Guest
home
beatmaps
rankings
community
store
help
forum
  1. Forums
  2. osu!
  3. Development
  4. Feature Requests

Add pp as a scoring system in standard mode

posted
Total Posts
318
This is a feature request. Feature requests can be voted up by supporters.
Current Priority: +2,788
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
As many people (including peppy) know, the current scoring system is highly flawed and inaccurate. Currently, there is a working system that most people find to be extremely accurate for rating scores, and that is the ranking system, pp.

It's hard to list all the reasons why this would be better than the current score system. A better question would be: Why not? Well, let's list some advantages of this system first:

1. It would make it so that scores wouldn't get beaten by a "worse" play anymore. People won't be able to easily beat 95% DoubleTime scores with Flashlight or Hidden+Hardrock. However, people will be able to beat their old HDHR or FL scores with new DT scores (which is a problem I'm currently having with some maps, that I can't beat my old HDHR scores because my DT accuracy is too low, despite the fact that the DT scores would be worth 2x the pp!). This also means that less skilled players can set scores with HDHR and not be afraid that a low accuracy DT might not be able to beat it in the future (the only reason I haven't HDHR'd Remote Control, for example), which would actually result in more accurate rankings.

2. It would more accurately represent what most people consider to be skillful. A lot of the time, scores that are ranked 1000 with the current system would in fact be high up in pp, perhaps in the top 50 of the map. Most people don't care to see low skill Hidden or HardRock scores in the top 50 when somewhere way below there's a DoubleTime with a nearly full combo which blows all of those plays out of the water.

3. If pp were to implement the three stats (aim, speed, accuracy) which tp has implemented, it could even open up a possibility to have 4 separate scoreboards on each map: Overall, Aim, Speed, and Accuracy, and the game could save the highest score a player has set for each scoreboard. This way, all players could show what they're truly capable of for each stat; a lot of the time a 95% DoubleTime+Hidden score might be worth an extremely high amount of pp because of the speed and aim required, but isn't worth much in terms of accuracy. With this system, a player could not only play that map with DTHD, but also with HDHR to get a high accuracy score as well.

There are a few disadvantages to this system, though, those being:

1. It would jumble the ranks around enough that a good portion of many scoreboards would have plays which lack replays, at least until those plays are pushed down by better scores.

2. If pp were ever to update its calculations, it could affect (possibly significantly) the placement of scores on the leaderboards. However, as far as I know, no significant changes have been made to either tp or pp in a long time, so I don't think this would be too much of an issue.

3. Any scores which were set in the past which didn't beat their highest score with the current system, but would have with the proposed system, obviously wouldn't be counted. This could cause a lot of issues with players asking that some of their scores be placed into the leaderboard because of the fact that the only reason they currently are not is because they were set before the system was changed.


In my opinion, the pros far outweigh the cons. If anyone has anything else to add, please feel free to post!


Edit: Because people keep asking, there are two reasons why mods can't be buffed or nerfed to more accurately represent skill. The first reason is that because mods don't literally multiply your score exactly, it would be impossible to recalculate old scores. This was one of the main arguments against my previous suggestion. The second reason is because a flat bonus on map-changing mods (such as DoubleTime or HardRock) doesn't work or make sense in terms of rating how difficult a play really is. Not all maps will be 1.12x more difficult with DoubleTime; some will be less, and others much more. That is what the star rating system is for.


Edit 2: Me and likely most of the supporters of the thread would be completely fine with there being an alternate "pp" scoreboard, so long as it's possible to set this scoreboard as the one viewed by default, and so long as high scores for both "pp" leaderboards and "score" leaderboards are saved.
Last edited by Ohrami , edited 7 times in total.
Quack
64 posts
Joined July 2013
Quack
Quack.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
I forgot to mention: In the case that two scores are tied for pp, the highest score for the old system will get the higher rank. This will make sure that "spinner wars" aren't annihilated, and ensure that spinner bonus isn't made completely useless.
Piine
1,573 posts
Joined March 2012
Piine
That good ol 69 stars. Nice idea Kyou. Wished I had stars to give
Ziggo
121 posts
Joined March 2012
Ziggo
That'd be awful. I hope this will never happen.
VoidnOwO
1,256 posts
Joined August 2012
VoidnOwO

Ziggo wrote:

That'd be awful. I hope this will never happen.
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma
I dont personally like it. Maybe I dont mind it, I just like the current system better I think.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
Can you offer reasons for why you wouldn't like it? Do you think the pp system is inaccurate, or perhaps the current score system is more accurate? If so, why?
Period
33 posts
Joined July 2013
Period
I wouldn't mind this if both systems were side by side. As in, you could select

Top 50s by Score
or
Top 50s by Performance

the same way you select which gamemode you want to see top 50s in, at least on the web interface for beatmaps. In-game could just work with the dropdown like selecting country and friend ranks does.

