive map creation, as long as it isn't a rule and these maps can still be allowed to go through
-GN
Champion Above Champions
734 posts
Joined June 2011
-GN

VINXIS wrote:

Fwiw Nyanaro I need you to prove to me these things like how those sliders are sightreadable and that people generally calculate thigns like these im not taking those at face value just as it is, and im mainly getting responses of the contrary so idk
it comes down to a gut feeling based on the speed of the slider essentially, 36ms is a very short time and the main effect of the magical placement of a true sliderend for fast sliders is just to prevent you from snapping directly to the sliderend without punishment. you can definitely train a certain accuracy for this for anything but the dumbest of sliders which aren't rankable to begin with, but what happens when you're wrong is that you have to know the magic number then either memorize the position from the editor or do trial and error until you know, which is simply put not fun

i'd support a visual cue if it's not intrusive; since it doesn't apply to most sliders there's no point in having it be displayed, but i think in any case there's a big movement from the true sliderend to the visual sliderend it should be possible to read where the sliderend is given you know what to look for (i.e. disappearance of the sliderfollowcircle or a visual point)
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
Oh if we are just talking about the gut feeling then yea, talking about the same thing in the 2nd paragraph of that reply then

ALSO WAIT I JUST fuicking realized why dont we just get the the whole area between the -36 sliderend and the visual sliderend itself count as 300 instead of just the -36 ms area thats so fucky howq it is currently not even actual sliderend leniency rn Lol
UndeadCapulet
1,100 posts
Joined April 2013
UndeadCapulet

VINXIS wrote:

ALSO WAIT I JUST fuicking realized why dont we just get the the whole area between the -36 sliderend and the visual sliderend itself count as 300 instead of just the -36 ms area thats so fucky howq it is currently not even actual sliderend leniency rn Lol
yeah this sounds like a much better approach to me, trying to reach out to a dev adjustment if possible before messing w/ ranking criteria stuffs
VINXIS
Featured Artist
3,161 posts
Joined April 2014
Topic Starter
VINXIS
have done https://github.com/ppy/osu-stable-issues/issues/808
yaspo
osu! Alumni
249 posts
Joined September 2014
yaspo
that makes so much sense, osuHOW
Nyanaro
185 posts
Joined March 2014
Nyanaro

VINXIS wrote:

ALSO WAIT I JUST fuicking realized why dont we just get the the whole area between the -36 sliderend and the visual sliderend itself count as 300 instead of just the -36 ms area thats so fucky howq it is currently not even actual sliderend leniency rn Lol
This is an idea and probably how it was originally intended to be when implemented back in ancient times. I support.
show more
Please sign in to reply.

New reply

0
1 / 25