This proposal aims to revert the hybrid set rules to revive the existence of hybrid sets pursuant to generally strong community and internal reception.

The main proposal is the creation of a "main game mode" determination system that applies to hybrid maps. This system works as follows:

- The main game mode of a mapset is defined as the game mode with the most difficulties (for mania, different keymodes are considered, i.e. a 3 diff 4k and 7k spread counts as 6 mania difficulties for purposes of determining the main game mode).
- If the amount of difficulties is the same between modes, the main game mode is instead the game mode which the host of the set mapped the most difficulties of.
- If the previous two rules do not produce a result, the main game mode is any game mode meeting the previous two criteria.
- The map must be nominated by two BNs qualified for the main game mode and one BN per other game mode.
- Probation BNs cannot be the only nomination on their mode.
- Hybrid BNs can place nominations for multiple modes as currently possible.

The spread rules for each mode still apply normally - no skipping difficulties.

Some examples:
I map 5 standard difficulties and get 4 taiko GDs from Hivie. The main game mode is standard and I need two standard BNs and one taiko BN to get the map qualified.

I map 3 standard difficulties and get 3 CtB GDs from Greaper. The main game mode is standard and I need two standard BNs and one CtB BN to get the map qualified.

I map 1 standard difficulty and 1 taiko difficulty. The main game mode is up to my choice, and I need one standard BN, one taiko BN, and an additional one either taiko or standard BN.

Common arguments:
- Hybrid set abuse a la CBCC/Black Rover/Quaver type sets - I believe this is addressed by the requirement to find two main gamemode BNs. Further, vetoes exist.
- Map checking may not be as good with only 1 BN - For MV-able issues, now there's 3 BNs looking at it, so the main concern here is mode-specific issues. See below.
- Map quality may suffer - True, but I believe the upsides of hybrid sets like newer players being introduced to other game modes, cross-mode collaboration and mappers having fun making sets with their friends outweighs this issue. In my experience if someone can find 1 BN for a turbo shitmap they can probably find 2 BNs anyway.

this currently has the full support of the NAT btw

sounds awesome

i agree this is epic

this would be epic :O

I can finally revive my plans for hybrid sets >:3c

i think this is good change, otherwise hybrid is just dead, no reason for mapper to make it harder to rank a map by making it hyrid

what are ur thoughts on not having the "main mode" system at all and just making it 1 bn/mode

(full support for the proposed system btw just curious about this)

yes plz finally

Rules regarding contribution should probably be the same to avoid issues where a std mapper maps 1 difficulty and can only get 1 BN, so instead of waiting for BN 2 they get 2 catch GDs and 2 catch BNs to make it so they can nominate their STD map with just 1 BN

Nijika Ijichi wrote:

what are ur thoughts on not having the "main mode" system at all and just making it 1 bn/mode

this is quite literally (almost) how the system was before it got changed to the current 2bns/mode, it's just way too prone to abuse

yeaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

Very nice!

arcpotato wrote:

Nijika Ijichi wrote:

what are ur thoughts on not having the "main mode" system at all and just making it 1 bn/mode

this is quite literally (almost) how the system was before it got changed to the current 2bns/mode, it's just way too prone to abuse

i feel like having 1bn/mode + 1 extra bn doesn't really make it harder to abuse than 1bn/mode. if someone had an abusive spread in one gamemode they could just load the other gamemode with more diffs so that they only need 1 BN for the abusive spread

either way, such a set would just get vetoed anyways since we're recently been very proactive in vetoing low quality sets

plus, wasn't the issue with the past system that you just needed 2 bns regardless of how many modes there were? 1bn/mode ensures that every mode gets checked which is much better than the previous system

Nijika Ijichi wrote:

arcpotato wrote:

Nijika Ijichi wrote:

what are ur thoughts on not having the "main mode" system at all and just making it 1 bn/mode

this is quite literally (almost) how the system was before it got changed to the current 2bns/mode, it's just way too prone to abuse

i feel like having 1bn/mode + 1 extra bn doesn't really make it harder to abuse than 1bn/mode. if someone had an abusive spread in one gamemode they could just load the other gamemode with more diffs so that they only need 1 BN for the abusive spread

either way, such a set would just get vetoed anyways since we're recently been very proactive in vetoing low quality sets

plus, wasn't the issue with the past system that you just needed 2 bns regardless of how many modes there were? 1bn/mode ensures that every mode gets checked which is much better than the previous system

