Hello Everyone!

We from the mania NATs already discuss regarding making 12-18K rankable, and pretty much after long discussion, we decided to finally open the possibility of making 12-18K rankable!

To give a big emphasis on this, Peppy already allow us to rank 12-18K, so basically all the process left to do is to get community support, and also testing on rank 12-18K map to make sure the PP works on those keymodes.

1. Change the rule from this:
   
   Beatmaps can only use 4 to 10 keys. Anything else is not supported for the Ranked section.
2. Into this new rule:
   
   Beatmaps can only use 4–10, 12, 14, 16, or 18 keys. Anything else is not supported for the Ranked section.

-----------------

Please note that 11K, 13K, 15K and 17K still won't be rankable due to the game not supporting it. So it's only the even number of keymodes as of now.

So yeah, we will attempt to formalize this very soon! So we do hope the support for making these keymodes rankable!

+1

+2

yay!

plus

I'm in support of this, but are you able to show some of the proposed layouts for 12K-18K for those who are completely brand new to it but would like to try it out? And possibly links to a 12K, 14K, 16K, and 18K map for people to explore and familiarize themselves with. I think it'd help with accessibility and getting more community members on board if they have an idea of how to play it and some maps to start with.

awesome!

Protastic101 wrote:

I'm in support of this, but are you able to show some of the proposed layouts for 12K-18K for those who are completely brand new to it but would like to try it out? And possibly links to a 12K, 14K, 16K, and 18K map for people to explore and familiarize themselves with. I think it'd help with accessibility and getting more community members on board if they have an idea of how to play it and some maps to start with.

The idea is that we have a test run in every keymode where only those can be ranked and we will gather feedback from those to then make an RC for these that will allow them all to be ranked

dont let pp stop players from getting new maps

The only thing I can remember for double play layout are these:

1. EZ2 DP w/o scratches and pedals [14K/5K+4+5K] (5+4+5)
2. EZ2 DP w/ scratches but w/o pedals [16K/1+5K+4+5K+1] (1+5+4+5+1)
   
3. beatmania IIDX w/o scratches [14K] (7+7)
   
4. beatmania IIDX w/ scratches [16K/14K+2] (1+7+7+1)
   

One of the obvious hard rule I can think for both layouts are simply to disallow 6+ chord in one hand.

Protastic101 wrote:

I'm in support of this, but are you able to show some of the proposed layouts for 12K-18K for those who are completely brand new to it but would like to try it out? And possibly links to a 12K, 14K, 16K, and 18K map for people to explore and familiarize themselves with. I think it'd help with accessibility and getting more community members on board if they have an idea of how to play it and some maps to start with.

Not a comprehensive list by any means, but here are a few mapsets that I think are worth mentioning.

10K2S
beatmapsets/1976303
beatmapsets/1818921
beatmapsets/1697324
beatmapsets/2010056

6+6K
beatmapsets/608318
beatmapsets/1245541

3+6+3K
beatmapsets/609187

14K2S
beatmapsets/1758516
beatmapsets/1137537
beatmapsets/1307155
beatmapsets/1231943
beatmapsets/1185947
beatmapsets/1355288

EZ2AC 14K2S
beatmapsets/2017439
beatmapsets/2015672
beatmapsets/2014963
beatmapsets/1972713

18K (9+9K)
beatmapsets/1989002
beatmapsets/1940015
beatmapsets/1817234
beatmapsets/1885716
beatmapsets/1805748

18K (4+10+4K)
beatmapsets/1863940
beatmapsets/1356948
beatmapsets/1872757
beatmapsets/980384
beatmapsets/733916

14K (7+7K)
beatmapsets/1974347
beatmapsets/1709986

Pianomania
beatmapsets/1723984
beatmapsets/1726566
beatmapsets/1725315

TopRow
beatmapsets/2018489
beatmapsets/772397
beatmapsets/1061387
beatmapsets/2019173

Multiple styles
beatmapsets/1816119 [9+9K, 4+10+4K]
beatmapsets/1723637 [10K2S, 9+9K]
beatmapsets/2000775 [10K2S, 14K2S, 4+10+4K]
beatmapsets/2018566 [10K2S, EZ2AC 14K2S, 666K]

hard +1, people have been wanting for this for years now. I dont think there’s many negatives with going forward with this proposal so im all for it.

lets goooooooooooooooo

+1 yes

hip hip hooray!

