Instead of having autogenerated star rating, let mappers pick it on their own.

Add an entry in .osu files under [Metadata] StarDifficulty:int
Create a textbox or slider in the GUI for song setup in edit where you can pick a value for the star difficulty property.
So basically, mappers get to pick their own star difficulty display per map difficulty.

Addressing corner cases:
Q. What about old map formats that lack the StarDifficulty metadata property?
A. They will use the old star difficulty

Q. How can a player who is not a map editor tell new star difficulties apart from old ones?
A. Apply a color filter. Make old star difficulties grayed and new ones golden, so that players can tell apart auto-generated and mapper-selected stars.

Q. What if a mapper picks wrong star difficulty?
A. This will be addressed during the modding process. BATs are experienced at this game and will be able to guide mappers on what correct star ratings to pick for their maps (honestly this isn't hard).

Now,

How will this help?

Right now maps pretty much clump together with a ton of 5 stars difficulties, that are all of a completely different level.
For example, if a map has an [Easy][Normal][Hard][Insane][Extra] spread, it often tends to be:
3 stars, 4.5 stars, 5 stars, 5 stars, 5 stars.

With this new system, mappers will be able to manually choose stars and then it will be:
1 star, 2 stars, 3 stars, 4 stars, 5 stars instead

This will make it easier for beginners to navigate across a mapset and find the spread of difficulties. It will also help everybody when a map has very vague difficulty names, i.e. [Snow][Ice][Rain][Storm][Winter]; having mapper-chosen stars will make it clear what was intended to be the easiest through hardest from those.

Also, right now maps are not being sorted correctly. For example an [Insane] might appear before a [Hard] simply because they're both 5 star on the autogenerated star difficulty. This solution would solve this.

So, to sum up this suggestion:

Add to the game: Mapper-selectable star difficulty per map difficulty.

Advantages of new system:
1. Easier for beginners to see the spread of difficulties in a map, instead of a clump of 5star diffs.
2. Easier to tell difficulties apart when difficulty names are vague
3. The sorting of difficulties will always be correct, making it much easier for everyone to navigate on mapsets
4. Since experienced players and mappers (XAT team, etc) will help in the selection of stars, stars will represent difficulty much better than an autogenerated system, which only works sometimes, and often fails to provide a correct value.

Disadvantages of new system:
1. Mappers have to go through an extra step, of picking a star difficulty to each map in the mapset.

This step takes approximately 5 minutes, which is a tiny fraction compared to the process of making a map which is already hours and weeks.
I believe the advantages far outweigh the disadvantages.

So, support?

support o3o *first*

This makes sense but implementing something like Tom's diffs calculator - you know that lvls, would be much better.

Sieg wrote:

This makes sense but implementing something like Tom's diffs calculator - you know that lvls, would be much better.

But do you agree that for now, implementing this would be a step forward?

And it's not a huge feature to implement. I can't imagine this taking more than one day to code into the game.

Aqo wrote:

Sieg wrote:

This makes sense but implementing something like Tom's diffs calculator - you know that lvls, would be much better.

But do you agree that for now, implementing this would be a step forward?

And it's not a huge feature to implement. I can't imagine this taking more than one day to code into the game.

sure
+support

plzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplzplz

i like this idea

and better to implement 10 stars together with this, so people won't find they suddenly can't play 4 star maps, 233

(i mean like, [Easy][Normal][Hard][Insane][Extra] = 3 stars, 4.5 stars, 5 stars, 7 stars [mapper-picked], 8.5 stars[mapper-picked])

t/124644

Moving Aqo's thread into the previous one would help, since he/she got some good points I guess.

Whether this gets merged with Ono's thread or not, I highly support this idea.

[CSGA]Ar3sgice wrote:

and better to implement 10 stars together with this

yeah

I think the current number of stars too low, at least...

OnosakiHito wrote:

http://osu.ppy.sh/forum/t/124644

Moving Aqo's thread into the previous one would help, since he/she got some good points I guess.

Except your thread only specifically discuses star icons. My thread is a more elaborate suggestion regarding stars in general for all maps and not just mode specific ones. It's similar in nature and I support what you're going for, I think it needs to be done in a more encompassing way.

Actually with Ono's Idea it would be already much better. But I think this Idea is fine tho. +Support

I'm amazed you took the time to write a long post like that but didnt took the time to look for duplicates:
t/58826

Sakura wrote:

I'm amazed you took the time to write a long post like that but didnt took the time to look for duplicates:
t/58826

Times changed. Today is not 1 year ago.

Star rating does NOT work
Nobody thinks it works
It's bad

People can pick stars way better than an automated system
It would only benefit everybody -
new players
adept players
mappers

We need this.

I'm amazed you took the time to dig up a 1 year old topic but did not read my OP which clearly explains why we need this today and how it would be in everybody's favor to apply this change.

