Ranking Criteria wrote:

No guest mapper should have more difficulties in the mapset than the creator, all modes together. For collaboration mapsets, the uploader should take part in all collab difficulties. The person who should upload the map is the one who contributed most to it.

This is confusing, as it does not cover mapsets where a few difficulties are a collab.
As the previous rule stated (which can be found in this thread), "you should fully map at least one difficulty of a mapset if you are the person who is uploading it", which means that apart from collaborative mapsets where all diffs are a collab (and uploader should take part in all difficulties), it is not allowed to have mapsets like Natteke's Let It Go, but the wording as it is now does not state whether or not it actually is allowed. We need to add something in the guideline to clear up these cases.

So the question is: how do we treat collabs from a accountability point of view ?

Any post discussing collab mapsets (where every difficulty in the mapset is a collab) will be removed. This is already covered by the guideline and is clearly allowed, as long as the uploader has a part in every difficulty.
Any post questioning the concept of accountability will be removed as well. It isn't contributing to the subject.

I know this is a hot subject, so try to keep the discussion civil. No need to attack me for bringing it up or each other for your opinions.


-------------------- personal opinion --------------------

I personally both like and dislike that version and would like to see it enforced. I don't see why someone who hasn't even mapped an entire difficulty should upload the mapset, unless it's the theme, like collab mapsets. On the other hand, it reduces the field for collab mapsets, making things like two mappers per diff on a gigantic collab mapset impossible, and encourages to refuse guest difficulties for different gamemodes if the uploader cannot map said gamemode at all.

One solution would be to consider that as long as the uploader participates in a collab (only the uploader, not guest mappers), that collab counts as a whole difficulty, but I dislike this because it is not a whole difficulty.
The other solution I can think of would be to consider non-integer accountability, taking the number of people participating in a collab into account. For example, a three-mapper collab would account for 33% of a difficulty for each mapper. With that logic, collab mapsets would be more lenient and allow for guest difficulties for different gamemodes, while still retaining the original idea behind the guideline, but it can make things awfully complicated, and there could be stupid borderlines cases (like a guest mapper/collaber having mapped 41% of the mapset and the uploader 40%, I mean what).

I'm waiting for your ideas on this matter. The goal is to explicitely state how collabs are handled (specifically on mapsets that aren't entirely collaborative) in the simplest possible way.

I like the integer idea

I don't think a collab itself alone should count for an entire mapper's accountability

honestly I think the original mapper should at least have 50% of the difficulties in a mapset but that's not the topic ;D

I wouldn't strech the guideline too much. Collabs don't count as "uploader's diff" but they add to the overall contribution of the uploader. I still think the current wording does not allow much of a strech:


X Uploader contributed in collab diff, all other diffs were made by different guests
O Uploader contributed in collab diff and made one other diff, all other diffs by different guests
O Uploader made one diff, all other diffs by different guests
X Uploader made one diff, one guest made either two diffs OR one diff and contributed in a collab without the uploader

By applying the current guideline all the borderline cases can be solved easily. The only time I would personally allow a breach or strech of this guideline are giant mapsets that cover all game modes (and only if it is very close):

O Uploader contributed in (standard) collab diff and made 2 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs
X Uploader contributed in (standard) collab diff and made 1 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs



TL;DR: I don't think we need a re-wording unless you want to specifically allow special cases.

p3n wrote:

I wouldn't strech the guideline too much. Collabs don't count as "uploader's diff" but they add to the overall contribution of the uploader. I still think the current wording does not allow much of a strech:

No guest mapper should have more difficulties in the mapset than the creator, all modes together.

X Uploader contributed in collab diff, all other diffs were made by different guests
O Uploader contributed in collab diff and made one other diff, all other diffs by different guests
O Uploader made one diff, all other diffs by different guests
X Uploader made one diff, one guest made either two diffs OR one diff and contributed in a collab without the uploader

By applying the current guideline all the borderline cases can be solved easily. The only time I would personally allow a breach or strech of this guideline are giant mapsets that cover all game modes (and only if it is very close):

O Uploader contributed in (standard) collab diff and made 2 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs
X Uploader contributed in (standard) collab diff and made 1 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs



TL;DR: I don't think we need a re-wording unless you want to specifically allow special cases.