12-18K (12-18 keys) (use all 12-18 fingers)

+1

+1

18k ranked gonna be crazy (dessie farm)

+1

+3

May I ask if the insularity of the 10+K community will be addressed? At this moment, only a small handful of people even bother to rank 10K stuff. How are we to expect these higher keymodes to not end up in the same dashcon ballpit?

I think everyone should be in support of these keymodes being rankable if there is a strong enough demand for it, and it was pretty much the same with 10K when that was going through its trial phase, but I don't think we've seen enough growth for this expansion to be considerable just yet.

I'd like to be proven wrong, so I'd like the following questions answered:
- How many mappers have ranked 10K? How many of those mappers are *not active* in the 10K Discord server?
- How many mappers have already mapped 12+K? How many of those mappers are *not active* in the 10K Discord server?
- Will BNs need to pass checks to ensure they can accurately judge these keymodes?
- To what extent will it be *required* that BNs have knowledge on these keymodes, even if they don't plan on nominating any of these maps?

Is the "insularity" of the 10K+ community really a concern? The same can be argued for 5K, 8K, and 9K where only a small handful of people rank those keymodes yet there are no complaints about those communities being insular and closed off. It's better that there is a community around higher keymodes to begin with as it indicates that there are people interested. I imagine that if this goes through, you'll see more people pushing higher keymodes now that there is incentive to do so (getting the map ranked instead of graveyarded or loved which gets less interaction than ranked).

To answer your first few questions regarding how many people have ranked 10K+, K0nomi's post here gives a good handful of 10K+ maps to start with community/forums/posts/9181718. I don't understand why it's such a concern if they belong to the 10K discord server or not, it's just a platform to gather like-minded people about upper keymodes to help each other out and push 10K+ maps.

Regarding BN judgement, BNs have never been required to know more than just the keymodes they nominate, i.e. there are many BNs who are 4K only and don't nominate other keymodes, so it'd likely follow that BNs don't need to be capable of nominating 10K+ maps if they decide they don't want to, assuming they're proficient in another keymode of their choice. As for checks to whether the BN is capable of judging or not, I'd imagine it's the same as it is now where their judgement is evident in the quality of the maps they push - regardless, this is something that is up to the judgement of the NAT come evaluation time for BNs as to whether they're capable or not.

+1

Here's some video precedent for 10K2S, EZ2AC 14K (with or without scratches), and 4+10+4 style 18K in mania.

https://www.youtube.com/watch?v=_BtLOOXzUN8

https://twitter.com/Hugged10K/status/1674141707519115274

https://www.youtube.com/watch?v=iNTyLLFdc8s

Hello !!!! Really big fan of 10K+ here ! I'll be writing down my thoughts here and also let you guys know on what the 10K+ community is up to at the moment!








I hope I didn't miss anything too important!

> Is the "insularity" of the 10K+ community really a concern? The same can be argued for 5K, 8K, and 9K where only a small handful of people rank those keymodes yet there are no complaints about those communities being insular and closed off.

this proves my point btw

> I imagine that if this goes through, you'll see more people pushing higher keymodes now that there is incentive to do so (getting the map ranked instead of graveyarded or loved which gets less interaction than ranked).

I take it this is happening for 5K, 8K, and 9K then?

my whole point is there needs to be *more* incentive than just having the option to rank these keymodes. this applies to 5K, 8K, and 9K as well as 10K and higher. more *active* encouragement of participation. otherwise, insularity is going to be a problem, it has been a problem for 5/8/9K. it's a bit less of a problem in 10K but it's still noticeable.

EDIT: the approach that the person above me is taking is a very good start btw. consolidation of resources for approaching the gamemode is good. what would also give this proposal more merit would be a *fully drafted Ranking Criteria* for all the gamemodes.
same applies to something like 3K, which i have discussed on here before.

I'd argue being insular is a symptom of being niche. For playstyles with small playerbases, people will naturally just congregate towards dedicated hubs (in this case, the 10K+ discord server) rather than forming "diasporas", like how 7K and 4K have many, many subcommunities. It's the dedicated hubs that provide the majority of resources for the niche playstyles in question.

But the thing is, the 10K+ community is doing all that it can to be as accessible to a wider audience as possible. We're compiling resource sheets, we're linking skins and keybind layout info in our map descriptions, we're spotlighting other obscure playstyles, and we're bundling our maps into packs for easy content access. 10K+ has an active community doing all that it can to break out of insularity.

