forum

osu!(lazer) game mechanics and balance discussion

posted
Total Posts
32
Topic Starter
Ephemeral
On 2023-02-12, a community meeting was held focused around handling a few remaining issues that are hampering user acceptance and uptake of the lazer client.

To keep things focused, the meeting was hosted by a panel of members from various different spheres of the community. As such a panel can be by no means entirely comprehensive, we've also decided to open up a focused discussion thread on the issue to get people's feedback on a few things covered.

Please read the meeting notes linked above for the full details and what to discuss. This can include things mentioned in the previous discussion, or new points related to the matter. There is also a VOD available if you want to listen instead.

When contributing to this thread, please stay focused on the topic of the thread and provide detailed responses. Low-effort or unproductive comments may be removed without warning.
moonpoint
To get the discussion started, I would like to get some opinions on the "First Place Ranks" that appears on player's profiles. From which leaderboard should a first place ranking be determined? Lazer-only leaderboards? Combined (stable+lazer) leaderboards? My gut feeling is that it should absolutely be the combined leaderboards - but that immediately implies that that HAS to be the default leaderboard that players see when they launch the game for the first time ever.

Tangentially, is the team currently expecting to keep score as the primary default metric for sorting scores on a leaderboard? i.e. first place rankings would *never* be determined by pp.
molneya
For the point I feel most strongly about: I don't think it's a good idea that lazer scores should have an inherent advantage over stable scores just because of different mechanics like slider acc, even if they technically make the game harder. I think making all stable scores beatable for the sake of "freshening up the leaderboards" would ruin the history of the vast majority of leaderboards.

Instead, I think that the way to "freshen" up the leaderboards is to give the ability to sort scores by score or pp, and have separate leaderboards for each one. This way, stable scores maintain their history but can still be beaten if someone does a play with better mods or acc (with or without classic mod), and pp leaderboards can remain competitive through things like slider accuracy giving more accuracy pp or rate adjust and difficulty adjust mods. For example: a stable HDDTHRFL 0 spin SS score would be beaten by the same play but done on lazer on the pp leaderboard due to slider accuracy, but would tie on the score leaderboard. This would also avoid having to deal with the nightmare of what mod multiplier should be given for what rate or difficulty adjust setting.

moonpoint wrote:

From which leaderboard should a first place ranking be determined?
With my idea, I think the best option would be to display both pp first places and score first places, if possible. Leaving first places sorted by score would make sense to me from a historical perspective though.
Tactic
I would really love to see all the issues related to spinners in lazer I have mentioned in this twitter thread to be dealt with properly.

Other major worries/issues I have with leaderboard related things with lazer are:
What will happen with maps like Clocks?
While on stable getting maximum score on a spinner on certain maps for over a decade has been almost impossible, isn't it kinda unfair to now be able to max out the spinner with ease?

Peppy thinks recalculating scores from replays is a good idea but the truth is replays can be VERY inconsistent and I would prefer not seeing random scores get ruined by a replay showing something that didn't happen during the actual score
Example 1: Sliderbreak on a PF score
Example 2: blejd, Shaper and probably some other people have their inputs delayed by a random amount on this map
These are only some very obvious flaws, who knows what other weird almost unnoticable inconsistencies happen in random replays cough spinners cough
peppy

molneya wrote:

I think making all stable scores beatable for the sake of "freshening up the leaderboards" would ruin the history of the vast majority of leaderboards.
Just to confirm you understand what you're saying: this would mean that anyone playing without classic mod into the future (ie. the default when playing on lazer) will always be at a disadvantage.

Is the better approach to not "ruin history" to make a way to view leaderboards without lazer scores, as was proposed and unanimously agreed on as a good thing to have?

Tactic wrote:

Peppy thinks recalculating scores from replays is a good idea
I do not think it's a good idea. If you listen to the VOD you will find that I say "yes it is feasible", but also touch on how much extra effort it would be, and as you say it will likely not be perfect.

We've done everything we can to avoid this so far, and hope to not need to resort to it if we can avoid it.