Then bump that one instead of creating a new thread?

It was denied already and lacked a lot of explanations needed, which Aqo provided here.

Sakura wrote:

Then bump that one instead of creating a new thread?

No.

I believe my OP is much better written and the OP is the main part of a feature request. Reviving an old thread will not progress a request anywhere.

Also, Ekaru's post in the other thread is flawed in every possible way. Mappers should not be restricted in how they can make their maps based on what a flawed automated system thinks about their map difficulty. The fact that freedom dive got 2.5 star rating shows just how bad it is. It is very much possible to make a map that gets 5 stars and yet is very easy to play for beginners simply because it's intuitive and designed well, or a map that gets 3 stars but is insanely challenging and deserves a higher rating. A flawed automated system should not restrict mapping. You have XATs with experience, and mappers that know what they're doing, let them be the judge. Mapping in 2013 has evolved far from previous years maps.

Can you give me an example then of an Easy map that's above 3.5 star rating?

Sakura wrote:

Can you give me an example then of an Easy map that's above 3.5 star rating?

I'm sure I'll be able to find one if I searched enough, but instead of wasting my time on this let me ask you what good is there if an Easy is even 2 stars when it's supposed to be the lowest difficulty in the game?
Why not make it

Easy = 0~0.5 stars
Normal = 1~1.5 stars
Hard = 2~2.5 stars
Insane = 3~3.5 stars
Another = 4~4.5 stars
Extra = 5 stars

and a wider range if possible, like Ar3sgice suggested, for EXX maps like Big Black.
This will look better and will be way more convenient for everybody to navigate on when inside the song selection menu.

I don't see this happening, I'd rather have a broken automated system that can be adjusted & fixed than inaccurate ratings from mappers biased by their perception of difficulty in an ocean of inconsistency.

RBRat3 wrote:

I don't see this happening, I'd rather have a broken automated system that can be adjusted & fixed than inaccurate ratings from mappers biased by their perception of difficulty in an ocean of inconsistency.

Your misconception is that mappers don't know what they're doing.

They do.

And if they need help, there's a whole XAT team for this. It's not hard to set an anchor on a few specific maps and then base difficulty rating from those on all others, having player-selected star rating would be a ton more consistent than the automated system.

Why the hell would you rather have a system that is known to be flawed. It can't be "fixed and adjustment". Having had to change the OD and drain values on Freedom Dive to make it 5 stars instead of 2.5 is stupid as hell and does no go in line with sensible mapping.

Sakura wrote:

Can you give me an example then of an Easy map that's above 3.5 star rating?

This is more meant to the Normal Diffs which would reach the 3.2 Star Limit but are the lowest Diffs in their Mapsets.
Also, an Easy above 3.5 Stars.. sorry but if this happens or happened then the Mapper fucked up with it.

EDIT: Oh god, really? However, newer Maps cannot have more as 3.2 Stars for the lowest Diff. So the argument isn't really valid for the Maps today. More to these Maps in the past.

I'd be worried if said maps existed, but the thing is... as you can see Easy maps never go over 3.5 stars, so beginners know that anything 3.5 stars or lower should be fine for them, the issue only comes when comparing Insanes to Super Insanes

Having bats do their role instead of debating over what star difficulty countless maps should brandish is more productive. Adjusted, Fixed, Completely replaced, do what you will with it as long as it goes by the numbers and not freely tarnished by the hands of a human.

Sakura wrote:

I'd be worried if said maps existed, but the thing is... as you can see Easy maps never go over 3.5 stars, so beginners know that anything 3.5 stars or lower should be fine for them, the issue only comes when comparing Insanes to Super Insanes

So beginners are stuck playing Easy and Normal maps forever because one time in their life they played a "5 star" map that was very insane when a top of 5 star maps are really just easy [Hard]s that they could play with no problem and help them transition into harder maps.

How about ditch the "it's more or less fine" attitude and just accept that the situation needs improving upon?

RBRat3 wrote:

Having bats do their role instead of debating over what star difficulty blah blah

booo

bzzt. BATs already spent more than several hours on checking things in a map before ranking it. The amount of time to decide what star rating every difficulty needs is far below 5 minutes in total. (it would usually take more like 5 seconds)


oh hey check out this map https://osu.ppy.sh/s/76825
[Lan] must be an easy diff right? I mean it's only 3 stars... whoops nope it's harder than big black nevermind.

AUTOMATED
STARS
ARE
BAD

Mappers DO KNOW how to do it better
EVERYBODY knows
Even if player-selected stars will be sometimes even 20% off I'd take that any day over the current automated system.


It seems like the entire "defense line" for keeping automated stars is that they "help beginners" but in reality they do exactly the opposite, they mislead beginners and prevent them from progressing in this game. Having human-selected stars will finally provide beginners with the progressive learning curve that they need and benefit them the most.