These edge cases make sense, but then Let It Go can't be ranked. Does Shiro have any cases that may be ambiguous?

english is so hard

What about saying "half a difficulty" for any collab diff, regardless of the number of people contributing to it? It would make the whole 41%/40% problem vanish by doing this. Of course this means the below would be 'acceptable'

O 2 Collab diffs made by 3 people + one guest diff that doesn't contribute to it. see below for easiest numbers.
Diff 1: A,B,C | Diff 2: A B C | Diff 3: D (A being map creator)
Would either be:
0.5 + 0.5 = 1 with the above proposed rule.
0.33 + 0.33 = 0.66 with taking the amount of contributors in each collab to account.

Just throwing out more ideas. I don't really 100% support this, but with more options to choose from, it might help in some way.

xxdeathx wrote:

p3n wrote:

I wouldn't strech the guideline too much. Collabs don't count as "uploader's diff" but they add to the overall contribution of the uploader. I still think the current wording does not allow much of a strech:

No guest mapper should have more difficulties in the mapset than the creator, all modes together.

X Uploader contributed in collab diff, all other diffs were made by different guests
O Uploader contributed in collab diff and made one other diff, all other diffs by different guests
O Uploader made one diff, all other diffs by different guests
X Uploader made one diff, one guest made either two diffs OR one diff and contributed in a collab without the uploader

By applying the current guideline all the borderline cases can be solved easily. The only time I would personally allow a breach or strech of this guideline are giant mapsets that cover all game modes (and only if it is very close):

O Uploader contributed in (standard) collab diff and made 2 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs
X Uploader contributed in (standard) collab diff and made 1 own (standard) diffs, one guest did 3 Taiko diffs, another guest did 3 Mania diffs



TL;DR: I don't think we need a re-wording unless you want to specifically allow special cases.

These edge cases make sense, but then Let It Go can't be ranked. Does Shiro have any cases that may be ambiguous?

The ambiguous cases are the ones like Let It Go and 2-way collabs with different mappers for each diff (uploader only participating in one), mostly.

Shiro wrote:

The ambiguous cases are the ones like Let It Go and 2-way collabs with different mappers for each diff (uploader only participating in one), mostly.

Yeah but that is why it is a guideline. It doesn't stop the ranking BAT if it is a close call with multi-way collabs. There are more clear cut cases where we can apply the guideline as is because blatantly overused guest difficulties should be easy enough to spot.

If you want to tie this guideline to a measurable numeric value, I think it would be best to move it to the RULES instead.

p3n wrote:

If you want to tie this guideline to a measurable numeric value, I think it would be best to move it to the RULES instead.

My grain of sand: I think this should have been a rule from the start.

As for a numerical value:
Each diff will be considered as accounting the same value on the spread.
That is, if a map has a E,N,H,I spread...
E- 25%
N- 25%
H- 25%
I- 25%
Or also, if a map has a N,N,H spread...
N- 33,3%
N- 33,3%
H- 33,3%

And so on...

For collab diffs, each mapper accounts for the same value on the diff.
That is, if a diff has 3 mappers, each mapper accounts for 33,3% of the diff, and so...
(N,H,Collab)
N- 33,3%
H- 33,3%
Collab- Mapper 'a' 11,1%, mapper 'b' 11,1%, mapper 'c' 11,1%.
If a diff has 4 mappers...
(E,N,Collab,I)
E - 25%
N - 25%
Collab- Mapper 'a' 5,25%, Mapper 'b' 5,25%, Mapper 'c' 5,25%, Mapper 'd' 5,25%.

And so on...

That rule can be ambiguous if one of the mappers on the collab has clearly mapped more than the others, for that, I propose that if a mapper on the collab diff has mapped more than 50% of the collab, his parts account for 50% of the diff, while the rest will be divided evenly for the other mappers.

...or something like that. I dunno, I'm just brainstorming ideas. DESUUUU >/////<!

*moes out*!

Chloe wrote:

english is so hard

中文很难

p3n wrote:

Shiro wrote:

The ambiguous cases are the ones like Let It Go and 2-way collabs with different mappers for each diff (uploader only participating in one), mostly.

Yeah but that is why it is a guideline. It doesn't stop the ranking BAT if it is a close call with multi-way collabs. There are more clear cut cases where we can apply the guideline as is because blatantly overused guest difficulties should be easy enough to spot.

If you want to tie this guideline to a measurable numeric value, I think it would be best to move it to the RULES instead.

It was a rule in the first place, then got moved to guidelines after community input for my first rewording thread (even though I was strongly against it). The problem is that the previous rule did not allow cases like Let It Go, while the new guideline seems to.
Giving a numerical solution like Krisom said and moving this to rules seems to be the best course of action to avoid ambiguous cases.

Anyone has any objections ?