For game mechanics not brought up in the meeting, you can look forward to future discussions. You can find a list of known game mechanic changes which are up for discussion at https://github.com/ppy/osu/wiki/Gameplay-differences-from-osu!stable and (possibly more complete) https://github.com/ppy/osu-wiki/pull/8791/files?short_path=a6f5e72#diff-a6f5e725e2b9c1f9f6e405b7643703559aebe041c3651e8e9e08e48250993604

Note that I will be reading this thread, but not heavily participating. We intentionally held the panel with a very small speaker group because trying to digest opinions from hundreds or thousands of players is counter productive at this point.

But please do continue the discussion :)
molneya

peppy wrote:

Just to confirm you understand what you're saying: this would mean that anyone playing without classic mod into the future (ie. the default when playing on lazer) will always be at a disadvantage.
I think if you are playing without classic mod or with other lazer mods to improve your pp that is a fine sacrifice to make for me. If you care about your score being on the leaderboard you would either switch it to pp or play the map again with classic mod to get a score on the score one.

peppy wrote:

Is the better approach to not "ruin history" to make a way to view leaderboards without lazer scores, as was proposed and unanimously agreed on as a good thing to have?
After stable becomes non-viable in the far future, wouldn't all these leaderboards be completely dead? I think it should still be possible to compete with stable scores if the gameplay mechanics are identical.
Rekunan
Having mod multipliers for rate adjust be bracketed to 0.1x while still allowing 0.01x seems unintuitive, especially for someone new, as players won't see their score/mult increasing despite the difficulty increase, only seeing so when performing 0.1x changes. I'd argue you should also limit accessibility to be the granularity that the mod multipliers are bracketed to.

I'd also argue that 0.1x causes too big of a difference and would much rather prefer 0.05x. This is more clearly shown on higher BPMs, such as 270, where 1.1x increases by 27 to 297 for 1.1x, while 1.05x only increases by 13.5 to 283.5.
peppy

molneya wrote:

After stable becomes non-viable in the far future, wouldn't all these leaderboards be completely dead? I think it should still be possible to compete with stable scores if the gameplay mechanics are identical.
As discussed and agreed upon in the meeting, we don't want to hide stable scores behind a filter. We want them to be visible in the default view. So while you will be able to filter to *only* stable or *only* lazer, I do believe the correct direction is allowing them to show in some combined fashion by default.

Rekunan wrote:

Having mod multipliers for rate adjust be bracketed to 0.1x while still allowing 0.01x seems unintuitive, especially for someone new, as players won't see their score/mult increasing despite the difficulty increase, only seeing so when performing 0.1x changes.
As discussed in the meeting, if we do this, the UI/UX would ned to convey this to the user, giving them the option of raising the setting to the next bracket, or playing with the knowledge that they are in a lower bracket for multiplier.
Sirek
I think it is reasonable that with the leaderboard merge, stable scores should get an applied classic mod, since stable's slider accuracy is inexistent. In the eyes of an average osu! player with barely any leaderboard scores there would be no major problem.

However, as a member of the leaderboard farming community, I would at least want to have my scores recalculated by the replay and converted into the new scoring system, since I am one of those who has spent years grinding for thousands of top 50 leaderboard placements (and I definitely do not want the majority of them to slowly fade away). I think top players would also prefer to have their scores converted, especially players with many, many #1 ranks. On the other hand, there are plenty of players who would not really care about the conversion, like lower ranked players, tournament players, players who play solely for map passes or for example - SS rank farmers, who would surely not want to have their scores getting converted, losing thousands of their SS scores. The osu! community consists of so many different types of players and the scoring effects on them will be dissimilar.

In conclusion, I think that with the lazer update, players should have an option to make a request for converting their stable scores (scores on top50 global leaderboards should be enough). Maybe the converted scores could get a special badge design next to them as well.
beef wellington

molneya wrote:

For the point I feel most strongly about: I don't think it's a good idea that lazer scores should have an inherent advantage over stable scores just because of different mechanics like slider acc, even if they technically make the game harder. I think making all stable scores beatable for the sake of "freshening up the leaderboards" would ruin the history of the vast majority of leaderboards.

