osu! | To kill a pp system (ppv2 Part 2)

[@BigDBrian]

getting rid of pp entirely is like cutting your leg off because you had a sore foot

[@momokomomose-virtualyoutub6677]

like cutting your head off because you had a headache

[@bananya6020]

i do that

[@aaroningots]

@@bananya6020 lets go! me too. Legless gang rise up.

[@bananya6020]

@@aaroningots yeaa- wait how are we supposed to rise without legs

[@aaroningots]

@@bananya6020 alright, FIRST... We gotta rent a fuckin construction crane. Then we will rise. Many are intimidated by my massive intellect.

[@TheEpicZJ]

I've been working on my own pp system for a couple of weeks now ands it's starting to become close to finished. What mine does differently is it accounts for "irregularities". Spacing irregularity = difference in spacing between two notes compared to the last two notes. Angle irregularity = same but with angles. This makes technical maps worth alot more than before. It also makes jumps/notes with lower angle worth more and there are some more things added that impoves it. I will probably post a gui for this somewhere pretty soon so people can try it out.

[@danieldragojevic3016]

The more complex = more lag

[@human3507]

Daniel Dragojevic Osu can be played in almost any computer, which is almost impossible to 90% of the games in this age.

[@bonfist7277]

human :3 I have an HP office laptop with windows 7 (and a shitty half broken mouse lmao).... Resolution is as low as possible, compatibility mode, etc. still freezes every now and then.

[@bonfist7277]

Some guy looks at his computer getting 50 fps on a very frame straining game and tosses it in the garbage while i would be in love if i could have something like that.

[@human3507]

Bonfist I have an normal hp laptop and it only freezes When I play in easy mode :v

[@danfoR.]

Damn man props to you. I´ve never watched a whole pp discussion because most of them throw so much bs at you, but you managed to make an objective discussion, remain neutral and give several povs on each argument. I´m really impressed. Very good video! Personally I feel like osu!skills is the way to go with a balanced way to calculate each skill section and a total pp score. This is also a possibility that Peppy was already thinking about (he answered a tweet or a reddit comment, I´m too lazy to find it rn) In the end all we can do now anyways is for lazer to be released and replace the old client, THEN and only then will we see big changes to the pp system and I´m sure that they will come since he confirmed that he also thinks that the pp system is broken.

[@namYa]

I also feel like accuracy is so underrated by this point. How is lewa's 50 unstable rate not more impressive than some dude playing a jump map with 97% accuracy??

[@namYa]

Yeah, slider being a certain 300, all patterns being valued the same, and the complexity of beat patterns not being valued really sucks. I feel like accuracy should be the main point of a rhythm/music game.

[@huanpham9024]

that's why sliders dont give pp since OD doesnt apply to sliders

[@hosthan8782]

namYa that's because you're good at it, think in other

[@namYa]

Hosthan Everyone should be good at accuracy in a game that's mainly centered on playing music...

[@namYa]

Huan Pham Huh... So that's why map with many sliders have lower stars?

[@phths]

i honestly cant imagine osu without one number determining your rank. that would be like dividing your csgo rank into aim, sprays, clutches etc... i guess im just too used to the current system to even consider change

[@luc222GH]

the difference is that in cs all your skills go towards winning the match, while in osu different skills are displayed achieving different scores.

[@johndough7994]

isnt that kinda like hltv rankings?

[@Troonald]

i like the idea of having a dan rank alongside the pp system, then it could be displayed on your profile or even batter, as a symbol in front of your rank or appended to your name

[@NightyFloes]

But then everyone would see im noob and bully :c

[@Troonald]

NightyFloes that’s the point of having the dan rank, it encourages ppl to get better at the game rather than only getting good at pp

[@Minib34ts]

Everyone keeps talking about shifting the mapping meta to something different when the basis of the whole problem is what derived a mapping meta in the first place. An ideal mapping meta would be to not have one.