12K+ shouldn't be rankable due to breaking out of insularity, it should be rankable to help it break out of insularity.

Just as an extra tidbit, since 10K became rankable in 2021, it surpassed the total playcount of 8K and 9K combined:

-mint- wrote:

the approach that the person above me is taking is a very good start btw. consolidation of resources for approaching the gamemode is good. what would also give this proposal more merit would be a *fully drafted Ranking Criteria* for all the gamemodes.
same applies to something like 3K, which i have discussed on here before.

The ranking criteria will come soon the plan is to rank one of each keymode here and then draft a ranking criteria for all other maps of those keymodes. I think getting feedback on different difficulties is important which is why we have asked the 10K+ community to provide us with a spread for each of the new keymodes for the ranking criteria. This way we can get more feedback from players and we have an example of what a 10K+ spread could look like.

hugged: i like your response. i see what you guys are doing (im in the 10k server) and its all good stuff. i do hope to see the 10+k scene as it expands in the future. the 10k community has clearly done more and has been more passionate for its keymode compared to 8k or 9k.
however, even 5k, a larger keymode than 10k based on your numbers, is suffering from the same treatment that im warning about. as a keymode, it is fairly neglected, even if it has a sizeable community. my skepticism purely comes from this angle: how can we ensure that there is the proper infrastructure and protocol to keep 10+k thriving when it very clearly is not being utilized for already existing keymodes such as 5/8/9k? why should more resources be devoted to new keymodes as opposed to already existing ones, when it could potentially create more problems (especially since multiple RCs would need to be drafted in the midst this process, which i think is very silly and absolutely should be done beforehand to mitigate concerns)

upupu

I just want to say for the record I’m 100% for this. This community of people does exist.

Also would like to say even if this area of people had less treatment than that of 5k, it’s still worth it because it’s a community of people. Nothing such as how much attention it’ll get should provide reasoning into why it should/shouldn’t get ranked.

(I’ll look at this more just saying my opinion then later today providing a longer explanation)

@mint I see where you're coming from but I personally don't think playerbase should be seen as so much of an issue with ranking a keymode. I feel that as long as there are people that play it and care about it, even if the number is relatively small, then it deserves to be recognised by the ranked section. Plus, spending time in the 10K+ server has shown me how eager everyone is to try their hand at ranking >10K.

The community is also starting work on drafting RC, I don't think it will be too much of a problem if the ranking test happens before the RC are fully implemented, as even if they aren't directly there, any concerns that the RC would implement would be brought up during the modding process of these beatmaps.

I'll also mention my personal thoughts on how >10K RC could look here, for sake of keeping this thread in the loop:

So, after you get to >10k all the keymodes become "dynamic", as in your fingers have to move between multiple keys. The thing is that this causes the idea if "keymodes" to break down in favour of "playstyles". For example, the difference between 16K and 18K becomes almost meaningless, as there are many different playstyles for each keymode. This means that just "16K" or "18K" alone are not enough to describe how a map is played. even within a keymode you can have many different playstyles, such as in 18K the most popular playstyles are '9+9K' and '4+10+4K'. Because of this, it ends up making a lot less sense to have keymode-specific RC, and you'd have have to make it playstyle-specific.

Now the problem with this is there are many playstyles, and if each was accounted for, it would bloat RC significantly more than if you would add one for each keymode. However, many playstyles share elements and can be categorised in to very few general categories. for example, '9+9K' and '7+7K' aren't even the same keymode, but they're played with generally the same layout and the RC can be generalised to include them all.
It's also very common for >10K maps to have scratch lanes such as with 'EZ2AC 14K2S', '14K2S', '10K2S', etc. The inclusion of scratch lanes can also be generalised into a section for how you should treat scratches when they're in >10K keymodes. This doesn't even have to have its own rules, as you can define scratch usage in terms of the existing RC pretty concisely.
Then the other way I would categorise playstyles is ones that are closer to the static <10k keymodes but with "rows". This would include playstyles like '666K' and '4+10+4K'. these playstyles have each key assigned to a specific finger, where in '9+9K' for example, multiple fingers can be used in a specific key and its more "free" in a way. Here you'd define what you are allowed to do with 2 keys that share the same finger, and how you are allowed to "switch rows" depending on the difficulty of the map.