Instead, I think that the way to "freshen" up the leaderboards is to give the ability to sort scores by score or pp, and have separate leaderboards for each one. This way, stable scores maintain their history but can still be beaten if someone does a play with better mods or acc (with or without classic mod), and pp leaderboards can remain competitive through things like slider accuracy giving more accuracy pp or rate adjust and difficulty adjust mods. For example: a stable HDDTHRFL 0 spin SS score would be beaten by the same play but done on lazer on the pp leaderboard due to slider accuracy, but would tie on the score leaderboard. This would also avoid having to deal with the nightmare of what mod multiplier should be given for what rate or difficulty adjust setting.

moonpoint wrote:

From which leaderboard should a first place ranking be determined?
With my idea, I think the best option would be to display both pp first places and score first places, if possible. Leaving first places sorted by score would make sense to me from a historical perspective though.
I completely agree with this idea as the history of these leaderboards is something I highly value. I feel having the scoring be the same between stable and lazer, as molneya gave an example for, while adding pp leaderboards that anyone can sort by is a great way to both preserve the history and same preferred competition for a portion of the player base (on the score side), along with having new mechanics such as slider acc/new mods/etc. be competitive on leaderboards as well (on the pp side).

Do note I also agree with molneya on displaying both score and pp first places.
WitherFlower
On the topic of scoring, I would like to share some of the work I've done in collaboration with Zyf and Morgh towards the implementation of a new scoring system.

For anyone interested, you can see the current development branch here

How our new scoring system works

The version of scoring we've implemented is similar to scoreV2, however there are some key differences :
  1. Accuracy no longer matters more at the end than at the beginning of a combo.
  2. An exponent (0.5 -> square root) is appled to combo to reduce the impact of missing on a long combo, and give more importance to smaller combos.
  3. The power on accuracy has also been reduced to match the lower exponent on combo, and to increase the importance of accuracy on lower acc scores.
Some ideas were proposed in the last meeting, though we chose not to use them :

Why not use the top X combos of the map ?

We do not plan to store the top X combos in the database as has been suggested by others in the past, for various reasons (bad for long maps, generates instabilities in the score derivative, ...) but we instead plan to store the computed ComboScore, in order to allow altering its weight (currently 70%) in the final formula if desired.

Why not use misses in the combo part of the score ?

Total misscount gives far less information on combo than what it would seem like, for a simple reason : it is impossible to predict how those misses were distributed throughout the play. If I have 3 misses, I don't know if they were spread evenly, or if the player only missed a triple.
It is also very hard to use misses for score calculation without making them very punitive. While testing this solution, we observed cases where a player could lose upwards of 50,000 score from a single miss, which feels very wrong in terms of player experience.

If you want more details, you can check this (pretty long) github discussion.

We plan to make a Pull Request for that proposal soon, so I won't go into technical details in this post, but I hope this post was helpful to answer some questions / clear some misconceptions about scoring.
Rekunan

peppy wrote:

Rekunan wrote:

Having mod multipliers for rate adjust be bracketed to 0.1x while still allowing 0.01x seems unintuitive, especially for someone new, as players won't see their score/mult increasing despite the difficulty increase, only seeing so when performing 0.1x changes.
As discussed in the meeting, if we do this, the UI/UX would ned to convey this to the user, giving them the option of raising the setting to the next bracket, or playing with the knowledge that they are in a lower bracket for multiplier.
On that note, if this takes place, would something similar occur on the pp side? I can see that being a good thing for score and pp alignment.

Sirek wrote:

However, as a member of the leaderboard farming community, I would at least want to have my scores recalculated by the replay and converted into the new scoring system, since I am one of those who has spent years grinding for thousands of top 50 leaderboard placements (and I definitely do not want the majority of them to slowly fade away).
Something like this was mentioned in the meeting, and peppy basically said it'd be too much work to recalculate everything from the replay, especially considering that they're limited to the top 1k on a map leaderboard, and that stable replays can be jank at times. I think as an alternative we should use percentages/proportions to estimate the 100's count, as it's safe to assume that if you dropped acc on x% circles, the same would apply for sliders.