[@joinmyjourney8407]

But there are plenty ways to elaborate on that kind of system, like stated in the video, one solution is to make the Ranking based on different categories, and there could be seperate leader boards for each map type, and a total ranking system, for the players ability in all of them combined, although it would be hgard to work out the maths in that, I see that as the best way

[@hypercyte]

mfw both pishifat and mo upload at the exact same time

[@DaxxelOsu]

rank gohan

[@austcol2010]

Both stopped uploading

[@beeschan7301]

@@austcol2010 :(

[@Gister]

"The system mainly favors maps with many circles and jumps" ooof how that aged well

[@Lance0]

xexxar in 2021:

[@Deastrisk]

@@Lance0 he still forgot about sliders ticks tho

[@Lance0]

@@Deastrisk I'm actually surprised slider ticks in straight sliders aren't considered as wide angle jumps (for the aim value only ofc)

[@Deastrisk]

@@Lance0 if that was the case, the meta would be tick rate=4

[@asonate]

I feel like bringing back Ranked Score and merging it with PP could work pretty well. Essentially still gaining ranked score by playing and amplifying that score depending on how hard the map was (we could use PP for that). With that short jump-heavy maps would still give a ton of points, and longer maps with high score numbers could compete. Now one could argue that people could farm long low-star marathons, but i think that the PP-scaling could work well to avoid this.

[@cheshire1]

Would the individual scores still be weighted before they are combined, so that only your top scores matter? If so, then you would basically just be buffing long maps, by multiplying their pp with a scaling factor (the score). Buffing long maps is generally a good idea, but score grows with the square of the object count. So an 8-minute-map would be worth 16 times as much as 2-minute-map with the same pp value, which is excessive. There should be some limit where further increasing length doesn't increase the end value by much.

[@puddingosu3326]

Been waiting for this :) Your videos are good, I watched all of them last month.

[@KorewaEden]

Really informative, got my sub :0

[@logmaster534]

KorewaEden??? Do you still play osu?

[@zjmullan]

Your videos are very well-made bro, I saw the other one about 0 bpm a while back and it was also very good. It's got 1.4 Million views now and you only have 11.1k subs! Major congrats dude you're definitely underrated right now and should hopefully be growing huge soon.

[@num-ber]

Yo, I REALLY like this Dan idea, imean players are already sort of categorized by how many stars they can play, i.e 5* pass is a milestone, 6* pass is a milestone, etc. Probably a whole ranking system would require a little more than just displaying each user's highest "Skill"-category achievement if you want to replace pp entirely, but I like the direction of that thinking honestly

[@iliamanolov5926]

How about having pp + Dan simultaneously? For example, every Dan contains 1000 players and inside each Dan you can compare the skill of each player using pp. This solves the farming issue for your average player because he can't climb ranks by only scoring pp. You can even make the Dan sizes exponential (i.e. Dan 1 has a players, Dan 2 has a*b, Dan 3 - a*b^2, etc). This would make Dans much more important for better-ranked players since when keeping the sizes constant you either end up with too many Dans or top players are all grouped in the same Dan. The other cool thing about exponential Dans is that you don't need to make too many of them because of how fast exponents grow which would make selecting the maps much easier. That's just my 2c.

[@AMo_osu]

Still requires people to actively participate in dan courses for it to be accurate, but it's a cool way to combine the two if we did end up getting dan courses.

[@charlesfreeze3997]

The only issue with a dan system would be what if the player doesn't LIKE the songs in the courses? It wouldn't be fun to play if you had to play something you didn't like. I have Touhou..metal, and such like that, I'm not a pop person and if a course was filled with pop or wub maps then I'd just not really be into it and not feel like playing it. That's just how I feel about Dan courses.

[@osutips]

I would be happy with any system that doesn't reward for FCing map only. That's really my main complain about the pp system in STD. The best solution imo would be something that sort pp by subcategories (as you said, tho there isn't really a good exemple of that now, osu!skills calculation can be abuse really easely in something catégorie) and most importently that you don't need to fc a map to get pp. The location of a miss across the map should not be a factor. I would rather take thé number of miss and the accuracy to juge the performance on a map

[@nihildwo4874]

i can definitely agree on this. get 1 shitty miss right and the middle and the score is basically worthless, like this one: kzbin.info/www/bejne/gWO4naqBmpyHq9E

[@osutips]

but if you get like 8 miss at the very end with a shitty acc you get pp... that's shit yeah

[@meta04]

@Nihil Dwo oh also that video is dead now that its creator got three copyright strikes :( edit: they got reinstated :)

[@nihildwo4874]

@@meta04 ah. the video was from circle people. wakson did a 1 miss 0x100 0x50 score on road of resistance, but because the miss happened to be right in the middle of the map, he only got like 390 pp instead of 600+.