I think being able to set a top 50 through your score opens the game up to be a lot more competitive to weaker players (Players who can HD+HR, for example, rather than DT), so it shouldn't be removed completely. If you removed top 50s by score, only the very best players would ever get a top 50s on the popular maps, so in the end, for a lot of maps, it would look very similar to the normal ranking system.

I think including both score systems would be a larger benifit. Maybe Top 50s by Performance could be the default, but I think maybe that would be too big a transition all at once. Including both seems like a better solution.

It would look pretty good on profiles too, imo:

[User] achieved Performance rank #42 on [Map] (osu!)
and
[User] achieved Score rank #14 on [Map] (osu!)
Drezi
217 posts
Joined January 2014
Drezi
I don't understand why people are against this idea. It's just a matter of disliking change I guess, being attached to legacy methods.

Cause there are no rational reasons other than the technical problems, for this not to be implemented. If pp is the most accurate measurement of skill right now, why would you only want global/country player rankings to use it, but not map rankings? Map rankings should display an order based on skill obviously, and the most accurate measurement of that is pp right now.

Too bad I don't have stars.

Also this is a little off topic, but don't you think country rankings should use the same weighting method to calculate an overall country pp based on it's player's pps, the same way our pp is calculated based on our best performances? You know the [performance above*0.95] weighting. Cause right now it's almost exclusively based on active player number.
Maxis
151 posts
Joined May 2013
Maxis

Period wrote:

I wouldn't mind this if both systems were side by side. As in, you could select

Top 50s by Score
or
Top 50s by Performance

the same way you select which gamemode you want to see top 50s in, at least on the web interface for beatmaps. In-game could just work with the dropdown like selecting country and friend ranks does.

I think being able to set a top 50 through your score opens the game up to be a lot more competitive to weaker players (Players who can HD+HR, for example, rather than DT), so it shouldn't be removed completely. If you removed top 50s by score, only the very best players would ever get a top 50s on the popular maps, so in the end, for a lot of maps, it would look very similar to the normal ranking system.

I think including both score systems would be a larger benifit. Maybe Top 50s by Performance could be the default, but I think maybe that would be too big a transition all at once. Including both seems like a better solution.

It would look pretty good on profiles too, imo:

[User] achieved Performance rank #42 on [Map] (osu!)
and
[User] achieved Score rank #14 on [Map] (osu!)
I agree with this. While I do like the idea of having scoring system based on pp, another con to this not yet mentioned is that there are thousands upon thousands of scores right now (based on, well, score) that people would be pretty mad about if they spent so much time trying to get and suddenly weren't top 50 anymore. Having both systems works best IMO, it keeps the old score system for those who prefer it more as well as a new and generally more accurate one.
jesse1412
1,457 posts
Joined January 2010
jesse1412
Sounds good. Change is progress.

Ziggo wrote:

That'd be awful. I hope this will never happen.
Hopefully you are in the minority and this does happen.

Period wrote:

I wouldn't mind this if both systems were side by side. As in, you could select

Top 50s by Score
or
Top 50s by Performance

the same way you select which gamemode you want to see top 50s in, at least on the web interface for beatmaps. In-game could just work with the dropdown like selecting country and friend ranks does.

I think being able to set a top 50 through your score opens the game up to be a lot more competitive to weaker players (Players who can HD+HR, for example, rather than DT), so it shouldn't be removed completely. If you removed top 50s by score, only the very best players would ever get a top 50s on the popular maps, so in the end, for a lot of maps, it would look very similar to the normal ranking system.

I think including both score systems would be a larger benifit. Maybe Top 50s by Performance could be the default, but I think maybe that would be too big a transition all at once. Including both seems like a better solution.