As an example, on Shige's FDFD play, he achieved 5x100's. Since that map has 1646 circles and 335 sliders, that comes out to about a 5/1 circle to slider ratio, meaning we can estimate to add 1x100, to give a final recalculation of 6x100's.

To provide an equation, it would be:
Estimated acc drops = current acc drops+(current acc drops)/(circles)*(sliders)

This comes with 2 caveats. The first is that SS scores will stay as SS scores (0% acc drops on circles means estimated 0% on sliders as well), which is a good thing if anything. The second is that there could be acc drops from dropped sliderends and sliderbreaks, meaning there'd have to be some way of estimating how many dropped sliderends and sliderbreaks there were, probably through the max combo, and also borrowing the algorithm of estimated misses from the pp code.
Sleepteiner

WitherFlower wrote:

Total misscount gives far less information on combo than what it would seem like, for a simple reason : it is impossible to predict how those misses were distributed throughout the play. If I have 3 misses, I don't know if they were spread evenly, or if the player only missed a triple.
I don't see why this is a problem. Why is completely messing up a whole triple assumed to always be less of a mistake than missing randomly on 3 different jump patterns? The player really has to mess up to miss a whole triple. I don't think we should make arbitrary judgments on different kinds of misses and use that as a basis for a scoring system.

WitherFlower wrote:

It is also very hard to use misses for score calculation without making them very punitive. While testing this solution, we observed cases where a player could lose upwards of 50,000 score from a single miss, which feels very wrong in terms of player experience.
Why does it feel very wrong for misses to be punished significantly in certain cases? The player can lose way more than 50,000 score total from a single miss using combo scoring. Also, what cases?

peppy wrote:

Just to confirm you understand what you're saying: this would mean that anyone playing without classic mod into the future (ie. the default when playing on lazer) will always be at a disadvantage.
Then how about just not forcing significant difficulty changing gameplay changes on already ranked maps while playing nomod? That would resolve this dilemma.

The main changes in this category that I know of are notelock, slider ends not really mattering, and slider accuracy.

The notelock changes usually don't really change how the map is played outside of trying to pass a high star map, so it can just be a setting for player preference. It doesn't have to be part of a mod.

The changes to slider ends making them almost optional are clearly horrible for the flow of slider maps, so I assume this'll be reverted soon.

The main change is of course slider accuracy. I have never understood why slider accuracy needs to be the default.

There are thousands of ranked maps that use slider accuracy leniency as an intentional mapping tool to remove unwanted reading and accuracy difficulty spikes. Some of the times where this would be used are:
  1. The first object of a map
  2. The first object after a break
  3. A sudden timing change
  4. A very complex rhythm
This is especially the case on low star maps. Hundreds of ranked map mappers over the last 15+ years have used a slider instead of a circle in these situations on ranked maps to smooth out the difficulty since players, especially new players, shouldn't always be expected to visually read the timing of the approach circles to time something that is rhythmically unintuitive.

The reality is that the vast majority of people will never use the classic mod and new players won't understand what it is or what it does. So, if slider accuracy is forced on all already ranked maps as the default, then the VAST majority of players will be playing those thousands of maps, that use slider accuracy leniency intentionally, in an objectively worse state. Forever. This is sad to me. This will discourage new players from playing non-new maps. And worse, it will discourage mappers from mapping to complex rhythms and timings even more than they already are since they will no longer have slider accuracy leniency to use as a tool.

People often phrase forcing slider accuracy on old maps as something that has to be done for the progress of the game, but for these reasons, with how it is currently planned, I can't see it as anything other than a step backwards for the game.

My proposal is having slider accuracy be a difficulty setting like AR and OD instead of including it in the classic mod. For new maps, mappers can choose to leave slider accuracy on if it doesn't matter for their map, or they can turn it off to use a mapping tool that has been used for 15+ years. For all already ranked maps, it'll be off by default. Players can instead change it using the difficulty adjust mod. If the map has it on and the player turns it off, there will be a score penalty. If the map has it off and the player turns it on, there won't be a score bonus because it is impossible to balance, but they can still use it to practice and also for competing in different ways.