[@meta04]

@@nihildwo4874 oh that's so sad 1 miss off an SS... reminds me of that one time I got 700+ combo towards an S+FC on osu.ppy.sh/beatmapsets/123593#osu/315552 for the "3 star S" medal it isn't enough for it to be an S... it has to be an S+FC misses already strike off an S (1miss 0x100 0x50 is an A on any map of 10+ objects) but you can't sliderbreak for the medals either

[@Spearra]

While unwieldy, maybe it might be possible to mix the Dan and osu!skills systems. For example, having different Dan courses for specific skills. Then use the osu!skills system to somewhat assess how well a player is in either specific Dan courses, or overall.

[@ExplosiveBrohoof]

No matter what pp system you create, players and mappers are going to do everything they can to promote maps that exploit it as far as possible and reward the most pp for the least amount of effort. Even doing something like dividing desirable skills will promote certain maps over others, either by favoring maps that isolate skill (so, for example, making a stream map with dead-easy jumps) or by favoring maps that are explicitly well-rounded (why play a map that gives me 200 jump pp when this map gives me 150 pp in everything?). You'll still have overweighted pp maps, they just may come in different flavors.

[@undisabled1552]

I like your ideas a lot, but what if instead of Dan Courses it was just a singular map you had to pass like a test. On Warframe they have a "test" that gets progressively harder that you have to pass every time you want to rank up. If we had the normal ppv2 system and just added 10(or whatever number) comprehensive Dan maps, couldn't we achieve a solid system where players that want to show off their skill, but don't want to grind pp can say "look, I'm a 10Dan" and the 5Dan that likes to grind can still say "look I did all these jump maps 8^)"

[@Menolifee]

i love the idea of the Dan system. not to replace pp but just to be there as a different kind of challenge kinda like a "campaign"

[@Menolifee]

could have different "courses" with different map pools for different skill aspects in the style of osuskills.tk. and theyre all split into lets say, 20 tiers? and the average of all these Dan tiers could be used to form an overall Dan lvl or something. :P think it could be a fun challenge to get your tiers up. and i dont know how it could be abused if the map pools are good. it could be a fun alternate way of competition.

[@Wutever]

before watching past 2:02 in the video i actually voted to kill off the pp system. was hesistent though because the question was both sides of the extreme but i would honestly have 0 problem with pp temporarily dissapearing and idk make score the ranking metric again temporarily in the meantime pending a new ppv3 algorithm build from scratch. thats not gonna happen though.

[@movagalbastomp]

R.I.P osu!'s PPv2 Score System 2017-2018 (1 year of living? WTF!!) *Emoctional Flute Meme Music*

[@Daniel-fk8cb]

no what?

[@dan339dan]

I think calculating the randomness, or the lack of pattern of the objects and sliders may be able to reflect the difficulty. And I think that can be added to the pp system. The higher randomness of a map, the more difficult it is. And randomness would be tied to the speed factor, then there is no chance for exploitation.

[@gobahgaber6909]

a lot of players are motivated to play to see their peepee rise :)

[@Haglax]

I can imagine overall pp calculate by suming up pp gained from several skillsets (from the osu!skills website) and withing each skillset there is a weighting system. That way it encourages to play all kinds of maps.

[@akirachisaka9997]

I feel like the ultimate solution should be someway that use a machine learning type of system to decide map pp purely on how well other players perform on them? As in, a system that completely ignores how a map is like, but solely decide map pp based on how well all players perform on a map? This do feel a bit like going back to ppv1. But I do feel that it’s really hard to make an algorithm that can tell the difficulty of a map without player input. And a new pp system might really want to watch how players do on a map to decide the difficulty of the map.

[@snalgae]

Why one pp system, what about many pp systems?

[@dacasman]

Idea 1: Rank is based on your average position on the leaderboards of each map. Pros: encourages playing a variety of maps and eliminates farm maps. Cons: based on play count and not raw skill. Idea 2: Each new ranked map is sent through a panel of judges that place a PP value based on how difficult they feel it is. Rather than an algorithm deciding the PP value. The average assigned PP value from each judge is averaged. Only con I can think of is it would require a trusted panel of judges to place a PP value on each new ranked map, which isn't too much different from how maps get ranked anyways.

[@maxtoosick]

1:48 I’m literally dyingggggg 😂😂💀💀😂

[@baccarah7010]

Imo the best solution is to have the osuskills system, and just have global rankings for each skill, that way the playerbase gets more diverse too. Plus, everyone can play the maps they actually enjoy, while still climbing in their respective skill, say stream for example.