It would look pretty good on profiles too, imo:

[User] achieved Performance rank #42 on [Map] (osu!)
and
[User] achieved Score rank #14 on [Map] (osu!)
I don't see why weaker players should even stand a chance, they're weaker, if they don't improve they will stay weak and die out of the ranking. Think of it like osu! natural selection, the weak die and the strong live on.

Maxis wrote:

I agree with this. While I do like the idea of having scoring system based on pp, another con to this not yet mentioned is that there are thousands upon thousands of scores right now (based on, well, score) that people would be pretty mad about if they spent so much time trying to get and suddenly weren't top 50 anymore. Having both systems works best IMO, it keeps the old score system for those who prefer it more as well as a new and generally more accurate one.
This argument really just keeps coming back as the biggest slap in the face to things I want to happen honestly. I recall making a thread about nerfing halftime because it's just so strong on hard maps but it got shot down for the reason that people spent a lot of time trying to get their scores and so I had to find an alternative; now the same argument is back again here. If a score is weaker then I don't really care how long you spent getting it, the score is weaker and shouldn't matter in a competitive game. Effort shouldn't be rewarded, performance should be rewarded and that seems to be the direction the ranking system wanted to go with the introduction of pp, ppv2 and tp.

I'll put it really harshly and plainly: I don't care if you spent 6 years trying to FL the unforgiving marathon if someone else double times it. Your score is clearly in my eyes the most well attempted when using flashlight but it's still just pointless now that someone has a DT score. If you want to win by effort then you should doubletime the map with flashlight and prove that you're not only better but willing to try harder.

Another example because flashlight gets bad press. I don't care if you spent 8000 retries trying to halftime fc fourdimensions, you aren't even close to someone who gets a 1000 combo and your score shouldn't even be considered a thing compared to them.

One last example because halftime is another taboo mod. I don't care if you spent 60 hours playing and fcing image material with 98% acc and hd when someone else has a 95% hr fc. Your hd score ISN'T as good no matter what your actual score says.
silmarilen
5,582 posts
Joined February 2010
silmarilen
the pp in standard may not have changed much (if at all) but the other modes are still a mess. this is a good idea, but i think we better wait untill those are fixed first.
VoidnOwO
1,256 posts
Joined August 2012
VoidnOwO
:oops:
Last edited by VoidnOwO , edited 1 time in total.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
And generally rhythm games become more difficult when they get faster. I don't really see your point, and I even mentioned that players will not only be able to prove their speed, but also their accuracy in my third point in the OP. I should also mention that most people would say pp represents accuracy better; would you like a 94% Hidden score to beat a nomod SS?

Period wrote:

I think being able to set a top 50 through your score opens the game up to be a lot more competitive to weaker players (Players who can HD+HR, for example, rather than DT), so it shouldn't be removed completely. If you removed top 50s by score, only the very best players would ever get a top 50s on the popular maps, so in the end, for a lot of maps, it would look very similar to the normal ranking system.
I don't understand why you see this as a bad thing. osu! isn't a game intended to hold your hand and make things easier; if it were, we wouldn't see beatmaps like FREEDOM DiVE and Rainbow Dash Likes Girls getting ranked. If the map is really so popular that most of the top 50 players on the game have tried their best on it, I think it's reasonable to expect that they fill the leaderboards. They are, after all, the best.
Last edited by Ohrami , edited 2 times in total.
Maxis
151 posts
Joined May 2013
Maxis

jesus1412 wrote:

This argument really just keeps coming back as the biggest slap in the face to things I want to happen honestly. I recall making a thread about nerfing halftime because it's just so strong on hard maps but it got shot down for the reason that people spent a lot of time trying to get their scores and so I had to find an alternative; now the same argument is back again here. If a score is weaker then I don't really care how long you spent getting it, the score is weaker and shouldn't matter in a competitive game. Effort shouldn't be rewarded, performance should be rewarded and that seems to be the direction the ranking system wanted to go with the introduction of pp, ppv2 and tp.

I'll put it really harshly and plainly: I don't care if you spent 6 years trying to FL the unforgiving marathon if someone else double times it. Your score is clearly in my eyes the most well attempted when using flashlight but it's still just pointless now that someone has a DT score. If you want to win by effort then you should doubletime the map with flashlight and prove that you're not only better but willing to try harder.