A number system representing difficulties would be easier to follow and simple, no confusion whatsoever. Every diff is 1 diff. Someone who maps a diff alone has mapped 1 diff. A collab diff would be split evenly among the mappers who participated; if 2 people collab on a diff, each of them has mapped 0.5 diffs. Add up each person's total diffs mapped and have a rule/guideline regarding them, for example the existing one "No guest mapper should have more difficulties in the mapset than the creator."

For Let It Go, the uploader has mapped 0.5 diff but two other people mapped 1 diff each, so it does not pass. No matter how you bend the rules/guidelines, Let It Go simply can't meet them if their basic idea is that the uploader has to do more mapping. Now I love that song just as much as the next person, but the uploader obviously didn't map as much as others, so it shouldn't be able to get ranked without an exception

xxdeathx wrote:

A number system representing difficulties would be easier to follow and simple, no confusion whatsoever. Every diff is 1 diff. Someone who maps a diff alone has mapped 1 diff. A collab diff would be split evenly among the mappers who participated; if 2 people collab on a diff, each of them has mapped 0.5 diffs. Add up each person's total diffs mapped and have a rule/guideline regarding them, for example the existing one "No guest mapper should have more difficulties in the mapset than the creator."

For Let It Go, the uploader has mapped 0.5 diff but two other people mapped 1 diff each, so it does not pass. No matter how you bend the rules/guidelines, Let It Go simply can't meet them if their basic idea is that the uploader has to do more mapping. Now I love that song just as much as the next person, but the uploader obviously didn't map as much as others, so it shouldn't be able to get ranked without an exception

Pretty much what I said, but with a different scale instead

So!...

For any mapper, each diff accounts for '1', and collab diff accounts for 1/'n' where 'n' is the number of mappers in the collab. The mapper that cotributed the most to the overall spread is the one who should upload the map. If two or more mappers account for the same value, any of them can upload the map.

Like that? I think that's pretty solid.

Now I regret getting my 10th map ranked. >.<
Also, 3000th post!

I can see that you lot are quick to gang up on my map. Not only has it been unranked unfairly before this is even a rule, which defeats the point of having guidelines. Anyway, I've spent more time with the mapset than anyone in it, that includes getting mods which in itself is slave's work, getting people to fix what ever they need to fix, get people to test the map, figure out what the fuck is wrong with the timing and on it goes.

There's more to it than just mapping, getting a map ranked is hard, I'm sure you will agree, due to my limited time I can't be mapping much any more. And who cares who contributed how much? It's down to the participants to decide, and I can assure you that everyone in my mapset is happy with how it turned out. Not even BATs give a fuck, do you know how I know? Here:

2014-08-09 22:02 dkun: and what if I gave you an easy
2014-08-09 22:02 Natteke: and cover it as my own?
2014-08-09 22:02 Natteke: can't do that
2014-08-09 22:02 dkun: in your name, easy
2014-08-09 22:02 dkun: or rather, submit as per-anon and it can't count as anyone else's except yours

See what I'm talking about? But no, due to the herd mentality that's present on osu everyone just has to fucking screw things up.

And if you're still interested in my opinion on the topic: I object, but you can do what ever the fuck you want to do because I simply don't give a shit anymore.

Natteke wrote:

2014-08-09 22:02 dkun: and what if I gave you an easy
2014-08-09 22:02 Natteke: and cover it as my own?
2014-08-09 22:02 Natteke: can't do that
2014-08-09 22:02 dkun: in your name, easy
2014-08-09 22:02 dkun: or rather, submit as per-anon and it can't count as anyone else's except yours

Funny, dkun enforced that rule and unrank map and proposed to cheat. Shows the full value of that rule.

-------------------- personal opinion --------------------

Stupid rules like this shouldn't exist. If mappers who contributed to mapset are agreed on uploader person why you even should be care?

Krisom wrote:

xxdeathx wrote:

A number system representing difficulties would be easier to follow and simple, no confusion whatsoever. Every diff is 1 diff. Someone who maps a diff alone has mapped 1 diff. A collab diff would be split evenly among the mappers who participated; if 2 people collab on a diff, each of them has mapped 0.5 diffs. Add up each person's total diffs mapped and have a rule/guideline regarding them, for example the existing one "No guest mapper should have more difficulties in the mapset than the creator."

For Let It Go, the uploader has mapped 0.5 diff but two other people mapped 1 diff each, so it does not pass. No matter how you bend the rules/guidelines, Let It Go simply can't meet them if their basic idea is that the uploader has to do more mapping. Now I love that song just as much as the next person, but the uploader obviously didn't map as much as others, so it shouldn't be able to get ranked without an exception

Pretty much what I said, but with a different scale instead

So!...