[@huuishuu]

The ending made me chuckle, nice one.

[@BassChip854]

possibly giving a rank for each skill like you could be #100 out of 1000 for jumping and then #520 out of 1000 for streaming and avg them out which would mean you can just be good at one thing but wont be high ranked as you are not so good at other things therefore stopping jumps maps booting peoples pp so much but i cant see them doing this as it would be fairly difficult to create and make a system that decides the performance of the skill

[@swerro9978]

Honestly the way you're talking/theorising for the best possible ppsystem is basically heading towards the almost exact same goals as pp+. people spend a damn load of time in jumps and therefore even on pp+ you're going to see jump maps in the topscore rankings , simply the most popular practiced skill.

[@kingsiracha4565]

Etterna kind of does what osu skills already does but for DDR although the one difference is that for the overall rating each of the skills get averaged for your overall average rank. That way ranking is more so based off of who has mastered the most skillsets the best and there would still be one main leaderboard but you would also be able to see how people rank in each individual skillset.

[@AnnexVideos]

Maybe we could objectively differentiate and categorise every osu map, just like the genre of the map is determined when the map actually hits ranked status, the bn's categorise the map to different categories (Stamina, Aim, Reading, Flow, etc). Say a map is categorised as: stream and stamina heavy map, then if a player gets a good score on them they get stamina points that will increase their total pp in a way. But then we have to categorise every osu map that ever been ranked in the database which will be death. Unless the original mapper is tasked to categorise their own map, but there are still mappers that don't play the game anymore so to categorise 100% maps out there is still a problem.

[@nintailsTF]

That outtake was great

[@yorimirus]

1:20 whats the map in the background? Or the song at least

[@DoofusLoopus]

The dan system has a fair share of its own disadvantages (a bms player here): -The maps that are included in a dan are of various difficulty and the gap between them can be neglectable if you are a beginner (and are simply going for a KYU clear) and VERY NOTICABLE if you want to test yourself on an insane scale. Imagine having a 0.5* gap in every single dan between the songs. -However, there is no such thing as 'grading' your dan clear. It is either PASS or nothing which goes to show that you can SOMEWHAT pass certain patterns but that's about it (despite having a pretty harsh life gauge system). -Since charts in a dan are supposed to get progressively more and more difficult, in bms dans are usually nothing more than your typical stamina test. Is it better than a pp system? No, these are 2 different things for different purposes. In iidx, as well as in bms, your skill is usually measured for each map separately in so-called EX points which indicate how well-timed your button presses are. Basically BAD/MISS=0pts GOOD=0pts GREAT=1pts PGREAT=2pts. The sum of all notes that you hit on a map is your EX score. Then players simply compare their score on different maps.

[@TuneEditsfx]

The osu skills system is great, just finely tweaking it over time can make it very balanced. Other games have achived this already, it just needs a few months as an adjustment period.

[@dred4dead551]

It's been a few years since I've even played osu. How the hell did we get here???

[@fishywishyfishfisk]

Oh how I've waited for this

[@peachy7776]

very good videos and really helpful to show to new players asking wtf is "pp mapping"

[@peachy7776]

also the pp separated into skillsets is honestly the only way i can see to improve the game in a fair way its just implementing it

[@rocketlanterns]

Simple solution to this. Do what Stepmania Etterna does. Etterna doesn't have a single overall player ranking system... It has seven. As a Stepmania game, its not directly comparable to osu!standard, but a system could easily be designed for it. Etterna grants a different song difficulty and player rating for: Stream: long streams of single notes JumpStream: long streams of 2 note chords HandStream: streams of 3 or 4 note chords Stamina: long times with high density and few breaks JackSpeed: mashing the same button fast JackStamina: mashing the same button for a long time Technical: ability to play different technically challenging patterns. It also calculates Overall difficulty and skill, but I actually don't know how that is calculated. This could easily be implemented into osu by having-- Oh, just saw that this is actually brought up in the video lol.Curiously though, the system Etterna doesn't reward mods... There is one additional point in favour of this though, You aren't just getting ranked on different stats, maps will also have difficulties on different stats, so you might see a 4 star map, and then it breaks down into 6 stars stream and 3 stars aim/stamina.

[@EmiliaHoarfrost]

Dans by beatmap packs could be an impressive innovation :D !