Another example because flashlight gets bad press. I don't care if you spent 8000 retries trying to halftime fc fourdimensions, you aren't even close to someone who gets a 1000 combo and your score shouldn't even be considered a thing compared to them.

One last example because halftime is another taboo mod. I don't care if you spent 60 hours playing and fcing image material with 98% acc and hd when someone else has a 95% hr fc. Your hd score ISN'T as good no matter what your actual score says.
I won't lie, I don't disagree. It's obvious the score system right now is very flawed, e.g. like you said HD is nothing compared to HR, DT is some tough stuff whereas FL is just spam retry until you memorize, and overall the system's a mess. PP right now is a much better system by far and when it comes to actually comparing scores and plays, the amount of pp given is a much better way to go. I worry more about how the rest of the community would react to this kind of change, for example a person who got #1 on a song with HDFL topping HDHR and DT scores by spamming it for hours won't suddenly like to hear they moved to #638 or... something. I dunno, but point is, there's people who spend hours trying to get top scores, not to get #638.

I think of this from an outsider's perspective, I myself agree that performance > effort and that's why I support pp as a scoring system. If anything, the problem is just avoiding making people mad, since nobody likes change.

I've thought about what you said and that made me think of a more of a compromise of an idea: if this pp scoring system gets implemented, then when it goes live, all top 50s and replays based on the old scoring system are archived somewhere else, like on the site, ingame in a separate tab (like Global Ranking: old or something), or just somewhere where you can still see it, but you can't submit score anymore to that system. In other words, if someone comes around and SS's Big Black with HD, rrtyui will still be #1 on the old scoring system because it won't submit there (though they might of course, be #1 in pp).

Either way this kind of system may not be quite ready to be implemented, as the pp system is still improving especially on gamemodes other than standard, but I would like to see this replace our scoring system at some time in the future.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
Updated the title to specifically call for the scoring system to change in standard mode only, at least for now.
Yarissa
587 posts
Joined September 2010
Yarissa
Have some stars, I'd like to see this implemented as the main scoring metric. It would finally allow us to supercede our old best ranks on a map with better mods

Plus I think that it would be easier to find impressive replays.

My only problem with this is that it might be hard to implement.
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma
I talk for myself here and its just my opinion on it.
I don't want to be tied to the pp system that much to be honest. I think mainly because this would allow you to miss more often with less consequences (but at a higher difficulty, yes I know). For example:

I play 'x' song with no mod, get FC with 1000x combo. , and get no pp from it.
I play 'x' song with DT, get 600x combo, with 3 misses, and get 5 pp from it.

In my eyes, this would be annoying. As it would diminish a full combo, by making a non full combo potentially be much higher in rank. (I understand the difficulty would be a lot higher with dt), Its just the fact that a score wouldn't need to be as "perfect" as it needs to be now.

In addition to that, top 50 rank means nothing to be honest. Your real ranking is already measured by the pp system. The top 50 rank is more of a honoring feeling rather than a skill measurer to begin with in my opinion . (You can look at any easy difficulty top 50 to realize this)

And, again, to me , I like that not everything is tied to the pp system. Besides it is more straight forward with scores, because in the end, no one will argue that someone is first on a song if they have the highest score, whether someone could argue that the pp system is inaccurate in some instances (which I am not saying it is, however I don't think its 100% perfect and that would be seemly impossible to accomplish) , and say that 'x' song isn't that hard or blabla with 'this' or 'that' mod.

That is why I don't particularly like the idea.
Notice how I emphasize that this is My opinion , you are all free to think the exact opposite about it for all anyone cares.

Edit: And we will probably turn to the double scoring as mentioned earlier. See it here t/218559
Tom94
osu!dev
750 posts
Joined August 2012
Tom94

Kyou-kun wrote:

2. If pp were ever to update its calculations, it could affect (possibly significantly) the placement of scores on the leaderboards. However, as far as I know, no significant changes have been made to either tp or pp in a long time, so I don't think this would be too much of an issue.

3. Any scores which were set in the past which didn't beat their highest score with the current system, but would have with the proposed system, obviously wouldn't be counted. This could cause a lot of issues with players asking that some of their scores be placed into the leaderboard because of the fact that the only reason they currently are not is because they were set before the system was changed.
Those two are actually very big concerns in my opinion. Whenever the algorithm would be updated there would also get even more scores into the top50 which do not have a replay.