Anyway, if something similar to this is considered for the RC it would allow the RC to consider >10K without being completely bloated by the amount of keymodes/playstyles >10K brings. It would also be good to hear other's thoughts on this and there could be better ways to categorise playstyles.

-mint- wrote:

At this moment, only a small handful of people even bother to rank 10K stuff. How are we to expect these higher keymodes to not end up in the same dashcon ballpit?

Is getting ranked maps the goal for a keycount to be relevant? Since people might be playing unranked maps, bms converts or even ez2ac converts if they have access to them, time before 10K was considered to be rankable as a keymode, even playing those charts in their native platform (bms clients/ez2ac cabs), same for 14K and above with bms and IIDX, also not everyone has an osu account to tell about "a small handful of people"

With my knowledge and experience in DP styles, I would be happy to help community by mapping some songs with 14K/14K+2 styles. But a quick question; do we form the RC first or we evaluate the maps first before forming the RC? Due to the broad amount of playstyles, it's quite hard to write the formal RC rules and guidelines.

While there is some overlap with superusers that play (and map) several different keymodes, I would argue that opening new keymodes to be rankable would simply urge more users to be active, as most of them prefer to stay at a single keymode.

Ryu Sei wrote:

With my knowledge and experience in DP styles, I would be happy to help community by mapping some songs with 14K/14K+2 styles. But a quick question; do we form the RC first or we evaluate the maps first before forming the RC? Due to the broad amount of playstyles, it's quite hard to write the formal RC rules and guidelines.

From what we know, we'll be Qualifying the maps before finalizing any sort of RC. An RC can be worked on and published afterwards once we've proven nothing breaks, specifically pp wise.

Vouch

This is an amazing decision
Goooooo

The idea is to get more players to try a wide variety of keymodes, so getting more keymodes ranked will not only attract more players to play these, but also give a try to map 12k or higher, therefore a new player that is introduced to XK will have a whole playlist of maps to try out. I mean, hey, you can't just stay at one keymode, you have to try others. This is more like an invitation for them to try out and challenge themselves for higher keycounts!

Also my answer is a +1 to Sebaex. There is a high percentage of players that would rather search by ranked other than search by pending or graveyard (the default setting for beatmap listing is "Has Leaderboard"). I believe one of the objectives is to get these keymodes in a ranked state so that it has greater visibility for the average players. I am not totally aware of this part, but I noticed that 10k got higher visibility since maps of this keymode started to get ranked, so it can apply to other higher keymodes as well.

if this gets implemented, gonna start learning these

As a 10k average player I would say that it could encourage more players (even more than 10k is already doing) such as me to try higher keymodes. I share the point of view of Pachekin since trying to make them rankeable would make it more visible for those who randomly check the latest ranked maps in maps page. Also because the "ranked" status gives a sense of quality in maps, so people are more attracted to them (at least that is my perception)

Besides that, the 10k+ community will grow even more and for those who are already in they will be more attracted to the game since their playstyle are being considered as a way to get in the leaderboards.

I'm also aware that making a RC for every keymode style will be quite large due to the great amount of playstyles and human hand boundaries lol

So.. I'm positive to say that my vote is affirmative

let's goooooo

+1

Glad to see ample documentation and working samples being brought to the table. Best of luck to everyone involved!

Overall, I'm ready to play 10+K(use no less than 10 fingers), so +1

K0nomi wrote:

you can have many different playstyles, such as in 18K the most popular playstyles are '9+9K' and '4+10+4K'. Because of this, it ends up making a lot less sense to have keymode-specific RC, and you'd have have to make it playstyle-specific.

Now the problem with this is there are many playstyles, and if each was accounted for, it would bloat RC significantly more than if you would add one for each keymode. However, many playstyles share elements and can be categorised in to very few general categories. for example, '9+9K' and '7+7K' aren't even the same keymode, but they're played with generally the same layout and the RC can be generalised to include them all.
It's also very common for >10K maps to have scratch lanes such as with 'EZ2AC 14K2S', '14K2S', '10K2S', etc. The inclusion of scratch lanes can also be generalised into a section for how you should treat scratches when they're in >10K keymodes. This doesn't even have to have its own rules, as you can define scratch usage in terms of the existing RC pretty concisely.
Then the other way I would categorise playstyles is ones that are closer to the static <10k keymodes but with "rows". This would include playstyles like '666K' and '4+10+4K'. these playstyles have each key assigned to a specific finger, where in '9+9K' for example, multiple fingers can be used in a specific key and its more "free" in a way. Here you'd define what you are allowed to do with 2 keys that share the same finger, and how you are allowed to "switch rows" depending on the difficulty of the map.