[@BEEEELEEEE]

I have very little interest/skill in rhythm games, but this was still fascinating to watch.

[@JacquesSnacques]

I think a good idea would be to only count your total pp for your best play on every map. So, if you play 1 map a bunch of times, only your highest score will be counted toward your pp. This way, Farming is avoided (or at least gives marginally less reward) and it incentives players to be good at the game in general rather than spending all of their time mastering 1 really hard map and playing it a bunch of times. It could also be updated after every map you play as to avoid searching every map's highest score all at once and lagging your computer a bunch.

[@Tenshi_on_Val]

wait, what if there were 2 seperate ranks? one could be the dan rank, one could be the "free play rank" or something like that, then it wouldnt neglect the other maps while still having the dan ranks

[@phronesis2471]

I think a mix between Dan courses and osu!skill is the way to and for overall you can just take the average of all the skills

[@nerassmallerchannel5619]

"hello, let's cookiez- clear a few things up." well...

[@MasterAnimit]

4:54 What is this page? I searched her for a long time...

[@AMo_osu]

Ah I knew I forgot something. I took that from redd.it/86xo9i and the site is in the comments.

[@shinzuno8267]

Isn't it possible as well to change ppv2 (or 3) by making it so if a map only has jumps, the pp value gets lowered, whereas if the map has a very balanced level of jumps and for example streams (and thus requires more seperate skills to fc) it would give more pp. If done correctly this nerves short jumpy maps that are average bpm, but high level DT maps with massive jumps would still get a pp boost because of the seperate strain it requires because of the high bpm (the high bpm combined with the jumping adds more pp because again more skillsets are needed than only 1).

[@BowieProYT]

the first thing they need to do is just add acc value to sliders

[@luxxias]

a dan system seems like it would be good for tournament players, as rank limits are kind of a weird thing to go off of for tournaments.

[@velldurn5578]

I think the dan course system is somewhat complicated and hard to understand (I honestly had no idea what you were talking about when you said the whole dan system) I think the osu skills what come out a lot better and not cause it eliminates pp from just jump maps but you will be able to see other maps besides jump maps that give out pp such as an AR 8 four Dimensions map which no one would ever see cause no one would try it unless it gave out a lot of pp

[@pedrototolo8027]

instead of a system that rewards players based on all these complicated things like aim, reading, speed, etc, which are probably very hard to measure, why not reward scores based on uniqueness, for example, when someone manages to hdhr fc that map with nomod A's in the leaderboard, it's definitely a good score right? isn't that enough? (or maybe that would be even harder to program?)

[@AMo_osu]

Just because 20000 people can FC Hitorigoto doesn't mean we shouldn't reward anything for it though, in my opinion.

[@pedrototolo8027]

Well, of course we would still need some base numbers for calculations, but is 20000 people really that much to reward "anything"? Considering how many people are playing nowadays, i think rank 20k in an ultra popular map like hitorigoto ain't that bad... I really think this would be a very objective way of ranking players and creating some incentive for people to play different songs and mapping styles... Just some random thought i had while watching the video though. (Your content is really good btw, keep up the good work!)

[@AMo_osu]

Someone did have the same idea a few months back if you're interested in reading about it redd.it/78jm1b

[@thepixelpr3dator207]

lol thats gonnabfuck up so many things. what if a map is newly ranked and a good top player comes in with modded fc before there Is even a nomodded fc

[@pedrototolo8027]

yeah, plays' pp score would have to be recalculated over time, that's probably not even feasible lol

[@CapnSurvivalist]

I think pp should be based off of success rate/average scores of maps. It would have to be per mod selection as well, which would stop DT farming because if a map is easy to beat with DT, you'd get less pp for beating it with DT. Accounts for everything really. If a map is harder to do well on, it automatically gives higher pp. If it's easier, it automatically gives less pp. Doesn't matter what's in the map, just how hard it is to beat and do well on it.

[@tmrex9340]

I say we do a rating system, similar to chess. We put players who are similar in rank in a 1v1 format where they will play from a selected map pool, whoever wins the 1v1 will have their rating increased while whoever loses will have their rating decreased

[@AMo_osu]

Same problem as dan in that it requires active participation.

[@justinkim5106]

Why can't standard be more like mania where you are rewarded more for accuracy rather than combo?

[@ZARtheFREAK]

Run for your life lad, the PP Cartel is on to you!