I feel like keeping score as a stable base measure is not such a bad thing. Especially with the possibility of mod-specific highscores (no more losing pp by beating your score with mods but worse accuracy) and optionally sorting the top50 pp-wise I don't see any reason why this wouldn't suffice.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
How often is the algorithm updated in a way which would affect beatmap leaderboards significantly? While it's a problem, I don't see it to be a significant enough one to warrant the use of such an inaccurate system.
shavit
201 posts
Joined August 2013
shavit

pl0x
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami

xasuma wrote:

I talk for myself here and its just my opinion on it.
I don't want to be tied to the pp system that much to be honest. I think mainly because this would allow you to miss more often with less consequences (but at a higher difficulty, yes I know). For example:

I play 'x' song with no mod, get FC with 1000x combo. , and get no pp from it.
I play 'x' song with DT, get 600x combo, with 3 misses, and get 5 pp from it.

In my eyes, this would be annoying. As it would diminish a full combo, by making a non full combo potentially be much higher in rank. (I understand the difficulty would be a lot higher with dt), Its just the fact that a score wouldn't need to be as "perfect" as it needs to be now.
If a 1000 combo nomod is worth less pp than a 600 combo DT, it means that DT is probably a better score, thus it should rank higher. Why should a score have to be "perfect"? I guarantee that on any map where a 600 combo DT would beat a 1000 combo nomod, the player who did the 600 combo DT could probably also do the 1000 combo nomod, and much more easily at that.

xasuma wrote:

In addition to that, top 50 rank means nothing to be honest. Your real ranking is already measured by the pp system. The top 50 rank is more of a honoring feeling rather than a skill measurer to begin with in my opinion . (You can look at any easy difficulty top 50 to realize this)

And, again, to me , I like that not everything is tied to the pp system. Besides it is more straight forward with scores, because in the end, no one will argue that someone is first on a song if they have the highest score, whether someone could argue that the pp system is inaccurate in some instances (which I am not saying it is, however I don't think its 100% perfect and that would be seemly impossible to accomplish) , and say that 'x' song isn't that hard or blabla with 'this' or 'that' mod.
Why should you be "honored" if your scores suck? Not to be harsh, but it doesn't seem reasonable. You think nobody argues about first places with the current score system? That's just dead wrong. Almost nobody considers a FL or HDHR to be the "true" #1 when it's ahead of a DT FC. People already argue that the score system is far more inaccurate than the pp system ever was.
Bauxe
1,624 posts
Joined August 2012
Bauxe
I think a lot of mid tier players would lose motivation if something like this was introduced. I don't really like the idea at all.
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma
Kyou kun, I don't particularly want to argue about this. I was just elaborating on why I didn't like the idea.

From what I see though is that if this were implemented, some people would in fact feel uneasy.
I'll keep reading to see where this heads. :)
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami

Bauxe wrote:

I think a lot of mid tier players would lose motivation if something like this was introduced. I don't really like the idea at all.
Why would you lose motivation because of a more accurate ranking system? As a "mid-tier" player myself, I would be more motivated to see my scores get high ranks, since I know they actually mean something.
Bauxe
1,624 posts
Joined August 2012
Bauxe

Kyou-kun wrote:

Bauxe wrote:

I think a lot of mid tier players would lose motivation if something like this was introduced. I don't really like the idea at all.
Why would you lose motivation because of a more accurate ranking system? As a "mid-tier" player myself, I would be more motivated to see my scores get high ranks, since I know they actually mean something.
By mid-tier, I'm not talking top few thousand. I mean the range from maybe 5k to 50k. Getting a top 50 score will seem impossible to them.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
Why should it feel possible? Further, on what maps are 5k-50k ranked players getting top 50 scores frequently with the current system? I don't know of any myself.
Drezi
217 posts
Joined January 2014
Drezi
Why should anyone hold ranks they clearly don't deserve? This ain't charity. They can aim for country ranks.
Ziggo
121 posts
Joined March 2012
Ziggo

Kyou-kun wrote:

Can you offer reasons for why you wouldn't like it? Do you think the pp system is inaccurate, or perhaps the current score system is more accurate? If so, why?
The main reason is that I think the current scoring system is more fun. Some examples: There are more ways to get a good rank besides going super fast or having godlike accuracy. Also you don't need to restart a map every time you get some 100s at the start if you want to get into the rankings.