The main problem I see is that removing the automation involved with difficulty attribution is the very large variance in opinions regarding difficulty. This pretty much applies to all levels of difficulty as well. The unviased nature of the script, however crappy it is in some situations, at least gives us some sort of springboard with which to work from. Perhaps instead of being able to pick the star rating, being able to order the way the difficulties are presented in the map list would be a compromise.

Also when staying in the confines of the current 10 level system, having each difficulty level restricted to only one of two possibilities in your example wouldnt give an accurate preview imo.

Aqo wrote:

oh hey check out this map https://osu.ppy.sh/s/76825
[Lan] must be an easy diff right? I mean it's only 3 stars... whoops nope it's harder than big black nevermind.

>Unranked. These things are purposed to have low Diffs but being hard to play. You don't really think people who are doing that are taking their Diffs serious or am I wrong?

Garven wrote:

The main problem I see is that removing the automation involved with difficulty attribution is the very large variance in opinions regarding difficulty. This pretty much applies to all levels of difficulty as well. The unviased nature of the script, however crappy it is in some situations, at least gives us some sort of springboard with which to work from. Perhaps instead of being able to pick the star rating, being able to order the way the difficulties are presented in the map list would be a compromise.

Also when staying in the confines of the current 10 level system, having each difficulty level restricted to only one of two possibilities in your example wouldnt give an accurate preview imo.

r:

Aqo wrote:

Q. How can a player who is not a map editor tell new star difficulties apart from old ones?
A. Apply a color filter. Make old star difficulties grayed and new ones golden, so that players can tell apart auto-generated and mapper-selected stars.

I've addressed the point you're bringing up. Those two star systems need to be visibly different. Thus, it will be possible to tell what maps have "old stars" and what maps have "new stars". This solution is fine since there is no reason to not have a different anchor for two different systems when they're clearly distinct.

Deprecated features of the past should not halt the progression of the game forward.

Variance in opinions would not be an issue at all, since ranked maps today already follow a similar standard. It would simply follow the existing standard, and in a way that's better represented on the star display. (i.e. es=1, nm=2, et cetera)

wow Aqo, you completely misunderstood Garven's post.

Havent even read this thread yet (not gonna waste time getting angry at some probably-not-so-good points), just gonna throw my two cents:

The only problem in star rating rely on 3 things:
-Normal difficulties (yes, it's very hard to keep it right, most times when the BPM is over 180)
-HP 7 is glitched (often produces unexpected results like 2.5 star on Freedom Dive or 5 stars on Hard diffs)
-Separating Insane from Extra maps.

Easy diffs are certainly possible to map on star rating with no problem. If you're going 3+ stars without the BPM being 220+, then you're doing something very very very wrong and you shouldn't blame star rating for it.

The reason why such a thing is not a good idea is because we are all people and there is a lot of us.
What will happen is someone sets a star rating, people disagree -> star rating changes, people disagree. What to do then? Do you say to those people that they have no idea what they are talking about? What if those people are staff? Staff members are still human and errors may still happen, and this is just one more variable that would be really hard to set properly, especially when it comes to really hard maps, because some people find jumps harder, some find streams harder, some maps are a bitch to read but others read it fine, etc.
To a lot of people star rating means nothing, so making a huge deal about this isn't something we really need. It is broken, yes, but what it does is that it still takes into account how dense the timeline is, which can still give a relatively accurate description of what may be too hard for an easier diff.

tldr; too many variables to set, which is why it's best to have it automated (but improved I guess). Afaik the system which sets the star rating may get a change in the future too, and this is something I would want to see much more than someone being able to create problems for everyone.

This is far too open for abuse. I only agree with it in one of two cases:

1. Difficulty rating is set by BATs at ranking (similar to Genre and Language)
2. Mappers can choose which difficulty the map is supposed to be and get an icon to match (Onosaki's request pretty much)

I'd much rather revamp the star rating system - there needs to be SOME objective measure of difficulty, even if it's not fully functional. The effort put into this would be better suited for making stars work.

MillhioreF wrote:

1. Difficulty rating is set by BATs at ranking (similar to Genre and Language)

I still don't think this is gonna work because there are a lot of different opinions and you just can't have an absolute power over this if you're in the wrong, which is also really hard to even define because someone will always agree with you and someone will always disagree with you.
My point is there is no objective way of doing this correctly. When it comes to objectivity, automated system is always the best.

Do you prefer something that is objectively bad or something that is subjective but will always be better.

"better" is also subjective in this case

and there is a major disagreement in this thread from most people

we're not here to make what a VERY vocal minority wants, we want to make things better for those who don't complain about every single thing staff does

Aqo wrote:

Do you prefer something that is objectively bad or something that is subjective but will always be better.