no cuz like lets say sotarks makes a 15 diff standard spread and gets 3 taiko gds, he would need 13 additional taiko diffs to rank his standard set with 1 BN

yes

what if he actually gets 13 additional taiko diffs (or rather, makes a 4k 5k 6k 7k spread each with 4 diffs)

i think this is a fire idea, because hybrid set rules right now make getting a hybrid set so annoying people just don't anymore and i think this would help a lot with bringing them back

yes yes YESSSSSSSSSSS WE ARE SO BACK

Yes please

Nijika Ijichi wrote:

what if he actually gets 13 additional taiko diffs (or rather, makes a 4k 5k 6k 7k spread each with 4 diffs)

well that's at least going to be immensely more effort than checking 13 std + 3 taiko LMAO

Nijika Ijichi wrote:

plus, wasn't the issue with the past system that you just needed 2 bns regardless of how many modes there were? 1bn/mode ensures that every mode gets checked which is much better than the previous system

nah, pretty sure they informally had a system in place where if a hybrid was more than 2 modes, some bns from the other mode(s) would still check the spread of their mode (and leave some message if it's ready), despite not being listed as an actual nominator. it's not the main issue as compared to what nao described earlier

edit for example: MIIRO with the 4 praises on the discussion lol

arcpotato wrote:

Nijika Ijichi wrote:

what if he actually gets 13 additional taiko diffs (or rather, makes a 4k 5k 6k 7k spread each with 4 diffs)

well that's at least going to be immensely more effort than checking 13 std + 3 taiko LMAO

the point i wanted to make was that there will always be a way to make abusive sets with hybrids, so the extra bn requirement wouldn't really be useful since its primary purpose was to prevent hybrid set abuse

although i do see how the "main mode" system could potentially serve as a deterrent to hybrid set abuse it just feels unnecessary to me

how does this works on 4 modes (osu! on main, osu!taiko, osu!catch, osu!mania)

does this means

2 bn for circle
1 bn for tako
1 bn for frut
1 bn for paino ?

Arzenvald wrote:

how does this works on 4 modes (osu! on main, osu!taiko, osu!catch, osu!mania)

does this means

2 bn for circle
1 bn for tako
1 bn for frut
1 bn for paino ?

yes

Nijika Ijichi wrote:

arcpotato wrote:

Nijika Ijichi wrote:

what if he actually gets 13 additional taiko diffs (or rather, makes a 4k 5k 6k 7k spread each with 4 diffs)

well that's at least going to be immensely more effort than checking 13 std + 3 taiko LMAO

the point i wanted to make was that there will always be a way to make abusive sets with hybrids, so the extra bn requirement wouldn't really be useful since its primary purpose was to prevent hybrid set abuse

although i do see how the "main mode" system could potentially serve as a deterrent to hybrid set abuse it just feels unnecessary to me

it raises the barrier to it a lot, doubt people who are trying to cheat the system are going to go and organize some mega set for another mode instead of getting another bn or making a better map basically

lets goooo

Sounds cool, tho map quality will be deteriorating at some point

I'm all for this change since I love hybrid sets and it helps promote the other gamemodes.

i think we originally went to this because the other modes only needing one bn meant that the quality of hybrids would suffer and that you could "sneak in bad shit"

I'm not sure how much it applies nowadays, but as the proposal also needs development we would need to get someone to actually do that

I think relaxing this is fine, if it causes issues we can discuss the issues in detail again

Yes please

we are so back

oh hell yeah make hybrids great again

Yea I fully agree with this. I feel like currently there's no motivation to make hybrid sets.

Very good proposal to revive hybrid sets, really hope this actually happens given how much i personnally love those

When it comes to hybrids, needing 2 BNs per mode apparently becomes an issue (otherwise, this wouldn't be a proposal). However, single-mode mapsets do get ranked requiring the same two BNs, without that being an inconvenience. It doesn't seem that the number of BNs required per mode to rank a mapset is actually the issue (otherwise, the proposal could be extended to single-mode sets), but the way the system is streamlined.

I do think having to go through something like Clover wish (one nomination reset in a specific mode forcing other modes to be renominated) is undesirable, but in such cases, the proper solution would not be to reduce the amount of BNs required to rank the mapset, but actually improve the system so resets are focused on the relevant modes. A Bancho reset would only reset the nomination for the modes whose .osu files were updated. Manual resets on specific diffs should reset nominations on the relevant mode. As for manual resets on general tabs, I can quickly think of a few ways to deal with them (and maybe someone else could come up with something else): (1) having the popping nominator indicate which modes it will be reset (similar to when a hybrid BN nominates several modes at once), or have several general tabs, one that actually applies for all difficulties (e.g. stuff like metadata, timing) and one general tab per mode (issues that apply to all or more than one difficulty of a single mode) and have the nominations be reset accordingly.