What I'm trying to say is, changing slider accuracy is changing mapping, so it should be up to the mappers how it applies to their maps by default.

(I apologize if any of this seems too strongly worded. I have thought about and debated these topics for several years now, so I have strong opinions about them. Thanks for reading.)
Walavouchey
The changes to slider ends making them almost optional are clearly horrible for the flow of slider maps, so I assume this'll be reverted soon.
Changing slider ends to only affect accuracy solves the problem of having slider breaks a judgement that breaks gives combo but doesn't give a miss, which is super weird. (Edit: I wrote the wrong thing here.)

In the past, slider ends used to give misses in lazer. It similarly solves the problem, but after a lot of backlash slider ends were made more optional instead.

There are thousands of ranked maps that use slider accuracy leniency as an intentional mapping tool to remove unwanted reading and accuracy difficulty spikes. Some of the times where this would be used are:
  1. The first object of a map
  2. The first object after a break
  3. A sudden timing change
  4. A very complex rhythm
This is especially the case on low star maps. Hundreds of ranked map mappers over the last 15+ years have used a slider instead of a circle in these situations on ranked maps to smooth out the difficulty since players, especially new players, shouldn't always be expected to visually read the timing of the approach circles to time something that is rhythmically unintuitive.
Rhythm sense is a skill. It's difficult to accurately hit complex rhythms, which is why pp has already been going in the direction of awarding more for it.

The points you've mentioned don't apply to skilled players who know how to play complex rhythms and read approach circles. I don't think it's fair to assume that complex rhythms should be made easier to play for the sake of newer players, since you lose out on proper feedback for whether you've played played the rhythm correctly, which discourages players from improving.

This will discourage new players from playing non-new maps.
How? Wouldn't the only realistic difference in new maps be that mappers would more often consider placing a circle instead of a slider since their judgement windows would be the same?

It's not like old maps would suddenly become harder than new maps at some arbitrary point in time. Every map would use the same game mechanics, proper hit timing taken into account.

And worse, it will discourage mappers from mapping to complex rhythms and timings even more than they already are since they will no longer have slider accuracy leniency to use as a tool.
I see it as the opposite, that slider accuracy leniency discourages mappers from requiring complex rhythm skill from players. Instead of encouraging players to notice e.g. a rhythm change, lenient sliders make playing the complex rhythms optional.

I can't see why any mapper would be discouraged from mapping complex timings unless they just don't want the player to play the rhythms properly.
Sleepteiner

Walavouchey wrote:

Changing slider ends to only affect accuracy solves the problem of having slider breaks. Having a judgement that breaks combo but doesn't give a miss is super weird.
Slider ends only affect accuracy in stable too. This is unrelated to slider breaks. Missing a slider end almost gives no penalty in lazer which encourages players in competitive situations to ignore them on hard slider sections. This ruins slider maps. Also, I don't see the point in changing something after 15+ years just because it is "weird."

Rhythm sense is a skill. It's difficult to accurately hit complex rhythms
Forcing something to be more difficult doesn't always mean that it'll be better because of it.

The points you've mentioned don't apply to skilled players who know how to play complex rhythms and read approach circles.
Plenty of skilled players don't know how to read approach circles well. That's why low AR maps and maps with complex timings are so underplayed and hated by the average player. Plus, lots of people play with HD as well.

I don't think it's fair to assume that complex rhythms should be made easier to play for the sake of newer players, since you lose out on proper feedback for whether you've played played the rhythm correctly, which discourages players from improving.
Complex rhythms obviously shouldn't be forced to become easier with gameplay changes in lazer, but if a mapper intentionally made it easier so more people can enjoy their map, then why not? Why should we retroactively decide that how they made their map was wrong and should be forced to change? Not everything has to be about competition. The whole design behind most low star maps is to be much easier than it could be so new players can have a relaxing time learning the fundamentals of the game. Why should we get in the way of that?