[@Kaidxx06]

"Squeezing all of the pp juice out of it" -A Mo 2018

[@Machixo]

I'm not a big osu player but I'll play for weeks on and off. The whole reason I come back and continue playing is the satisfaction of +xXxX ranks. The satisfaction of beating a song and seeing my ranks go up. The PP system is broke but shouldn't be taken out completely. It'd stop casual players like me from coming back. Ign Machiko

[@Haganenno121]

My thoughts: 1. osu!Skills type of system will never work due to the fact that some aspects of the game are not measurable. There will probably never be a way to make HW's slider experience be worth a lot of... Anythinf, really. Same goes for Lindwurm. Even some aim elements are difficult to calculate - for example, antijumps, weird angle jumps, density. Like, even AR is not a measurable stat, because it is completely subjective - while one prefers AR10 for lower density, another might prefer AR9.5 for better reaction time. ppv2 values aim, and it still sucks at doing it properly. It would be impossible to actually make maps like Relentless. 2. pp+ is a version of ppv2. That means it is biased, and in itself bad. It just has different biases. Tl;dr making ppv3 would be impossible and balancing ppv2 would be impossible as well.

[@Krunschy]

The big problem with overweight skills in the ranking system is just how much they snowball. By that I mean when a skill is just slightly overweight, people will eventually figure out what it is and specialize in it. This in turn eventually leads to people mapping for that skill, which on the other hand helps people learn and specilize in that skill and have it pay off. So you need an exceptionally balanced system to not have that one skill which slightly easier farm snowball out of control until we arrive where we are now. Looking at things this way, maybe our current system isn't so bad and adjustments are in fact all that's required. Keep in mind that even small adjustments can make a big difference as they may allow for other skills to snowball too and catch up.

[@AMo_osu]

I believe the initial plan for ppv2 was that it would be constantly maintained and updated, but there just aren't the resources to do that it seems.

[@potatokitty]

Any new videos?

[@koopafou]

The Dan system would be better used for a training course rather than rankings, or maybe to handle tournament entry requirements, because it's more stable and players wouldn't feel too frustrated to be one rank too high or too low to enter (or they casually do another top score and suddenly they aren't eligible anymore). The beatmap packs and spotlights are somewhat similar to this, but all you can get is a medal (unless your name is Toy, in which case you also become a supporter forever), it works even with NF, and the rankings only last for a few months at most, so it's not really relevant in the long run. That aside, you haven't said a single word about rebalancing PP according to how players actually perform on the map. If the goal is to reward difficulty, then why not just reward less when the scoreboard is full of FCs or whatever relevant metric is validated by a huge playerbase (and/or more when it's a disaster) ? The only potential drawback is the computing power required to maintain all the PP buffs/debuffs and recompute the entire leaderboard, which unfortunately takes a few days (we could see that when the HD adjustments were implemented). It should be fine to do it once a month though ? Couple that with osu!skills or whatever to have a better baseline PP estimation than PPv2, and we're good to go. There is another problem though, even if you get a hold of the best system possible, people are going to bitch because they lost ranks even though they don't deserve it. And then, because they are just kiddies who don't want to put effort into anything after all, they will take petty revenges with unending DDoS attacks on bancho. This is really a touchy subject, even moreso when the current system has been running for years. Even when you look at scorev2 which is a pretty good idea to replace scorev1 aside from maps abusing slider leniency, there is a strong resistance to change and we still don't have any scorev2 leaderboard whatsoever. And then players still can't choose which score to use online even though they can have better combo and accuracy with a lower score. This is even worse when you account for mods and the dreadful score multiplier of EZ when higher AR is easier to play.

[@mooshiros7053]

A Mo: Jumps are the only thing that give pp right now Merami: HA HA HA HA HA

[@chaoticneutral1090]

Haha that outtake was good

[@arzim646]

The problem that could emerge form the osu!skills like solution would be that people will NOT try to improve in other categories. Obviously right now jumps are the most farmy, so the aim category would stay the same, people that want to be high in the aim ranking would keep doing jump maps. But let's say there is a player. I'm not very creative so let's say it's Wilchq, he hit his skill cap. He can't do fast maps, he's not that good at jumps. BUT, he's very consistent, good reading, good accuracy. Why should a player like that try to improve in speed, aim etc? Yes some sort of global ranking could be a solution, but what if the player thinks that he's just not going to be high in the overall ranking, let's focus on these smaller rankings. On one hand that system might stop the overall improvement, but on the other players that aren't going to improve as much might get their spotlight in other rankings. Either way, some sort of refreshment from the current system is very needed and having the pp system completely rebuilt will - if done well enough, it doesn't have to be perfect, just new - bring back players and make the current ones try even harder.