I agree, 10K+ can be categorised into three groups

1. Freeform type(or controller type)
   10K2S, EZ14K, 9+9, basically anything that implies arcade-like key layout(or piano-like like Pianomania)
2. Fixed type
   4+10+4, 666K, basically anything where you have a base <=10K keymode with some extensions above and/or below
3. Other
   Stuff like the osu!typing loved map

And since the last group is too experimental for rankability anyway, I think just having two sets of criteria for all of >10K(freeform and fixed) might be enough. Heck, I'd argue these could also extend to 7K1S vs 8K, but I'm biased because I play more with controller-type layouts(even 7K and 9K sometimes) and also am not that good.

Akira Morilas wrote:

Heck, I'd argue these could also extend to 7K1S vs 8K, but I'm biased because I play more with controller-type layouts(even 7K and 9K sometimes) and also am not that good.

I don't think these apply to 7K1S vs 8K as much, since the difference with that is the "dynamicness". For example, the difference between 10K and 4+6K is that 10K is static, you dont need to move your fingers at all and each key has its own finger, but in 4+6K, Each key has multiple fingers and you have to move around so its dynamic. In 7+1K vs 8K, they're both static so despite using different fingers, they aren't as different as some 10K+ playstyles are from each other.

Akira Morilas wrote:

I agree, 10K+ can be categorised into three groups
Freeform type(or controller type)
10K2S, EZ14K, 9+9, basically anything that implies arcade-like key layout(or piano-like like Pianomania)
Fixed type
4+10+4, 666K, basically anything where you have a base <=10K keymode with some extensions above and/or below
Other
Stuff like the osu!typing loved map

I do agree with your categorisation, but also think a subcategory needs to be included for scratches. Also, after looking at things closer it's become apparent that playstyles can have elements of both "freeform" and "fixed". for example, 14K2S is played with an arcade controller-like layout, but the way you position your hands on the keyboard is relatively static when compared to something like toprow. Despite this, there are some patterns that can be applied to almost every playstyle that "break the mould" such as crosshands (which force the player to move one hand to the other side of the keyboard), fatfingering (forcing 1 finger to press 2 keys at once), etc that can increase the freeform-ness of even row based playstyles. In the end I think most guidelines will be all encompassing, applying to all of these categories and thus all dynamic playstyles.

Hugged wrote:

Just as an extra tidbit, since 10K became rankable in 2021, it surpassed the total playcount of 8K and 9K combined: (Large image)

Hi!

Someone pointed that my image is quoted here so I need to do some clarifications:

1. First, I queried the data from one of the many data dumps available on data.ppy.sh, the dump I had doesnt take into account 2023 data, I can update this later if needed as there is a new data dump from days ago.
2. Second, while yes the playcount number on 10k ranked diffs on 2021-2022 is higher than 8k-9k ranked diffs in the same period, the issue on this is that 10k had a lot more diffs ranked but not as many plays per diff as noted here:

If we remove the ranked requirement (we add loved diffs) doesnt help that much:



While i dont particularly reject 12k-18k to be ranked I have similar concerns like mint pointed, for a keymode to be fully addressed things like RC, difficulty processor, pp awarding and others should be taken into account and not be problems that someone else might need to tackle, and currently those just add complexity to things that i guess are being only looked on 4k and 7k.

WalterToro wrote:

Second, while yes the playcount number on 10k ranked diffs on 2021-2022 is higher than 8k-9k ranked diffs in the same period, the issue on this is that 10k had a lot more diffs ranked but not as many plays per diff as noted here

Hello Walter, regarding the low play counts per difficulty on 10k. I would also like to add that the trend we see for 10k ranked maps is that the lowest difficulty almost always has a significant decrease in passing rate compared to that of higher difficulties. Note that this trend does not exist for the majority of 4k-9k maps (that I've checked). The map where you can most obviously see this trend is on Mad Piano Party: beatmapsets/1756251#mania/3594045

The lowest difficulty has a passing rate of 14.3% with 8371 playcounts and the highest has a passing rate of 31.3% with 1461 playcounts. From this statistic we can infer that most new players to 10k contributed to the play counts of the lowest difficulties and mostly seasoned 10k players contributed to the play counts of harder difficulties. If we check the play counts for a 10k map that is not ranked we do not see this statistic of lower 10k difficulties extremely low passing rates compared to ranked and this is likely because most people who play unranked 10k maps are already well-established within the 10k keymode:
beatmapsets/1341108#mania/2778096

This implies that ranked maps have brought a fresh new audience to 10k and that higher difficulties don't have enough play counts because these newer players have not obtained the skills to play harder 10k maps yet. In other words, the rankability of 10k has brought 10k to the attention of hundreds to thousands of new people when previously it was mostly people who stumbled upon the keymode who started playing.