How? Wouldn't the only realistic difference in new maps be that mappers would more often consider placing a circle instead of a slider since their judgement windows would be the same?
I didn't bring up all of the situations where someone might use slider accuracy leniency intentionally, so it is more than just that, but either way, judging by the comments on lots of maps, new players hate when they don't understand why they can't get good accuracy on a certain part of a map. Having a sudden and unwanted accuracy difficulty spike randomly due to a timing change for example will only push new players away from that map. When they realize that new maps will be easier on average due to that, they will avoid old maps in general, at least to a certain extent.

I can't see why any mapper would be discouraged from mapping complex timings unless they just don't want the player to play the rhythms properly.
osu players tend to avoid maps with complex rhythms. So, yes, tons and tons of mappers avoid complex rhythms because they want their maps to be played. This is just how things are in this game, unfortunately. If this slider accuracy change happens, then mappers and players won't suddenly just change their minds about complex rhythms, many mappers will just avoid them completely. They won't map to difficult rhythms at all. At least in stable, many mappers still map to those rhythms, even if some of them only do so using slider accuracy leniency.

Sure, this wouldn't be a catastrophic change for the game, but it would be a significant and noticeable downgrade for those who play the kinds of maps that are affected and also for the diversity of mapping styles.

Also, I should mention that I usually don't like when mappers use sliders to make a complex rhythm super easy. I like playing complex rhythms. But, I also believe in the importance of artistic preservation. So, even when I think it is a shame that a map is made rhythmically way too easy, as long as it still roughly follows the song, I am not against it being ranked because that is what the mapper intended and that is what average players enjoy. I would never agree to force thousands of ranked maps to change just because I personally disagree with some of the mapping ideas.
Walavouchey
Slider ends only affect accuracy in stable too. This is unrelated to slider breaks.
Ah, I confused myself there. In stable, slider ends affect accuracy and combo, while in lazer they only affect accuracy (they don't give combo). (+ in both versions they affect score, but that's not relevant.)

Because in stable they don't give miss judgements even if you miss them, in lazer they were initially changed so that slider ends would give miss judgements when missed, but then they were changed again (per backlash) such that they don't give misses but also not give combo.

Plenty of skilled players don't know how to read approach circles well. That's why low AR maps and maps with complex timings are so underplayed and hated by the average player. Plus, lots of people play with HD as well.
Maybe it's time to learn? With HD, you read the fade-out timing.

While low AR is pretty niche, it has nothing to do with rhythm. Having complex rhythms be less challenging is weird for a rhythm game.

Complex rhythms obviously shouldn't be forced to become easier with gameplay changes in lazer, but if a mapper intentionally made it easier so more people can enjoy their map, then why not? Why should we retroactively decide that how they made their map was wrong and should be forced to change? Not everything has to be about competition. The whole design behind most low star maps is to be much easier than it could be so new players can have a relaxing time learning the fundamentals of the game. Why should we get in the way of that?
Slider leniency is fundamentally an unintentional game mechanic.

OD should be lowered in order to make judgements easier, not have one specific hit object be easier for no reason. Taiko and mania don't have this kind of thing either, and why would they?

Having a sudden and unwanted accuracy difficulty spike randomly due to a timing change for example will only push new players away from that map. When they realize that new maps will be easier on average due to that, they will avoid old maps in general, at least to a certain extent.
I guess people don't want to play challenging rhythms? Yeah, farm maps and the like are popular and have trivial rhythms for a reason, but why should simpler rhythms be the norm in a rhythm game anyway?

tons and tons of mappers avoid complex rhythms because they want their maps to be played. This is just how things are in this game, unfortunately. If this slider accuracy change happens, then mappers and players won't suddenly just change their minds about complex rhythms, many mappers will just avoid them completely.
Well, what's the root cause here? Maybe that pp hasn't (and still doesn't to some extent) properly award points for complex rhythms. What do you think?

I didn't bring up all of the situations where someone might use slider accuracy leniency intentionally, so it is more than just that [...]

Sure, this wouldn't be a catastrophic change for the game, but it would be a significant and noticeable downgrade for those who play the kinds of maps that are affected and also for the diversity of mapping styles.