[@cavemann_]

Idea regarding this Dan system. Let's say there are 1000 players in each selection, then, top 10 players will go up each month if they can hold on to their rank in top 10 for a month. Although that is flawed as well, because some players will take years to go up the ranking system that way. However, I wonder if I am going in the good direction.

[@chenchen6150]

CircleNoob talking about most pp gain for a month they should separate it. So there is overall pp that you get from your entire years(your dedication). And there is how much pp you get from last month. So there is ranking leaderboard like the su website have and there is overall pp get by "bla" month. Just like that ranking chart "dat like october map ranking score chart".

[@multyrain]

how about we keep this ranking system for now and add additional ranking systems and look at the pros and cons between them to find out which one we should use?

[@Krunschy]

I mean following this algorythm of messing with some numbers and seeing how it goes and advancing with whatever works best definitely leads to great results, however the "whatever works best" part sadly basically boils down to the same problem we have now, which is that we really can't evaluate how much skill goes into a play exactly.

[@ReLoadedProject]

Big brain keep pp and add dans that replace the pointless leveling system

[@dangeroustoaster2688]

I think we should have two systems. We are far too invested in pp to take it away. My idea is that we should calculate pp based on standard deviation from scores, and number of passes. We will need heavy ai to calculate harder maps however. or lots of sample data.

[@huanpham9024]

so how do u calculate pp for new ranked map?

[@dangeroustoaster2688]

Huan Pham spacing between circles, cps, and combo basically

[@huanpham9024]

i mean u said they should calculate pp based on scores and number of passes right? then what about new map? and if someone fc a new hard map then noone does and a few years later a lot of players fc it with same acc will they get same pp?

[@sayorikirigaya]

FYI there's also Dan Course for osu!Catch. Just saying. ;)

[@fakeflake5925]

osu! Skills is the best one in my opinion aswell as that etterna has that way of ranking but i think there should be a cap on how well ranked you are in one skill set compared to another bc people in etterna just play one style and get rank 1

[@infiniterune7101]

Why not have map difficulty ratings be based off of the performances of players that play them? If a map has a very low average accuracy or success rate it clearly is probably hard. Low ranked players beating the map would lower the pp more than high rank players and visa versa for failing. You would probably need to pair it with the old pp system as a starting off point for maps that don't have a ton of plays but as the maps get played it would simultaneously raise the pp for difficult maps that don't conform to what gives lots of pp and would very quickly diminish the returns for pp farm maps that get super consistently cleared.

[@XtoGTA]

"Other rhythm games don't have a ranking system" except that they do?

[@MohabBlaze]

What about a competetive elo ranking system like dota 2 in multiplayer?

[@AMo_osu]

Same problem as dan in that it requires active participation.

[@geoidYT]

What would be the req for Kaidan? 👀

[@kaimuu]

if the pp system would get removed i would play a little bit but nowhere near as much as with the pp system. It's just a huge part of the fun i have with the game

[@Vladislavchooo]

Take a look at how accurate EtternaOnline 's ranking system is. Yes there are issues with some patterns however keep in mind that it's been developed by mina and some community feedback alone in the course of no more than a year. You might find cool concepts there however I doubt Mina would want to help osu! in any way.

[@vulcanprincess1584]

a dan system would be SUGOI but yeah, people shouldnt be required to do it to rank up, it would just be a matter of prestige then? idk

[@notkdoge4217]

i kinda finished anime ban with HDDTFL however, i tried using DTHRHD it doesn't get me to leaderboard because there's no spinner cause some overlap on other peoples rank Edit : it only on easy mode that caused some problems if ppl trying to get to leaderboard, ppl uses FL instead of HD

[@huanpham9024]

FL gives 12% extra scores while HD gives only 6% so use FL on a low star map is better right?

[@walmartpudding8047]

per-hit object will solve everything

[@rijrya]

what if the pp system used humans instead of an algorithm? for example there could be a team of experienced players who would decide the amount of pp a map would give if it was ss'd, and the system would calculate the pp for lower accuracies, with mods, and the combo based on the predetermined amount. it might not be very accurate, but at least it can't be exploited

[@AMo_osu]

Manual work just isn't practical.