And since map quality has been brought up as a potential issue with moving forward with this proposal - will the NAT keep track of the number/severity of disqualifications before/after the change to actually measure such an impact?

----

Nao Tomori wrote:

- Map quality may suffer - True, but I believe the upsides of hybrid sets like newer players being introduced to other game modes, cross-mode collaboration and mappers having fun making sets with their friends outweighs this issue.

All these upsides you mention are still there, whether you need 1 or 2 BNs for any of the accompanying modes. They are not a benefit of this proposal.

Nao Tomori wrote:

In my experience if someone can find 1 BN for a turbo shitmap they can probably find 2 BNs anyway.

This is basically screaming "Whoever wants to abuse the system will abuse it, so it's irrelevant if it becomes even easier to abuse it"

ZiRoX wrote:

Nao Tomori wrote:

- Map quality may suffer - True, but I believe the upsides of hybrid sets like newer players being introduced to other game modes, cross-mode collaboration and mappers having fun making sets with their friends outweighs this issue.

All these upsides you mention are still there, whether you need 1 or 2 BNs for any of the accompanying modes. They are not a benefit of this proposal.

Nao Tomori wrote:

In my experience if someone can find 1 BN for a turbo shitmap they can probably find 2 BNs anyway.

This is basically screaming "Whoever wants to abuse the system will abuse it, so it's irrelevant if it becomes even easier to abuse it"

1. The upsides are not being exposed right now because people don't rank hybrid sets, the proposal aims to revitalize them so I don't think attributing the upsides to them being revitalized is inaccurate.

2. Sure if that's how you want to interpret it then yeah.

can have a try ig.

But if 1 BN can nominate a single mode in hybrid mapset properly, we exactly don't need 2 BNs for nomination in even single mode mapset as well?

I just wanna sure what the duty of 2nd BN is.
cuz in my view, we need 2 BNs' nomination for keeping positive map quality, and avoid missing dq issues when checking maps.
If u remove the limit, it means 2nd BN duty under any mode is not necessary now, so why not remove the limit for single mode mapset as well? 👀

History lesson on demand: community/forums/topics/194226?n=253

We had such a system in place before which worked well. We abolished it soon after because technical-wise, the new Nomination System only allowed for two Nominations to take place without the possibility to set icons. Icons were back then something like a sanity check for the two actual ranking BATs(=nominating BNs) to know that the other modes got checked by mode specific BATs(=BNs) as well. When an unrank(=DQ) happened though, everyone got kinda yelled at. The latter part we can easily fix nowadays.

Other than that, there were not really huge drawbacks from this, besides the usual culprits that it required some more time. People through the modes where more connected with each other and exchanged knowledge through the modes as well.

It worked back then. So I would appreciate to try this out again. This gets my support.

Ono is correct.

Correct, when moddingv2 was first implemented, it was statically fixed to require 2 BNs to nominate it. Also the system did not differ between the game mode specialization of each individual BN, this was added years later. Therefore, an equal implementation to the status we had prior moddingv2 was not possible back in the time.

The current proposal is factually equal to the sanity-check implementation we had before moddingv2 became a thing, requiring 1 bubble and heart as usual (nowadays: 2 nominations) plus one icon for each non-main-gamemode set by specialized BAT. It worked quite well and we made good experiences with it. We were quite sad that we couldn't carry over this implementation to the new moddingv2 from the get-go. The system gave non-hybrid BATs the confidence that what they are about to rank/nominate is good to go, as someone else specialized mulled over the difficulties they couldn't assess themselves.

The only reasons I heard back then when hybrid system like this can't be implemented was technical issues. I'm all in with this proposal given the support of the system, we need more hybrid mapset!

no pls , make it harder with 5bn/game mode since everything feels same.

I'm all for a hybrid beatmap revival, there really isn't much incentive to map them currently. However, my main concern with this proposal is the quality reduction of hybrid maps, which is even alluded to in the proposal, "Map quality may suffer - True". I don't think things like "having fun" outweights the issue, since they can already happen in the current system. In general, I think that, if two separate maps would require 4 BNs, I don't see why it becomes 3 BNs when they are joint, if the maps are the same.