I also believe in the importance of artistic preservation. So, even when I think it is a shame that a map is made rhythmically way too easy, as long as it still roughly follows the song, I am not against it being ranked because that is what the mapper intended and that is what average players enjoy. I would never agree to force thousands of ranked maps to change just because I personally disagree with some of the mapping ideas.
This is the kind of discussion that needs to be had, and was touched on in the discussion panel, but probably (definitely) not enough.

Will new players coming into the game via lazer toss their keyboard into the wall because they couldn't hit a slider properly? No, I highly doubt that.

Will experienced players adapt to the new game mechanics and accept that they'll need to hit sliders accurately? Maybe, but even if they don't they can still play with the Classic mod (old game mechanics).

Will mappers be frustrated that their maps will have the new game mechanics imposed on them? I have no idea.
Sleepteiner

Walavouchey wrote:

Because in stable they don't give miss judgements even if you miss them, in lazer they were initially changed so that slider ends would give miss judgements when missed, but then they were changed again (per backlash) such that they don't give misses but also not give combo.
I will continue assuming that the state of slider ends in lazer is temporary because having them not at least give 100s when missing them is disastrous for the game. Much more so than even slider accuracy.

Maybe it's time to learn? With HD, you read the fade-out timing.
Most players won't use the opportunity to learn, they will simply play fewer kinds of maps to avoid learning. Also, the number of players who can use the fade-out timing to acc complex rhythms and timings on lower AR maps with consistency is extremely low. That is a super rare skill.

Slider leniency is fundamentally an unintentional game mechanic.
In my opinion, the original intention of a game mechanic should be valued less than the long term practical utilization of said game mechanic. Should sliders have required full accuracy from the start? Probably. But, that doesn't change that players figured out how to use that design choice to make their maps more enjoyable for their target audience.

OD should be lowered in order to make judgements easier, not have one specific hit object be easier for no reason.
But, in the cases where there is a reason to make a specific hit object easier, I don't think the rest of the map should be made easier by lowering the OD just to make that one hit object manageable.

Taiko and mania don't have this kind of thing either, and why would they?
It is easier to acc taiko and mania visually than it is in standard, in most cases at least, so I don't think it is exactly the same. But, yeah, I'm not saying that how slider accuracy leniency works specifically is a mechanic that should be added to other game modes, don't worry.

Well, what's the root cause here? Maybe that pp hasn't (and still doesn't to some extent) properly award points for complex rhythms. What do you think?
In my opinion, the primary root cause for this is combo. The way combo has always been implemented in every score system and every pp system makes accuracy largely meaningless. From this, new players are taught to mostly ignore accuracy. When a map is hard to acc, for many players, it is seen as an annoyance rather than a challenge and they often blame the map and the mapper for their own lack of accuracy and reading ability. This can be seen from players of every rank range. This is ingrained into the fabric of this game mostly due to reliance on combo.

Will new players coming into the game via lazer toss their keyboard into the wall because they couldn't hit a slider properly? No, I highly doubt that.
I agree. It will be no more than an annoyance for the players that I'm talking about (with the exception of the more rare kinds of maps that would become dramatically less playable with slider accuracy, of course). But, if old maps become more annoying to the average player, then they'll be played less and less. I'd like to avoid artificially contributing to the gradual obsolescence of older maps.
Rekunan
I want to clarify that dropped sliderends were just moved to a different metric that you can see at the score details screen, and that there are plans to add a visual indicator for this: https://github.com/ppy/osu/issues/8700 (similar thing for sliderbreaks). I believe this serves as a better alternative than having them give 100s or 50s because we can know now that any 100s or 50s are due to hit error and not due to only hitting part of a slider.
Sleepteiner
I do not see why 100s and 50s need to only represent hit error considering that that is not how it has been for 15+ years. Creating a whole new metric for missing slider ends both inherently complicates and devalues slider ends. If something similar to what lazer has right now becomes the standard for the game, then competition on hard slider maps will be ruined since the current slider end system in lazer encourages players to play sliders incorrectly.
moonpoint
How does it encourage players to play them incorrectly? Dropped slider ends affect accuracy in Lazer, and consequently score. Is it the lack of a visual indication that a sliderend has been dropped? That's something I could agree on it needing.
show more
Please sign in to reply.

New reply