For any mapper, each diff accounts for '1', and collab diff accounts for 1/'n' where 'n' is the number of mappers in the collab. The mapper that cotributed the most to the overall spread is the one who should upload the map. If two or more mappers account for the same value, any of them can upload the map.

Like that? I think that's pretty solid.

lol i can't math to figure out what you were going for, but i guessed it would only make sense if it was something like this

Sieg wrote:

Natteke wrote:

2014-08-09 22:02 dkun: and what if I gave you an easy
2014-08-09 22:02 Natteke: and cover it as my own?
2014-08-09 22:02 Natteke: can't do that
2014-08-09 22:02 dkun: in your name, easy
2014-08-09 22:02 dkun: or rather, submit as per-anon and it can't count as anyone else's except yours

Funny, dkun enforced that rule and unrank map and proposed to cheat. Shows the full value of that rule.

-------------------- personal opinion --------------------

Stupid rules like this shouldn't exist. If mappers who contributed to mapset are agreed on uploader person why you even should be care?

I was only talking about the rankability of Let It Go within the context of the existing rule, whose fundamental idea is that no guest mapper can have mapped more than the uploader. I haven't offered my own opinion on the rule itself so here it is:

@Natteke As a mapper who's never made GD's for others yet but got two mapsets of my own ranked, I know how much work the uploader does to get mods/BAT's/stuff. I told Shiro this when they first linked this forum thread in #modhelp; really I couldn't care less about how much the uploader has mapped, since they're responsible for everything else in the ranking process.

But it looks like the discussion at hand is about how people are gonna enforce the current rule as it is, and not about abolishing it or rewriting it completely. So unfortunately, as I said earlier, Let It Go can't be ranked as long as this rule stands in its current form.

Natteke wrote:

I can see that you lot are quick to gang up on my map. Not only has it been unranked unfairly before this is even a rule, which defeats the point of having guidelines. Anyway, I've spent more time with the mapset than anyone in it, that includes getting mods which in itself is slave's work, getting people to fix what ever they need to fix, get people to test the map, figure out what the fuck is wrong with the timing and on it goes.

There's more to it than just mapping, getting a map ranked is hard, I'm sure you will agree, due to my limited time I can't be mapping much any more. And who cares who contributed how much? It's down to the participants to decide, and I can assure you that everyone in my mapset is happy with how it turned out. Not even BATs give a fuck, do you know how I know? Here:

2014-08-09 22:02 dkun: and what if I gave you an easy
2014-08-09 22:02 Natteke: and cover it as my own?
2014-08-09 22:02 Natteke: can't do that
2014-08-09 22:02 dkun: in your name, easy
2014-08-09 22:02 dkun: or rather, submit as per-anon and it can't count as anyone else's except yours

See what I'm talking about? But no, due to the herd mentality that's present on osu everyone just has to fucking screw things up.

And if you're still interested in my opinion on the topic: I object, but you can do what ever the fuck you want to do because I simply don't give a shit anymore.

You got it backwards. I noticed this hole in the rule when I was checking your mapset. I wanted to clarify this to avoid a possible unrank - and I couldn't. I asked around and it seems that no one really knows how to handle this kind of case. As I said, this is a clarification, not me forcing my views on anything (if I were forcing my views, this would be a hard rule and the uploader would have to map strictly more than half of their mapset). Granted, I could have ninja edited it to fit what I thought about it, but I wanted to get the community's opinion on it to know what to do with it.

Mapset accountability should be dealt with a case-by-case basis in terms of Let It Go. We should keep the original wording but add a clause as a sub-bullet somewhere along the lines of...

1. In addition to this rule, maps with an ambiguous set of difficulties will be evaluated on a case-by-case basis to ensure that the submitting mapper is properly accounted for.

Collabs are and always will be ambiguous. The mapper can choose to map as much or as little as they want, and shouldn't ever count as a static value. The only efficient way is to keep this thread in mind for the future (even link it on the rule like some already do) and work from there.

tl;dr this does agree with p3n's suggestion, but adds a secondary clause to ensure nothing goes wrong on our end of being the ones that rank these maps.

Basically the thread that had the rule moved to guidelines didn't really come to a sound conclusion, since multiple BAT involvement in the thread died out by the end of it.
For the year leading up to today, I don't think BATs even knew about this thread, as zero sets were breaking this guideline.
Overall it was a mistake to have the rule put into guidelines, and now that the first set to use it has finally shown up (a YEAR later I might add), we should put the rule back where it belongs.
No need to do anything complicated.

Moving this back to rules and appending dkun's sentence.