WalterToro wrote:

Someone pointed that my image is quoted here so I need to do some clarifications:

Hey Walter, I feel really silly now for using your screenshot all willy nilly, I wasn't aware of the context that it missed, so I'm sorry about that.

In regards to the other concerns, mania RC is currently being worked on collaboratively between the NATs and 10K+ peeps to be more airtight in regards to 12-18K.

In terms of SR, I'm not convinced it should be a large obstacle for the higher keymodes. In you don't play the higher keymodes, it might come as counterintuitive because more keys opens the possibility for more density, but most maps for these keymodes are significantly underrated by SR because a huge portion of the perceived difficulty for these charts comes from the hand movement required to play rather than directly from the chart's density, which SR calculation fails to consider.

This means that it would be really difficult to make reasonably playable maps that also abuse the SR system.

Hi Yucky, I would love to put more eyes on how much players are playing a certain keymode, sadly I need some time and a little computing power to see if people are trying 10k from their main keycounts, will try to query that.

I don't doubt that people will find these keymodes interesting and will try those at least once, heck I even tried 9k in the past when looked into other keycounts over 7, but again, I think that adding keycounts into ranked should be more than just an idea to be approved (it will be like the rest of the other keycounts aside that 12k to 18k currently is played as coop in stable) but a good implementation of all systems related to ranking (RC, BNs, AntiMeta NAT, pp, diffcalc, rankings, etc.)

We will see how this goes.

@k0nomi @Akira Morilas

I'm pretty much in agreement with Akira's Freeform/Fixed categories (with the addition of a scratch category), but I have a different idea on the exact definitions/naming convention that may allow a much more compact RC.

@DesKurisu has in the past coined the term `Islands`, which I think would be a great term to adopt into the glossary as follows:

`Island: A set of keybinds that is physically separated on the keyboard from other keybinds for certain playstyles.`

`Island` would cover EZ2AC 14K, which I'd say is the biggest playstyle as of right now that requires moving your hands across the keyboard.

Next, I think the term `Multi-row playstyle` would be a good umbrella for styles like 4+6K, 666K, 4K10K4K 18K, BMS DP, and 9K9K, and it could be adopted into the glossary as something like this:

`Multi-row playstyle: A keybind layout involving the use of two or more adjacent rows on the keyboard.`

Then my idea for RC using these terms would be similar to as follows, giving these categories just one bullet with 2 or 3 sub-bullets containing general guidelines per difficulty:

```
Easy Guidelines:
- All other Easy guidelines
- For maps utilizing an n+2 special style:
---- Stuff about scratches played at the same time as normal notes
---- Blah blah
- For maps utilizing an island playstyle:
---- Stuff about the frequency of switching islands
---- Blah blah
- For maps utilizing a multi-row playstyle:
---- Stuff about cross-row chords on the same hand
---- Blah blah
```

```
Normal Guidelines:
- All other Normal guidelines
- For maps utilizing an n+2 special style:
---- Stuff about scratches played at the same time as normal notes
---- Blah blah
- For maps utilizing an island playstyle:
---- Stuff about the frequency of switching islands
---- Blah blah
- For maps utilizing a multi-row playstyle:
---- Stuff about cross-row chords on the same hand
---- Blah blah
```

etc.

The main potential con I see with this approach is that `Multi-row playstyle` might we way too general, because it would include, for example, 4K10K4K AND 9K9K, which are two different 18K playstyles that just happen to use two adjacent rows.

Edit: I have my own GitHub fork of the ranking criteria as a "what if" containing some of my own as well as others' ideas to be more accommodating for 12-18K, WIP: https://github.com/sinanates17/osu-wiki/blob/master/wiki/Ranking_Criteria/osu!mania/en.md

Edit 2: That github fork is now pretty much a complete draft covering everything 12-18K