Initially I had written a bigger text against this proposal, but decided not to post it. It seems that, due to all NATs agreeing + large community support, this proposal be approved. In that case, I ask the NATs to constantly analyze the quality of hybrids from now on to see if their quality (or more specifically, the quality of the modes' diffs that require 1 BN) will decrease. If they do decrease, the proposal should be reverted.

Since this proposal is being probably approved, I'll be positive and hope it is successful with reviving hybrids while not decreasing maps' quality. If it is proven that this system does not decrease maps' quality significantly I'll be all for it.

agree pls make hybrid easier to rank

I strongly disagree with this.

For the purpose of maintaining the quality of Ranking content, it is essential that two BNs check the map. This is the premise.

Loosening the rules because they are cumbersome should not be allowed.

If this is allowed, then a single mode mapset should also be able to be nominated by a single BN.

Overall, I think this proposal is well done, but I do have some concerns...

The issue about quality degradation has been addressed by iceOC, but I agree and think that nominating only one BN will only increase the likelihood of being DQ / vetoed due to quality issues.

If we can resolve this, I agree with nao.
To use a rough analogy, the map to be solo nominated has to be confirmed by the NAT, or the BN / NAT has to hype.
I understand that all of this is not realistic, but I think it is a bit risky to pass the proposal as it is.

I'm having a hard time understanding the motivation -- why are hybrid sets specifically desired to the point of making it easier to nominate them than if they were split into separate sets?

two "upsides of hybrid sets" were mentioned at the end of the post:

- "newer players being introduced to other game modes"
this doesn't even feel like an upside to me, player can play how they want and it's not my business

- "cross-mode collaboration and mappers having fun making sets with their friends"
yeah it's fun but you can already do that, with the same amount of nomination effort you would need from each mode normally, why do we need less?

the issues zirox pointed out seem much more relevant to me than the original proposal.

No clue why nothing has happened in here in a week, seems like this is generally perceived as a positive change by people who understand the proposal correctly. +1 from me

zirox's post got basically a non-response, firika and flashlight's question about why the second bn is suddenly unimportant here is unanswered, mine and iceOC's about differing nomination effort is unanswered, yet it's apparently been forwarded to devs because I see it's in code review now...

the types of suggestions zirox makes would help resolve the systemic tedium that is unique to hybrids without making any questionable compromises so I think it would be better to explore that first

or don't, and just ignore the thread again 😐

clayton wrote:

zirox's post got basically a non-response, firika and flashlight's question about why the second bn is suddenly unimportant here is unanswered, mine and iceOC's about differing nomination effort is unanswered, yet it's apparently been forwarded to devs because I see it's in code review now...

the types of suggestions zirox makes would help resolve the systemic tedium that is unique to hybrids without making any questionable compromises so I think it would be better to explore that first

or don't, and just ignore the thread again 😐

Since this is moving forward I will respond again to the concerns brought up. For what it's worth, I don't think anything here is being newly stated.

1. It's desirable to have hybrids be easier to rank because of the I reasons stated, and because the community wants it. You can see the primarily supportive messaging from most people.

2. Historically, "abusive sets" were always made by sticking 2 or 3 taiko diffs onto a gigantic standard spread. The modification to the system requiring two BNs for the main mode directly addresses that.

3. The job of the 2nd BN is to check the map as well. This is a matter of 15 years of mapping community norms, which for the most part also included the n+1 BN system for hybrid, as Loctav and Ono mentioned. As such, I don't agree that reverting back to that system should also mean we should go to 1 BN for single mode sets.

On quality, my perspective is that checking a map has two parts. The first part is technical items like timing, unsnaps, mp3 quality, etc.; these cause the majority of disqualifications. For this part, now we effectively have 3 BNs checking instead of 2. The second part is actually checking the map, and this is where I acknowledged that an issue can arise. Restating my previous arguments, I don't believe that a subjectively bad map's obstacle to getting ranked is finding the 2nd BN out of two most of the time - it's finding any BNs at all. Additionally, if we assume someone wants to rank a bad map on purpose, it will be less effort to find a 2nd BN than to map a different gamemode full spread to the extent that the other mode becomes the main mode. If the mapper is unaware that their map is bad (usually the case), it is possible that they find their single BN for their GD and the map gets ranked, similar to how bad maps find 2 BNs and get ranked from time to time. I don't think the delta between finding 1 BN and 2 BNs is consistently large enough that some huge contingent of bad maps that can get 1 BN but not 2 will start getting ranked constantly.

this is now implemented!

https://github.com/ppy/osu-web/pull/10984