[@NameAcht]

Let me give you this example, in BMS, charts are rated on a scale, there is the normal scale, insane scale and overjoy scale, normal scale starts at level 1 - 13, lv13 is like the equivalant of lv1 on the insane scale, insane scale then goes from lv 1 - 25. Where lv20 is the equivalant of lv1 on the overjoy scale. But there is no rating system in this game, charts were first of all chosen to be on one of these certain scales, and then their level on that scale was determined by community votes. This actually worked surprisingly well for BMS, i dont know if this could technically work for osu in some way, i guess the bms community was just so small, that the reasoning behind every vote could be considered to validate it. With osu having a massive community, that might not work that well, but i still wonder if it would be possible to refine that in a way to make community votes a thing for osu to determine difficulty of maps, just a concept of an idea i guess

[@deadfromhell1235]

You can easily stop almost of all this without ranking maps like pp compilation even tho a few problems will still exist

[@SerreVid-dt8sm]

New system when?

[@D1ttu]

I don't think overall ranking is necessary tbh. You can't compare players based different skills like aim and stamina. So different leaderboards for different skills like Osu skills.

[@iamzombie9889]

Equaius it’s my first ti using any software tho

[@minecraftermad]

Like in warframe?

[@Tenshi_on_Val]

honestly its just a issue that dioesnt need fixing, if you want to be higher rank than pp farmers, get better, you cant expect a free game to put so much work into something that doesnt matter when factoring in the cost of fixing, maybe its hiring new people, changing a program, etc.

[@ChristopherKim]

Tmw cookiezie's highest pp maps are stream maps and is still #2

[@SkterBi]

I think that we need to change the pp system with elo system and when you want to get elo or ranks you need to try to 1v1 another player for it and winner gets ranks and loser losses them or dosent get anything

[@Ryutchii_Wisteria]

and thats why osu!lazer existed

[@gdash6925]

Maybe put underweighted maps in The (now dead) approved system (give it a better name) and give it round 1.7x the pp?

[@AMo_osu]

How would a map be categorised as underweighted?

[@gdash6925]

A Mo i mean the underweighted 0 acc value slider maps( aka tech maps) or something of similar means

[@ephemera-noctourniquet]

Actually ppv2 being replaced with osu!skills seems like a good idea. You could have players that can be jack-of-all-traits, stream players, and the pp farmers will get qualified against other players with the same maps, which will put players that gain pp (or skill points) normally more of an edge in different skills instead of "this guy played harumachi clover for an hour straight with hdhrdt so now your rank plummets down to fucking -infinity"

[@alansmithee419]

The people who play these maps are not playing the game, they're just boosting their egos through some arbitrary points in a video game, which is just sad. How can someone have it in their head that the only important thing is to be as high up as possible, and not care about whether they worked to get there? It's so much more satisfying to work for your achievements.

[@JayeDidAThing]

I think it should be osu skills, but it takes the average of the numbers and that is what the ranks are based off of.

[@emanueleborghini3186]

Imma go help code some bits of ppv2

[@sosteneshion]

I started at osu because I wanted to improve my reflex to be able to play league of legends and counter strike. the first video I saw was a video of cookiezi playing rainbow dash like girls with HR, when I saw those jumps I got very hyped , because he had such a good reflex, something that I was looking for. When people start to say thing about jumping maps I kinda dont understand because that is what osu is about... you dont need a good reflex to do a stream you need stamina but still is easier than a 8 stars jumping map. Some players play a 6stars jumping map with DTHD and fc and they say that it was easy, OFC you just got good at the game but is not actually that easy. (just my opnion)

[@huanpham9024]

try a 8* jump and 8* stream map u will see the difference and reflex shouldnt be the only skill it take to get top score right?

[@sosteneshion]

yep 8 stars stream just means more stamina thats it. my point was that alot of players are like " they should nerf jumps because it rewards you with more pp" but its actually what it supposed to reward imo. its like rohulk playing Ne Obliviscaris - Painters of the Tempest VS vaxei playing Excision & Pegboard Nerds - Bring The Madness (Noisestorm Nightcore Remix). bring the madness in my point of view is much harder. maybe doing like osu!skills did, but putting it on the main system like different playstyles have their on categories