Objectively bad, because as I said, it still has it's ups which would be completely gone if it was only up to a person to decide.

What I would want is to be objectively good, which will come with time I guess. How long? No idea.

Solution?

Don't complain about difficulty ratings regardless of what or who sets them.

In a perfect world, maybe.

A BAT assigns a difficulty rating to a map, okay, leave it be because it's probably not off by miles, or if majority still disagrees then exclude said "BAT" from assigning ratings in the future. Regardless of the "human" variable, a difficulty rating by someone knowledgeable will still in nine out of ten cases be better than how it currently is. Exclude anyone outside of the MAT/BAT or whatever it's called. Don't make it more complicated than it is. The goal is more accurate ratings, forget about making it a democracy.

because there's always a vocal minority that disagrees, and people who agree often don't throw out their opinions like rules.

Blue Dragon wrote:

because there's always a vocal minority that disagrees, and people who agree often don't throw out their opinions like rules.

I would never have posted that I agree with this because I have no comment about it. Posts like "support" are beyond pointless in most peoples eyes. You're pretty right with this.

uh, well
star rating is still totally broken for taiko anyways
js
/me summons OnosakiHito

boat wrote:

A BAT assigns a difficulty rating to a map, okay, leave it be because it's probably not off by miles, or if majority still disagrees then exclude said "BAT" from assigning ratings in the future. Regardless of the "human" variable, a difficulty rating by someone knowledgeable will still in nine out of ten cases be better than how it currently is. Exclude anyone outside of the MAT/BAT or whatever it's called. Don't make it more complicated than it is. The goal is more accurate ratings, forget about making it a democracy.

You are still ignoring the fact that even BAT can disagree with each other, which makes this a hassle in the higher ranks as well as in the general community. It's a lose-lose situation.

but BATs can have different opinions about everything regarding a map. That's exactly why there are more BATs, so you can discuss stuff and find the best sollution possible.

I'll ad my personal opinion here and say that I think even with judgement difference from person-to-person, adding this would be beneficial over what we have, but may not be the best way moving forward if the star rating is to be improved in the future.

peppy wrote:

I'll ad my personal opinion here and say that I think even with judgement difference from person-to-person, adding this would be beneficial over what we have, but may not be the best way moving forward if the star rating is to be improved in the future.

the star rating was "to be improved in the future" a year ago already. So yes, if you have a concrete plan what you're going to do, that is good, but if not, this will at least be a step to improve the current system until we can get an even better one.

lolcubes wrote:

You are still ignoring the fact that even BAT can disagree with each other, which makes this a hassle in the higher ranks as well as in the general community. It's a lose-lose situation.

Which is why this shouldn't be a democracy, and not a fix, just an improvement. I'm neither for or against the suggestion, but I find it to be incredibly silly that you can't just leave things be the way they are, ever. It's as simple as setting up a set of guidelines to follow, half of which the automated system already does, but also with the variables which it doesn't take into account. The people assigned to the task will through these guidelines assign a difficulty rating to the map, and despite that some of you may disagree, I'm certain that it would at the very least be already by miles better than what the very much lacking automated system does, which is when you guys should stop nitpicking and leave it be. Something subjective can obviously never be perfect, but as of now it seems as if none of you even want things to get better.

If there would be no democracy, only peppy should be able to set the ratings, but I don't think he would be able to do it correctly on harder maps.

Everyone in the BAT ranks is equal, and this reason is enough for democracy. What you ask is for us to hold a meeting for every single diff when someone says something might be wrong and I find that unreasonable. Just as one BAT may set the diff settings himself, another BAT has the authority to change it, if it's deemed wrong, which requires discussion. How is that not democracy?

Also, if BATs only would be able to set those settings, unranked maps would be left blank mostly. There are so many more unranked than ranked maps, and if you let people set the difficulty rating themselves, you are asking for trouble because a lot of people might on purpose set wrong settings to troll others. Should BATs check all unranked maps too then? Over something that's not gamebreaking (because each diff has a name, which should be enough).

So, for me to not sound only like I am trying to crush all your hopes, let me propose a different system which I think it might work better in that case.

What I propose is that instead of star rating or people assigning difficulty just like, a new system may be created.
I wrote another suggestion about all this and you can view it here: t/124771

This topic seems a lot like this request t/124644

Why don't we just perfect the difficulty rating system first. We should have it automated, not subjectively determinate and then deal with mods telling you differently. It would just be a mess, like this feature request.

Me, lolcubes, MillhioreF, Saten and CXu are currently discussing general agreements and coming up with the best points for how the best difficulty rating would work. Please PM any of us if you want the link, so you can even request joining us with this work. It's mostly just suggestions and discussion for woc2006 to see, he's the one developing this new difficulty calculator. edit: He's taking our suggestions into consideration.

I wantttttttt