The second point is, that changes in the pp-system would change the scoring system every time. This would result in even less changes to the system. Also, I believe a consistent scoring system is better in general.
Bauxe
1,624 posts
Joined August 2012
Bauxe

Drezi wrote:

Why should anyone hold ranks they clearly don't deserve? This ain't charity. They can aim for country ranks.
You make it sound like the same 50 people should have rank 1 to 50 on every map...
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami

Bauxe wrote:

You make it sound like the same 50 people should have rank 1 to 50 on every map...
If they set the 50 best scores on the map, then yes, they should. That's sort of the point of an accurate, skill-based ranking system.

Ziggo wrote:

The main reason is that I think the current scoring system is more fun. Some examples: There are more ways to get a good rank besides going super fast or having godlike accuracy. Also you don't need to restart a map every time you get some 100s at the start if you want to get into the rankings.
Why should there be more ways to get a good rank besides going super fast or having godlike accuracy? Those are the skills that deserve high ranks, and currently, they often get low ones instead. As for the 100s at the start not mattering as much: Do you really think that makes sense? That a lower accuracy score with the same mods and combo often beats a higher accuracy score because of that?

Ziggo wrote:

The second point is, that changes in the pp-system would change the scoring system every time. This would result in even less changes to the system. Also, I believe a consistent scoring system is better in general.
Yes, I mentioned this. Why would a consistent but flawed scoring system be better than an evolving but accurate one?
Mythras
255 posts
Joined February 2010
Mythras
this sounds unfair to flashlight

though i do hate flashlight
Drezi
217 posts
Joined January 2014
Drezi

Bauxe wrote:

You make it sound like the same 50 people should have rank 1 to 50 on every map...
Well, if every top50 player achieved the best they are capable of on every map, why should others take their place? It's not like top players care about hards and below anyway as far as I know.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami

VioletMaid wrote:

this sounds unfair to flashlight

though i do hate flashlight
Flashlight already gives bonus pp, and not an insignificant amount.
haha5957
1,433 posts
Joined July 2010
haha5957
Even the simplest (new) scoring method that might be added is still going to be better than the current one.




I mean...... I also do think that current pp system is not exactly perfect and might change(which sucks, since your pp might drop randomly) in future.
However, does that mean current best scoring system makes more sense ans should be left unchanged?

A score system that doesnt even care the first 10 100s and that does care last one 100?
Score system that takes 120pp right away from you just because you decided to HD on a previously-DT-best scored-map?

hello?


be a person who knows how to think about how it can be better than before, not a crybaby who says NO all the time.

There is no such ing as perfect, but there is a thing called "better"
Lanturn
osu! Alumni
3,912 posts
Joined March 2012
Lanturn
It'd be interesting. Getting all 300's to the end of the song, you have only 4-5 notes left, You are placed 10th on the map, then completely f up the final stream. You'll end up dropping from 10th to 500th. If we take this route with the scoring system, this is exactly what's going to happen. Scores won't be just scaling up, they will be increasing / decreasing as you play through the song.

An issue that I think would have problems is the lower diffs where there is literally only 1pp or a few extra points in Aim/Acc/Spd difference between each mod. That means anyone who can spin would only have to run two mods (DT HR) and they'd have a first even on medium length spinners.
At this point, having an option to sort by PP earned would be better than having the scoreboard adjust since it wouldn't be the top PP players in first. The current system also has this problem, but I see it being even worse with the new system since mods no longer would give a set value, and would probably make the multipliers within only a few decimal places of each other because of how close they are in performance.
It depends on how it is calculated as I'm no mathematician to be able to even comprehend how it might be calculated, but the difference between mods on easy songs, and mods on hard songs is extremely big. We have to make sure all ranges of difficulties can be calculated properly.

Just some things I thought of while reading through this thread for a few days. I'm not against this, nor am I supporting it. I'm just going to stay neutral in this whole thing.
PP is out of my reach with this new system since it doesn't work to my style of playing so I've honestly just given up on it. I've never really been competitive anyways, no do I care really about improving my speed skills. I'd rather improve at something I have fun doing (Flashlight and 100%ing slow songs no one has before.)


Is this a good idea? Yes, very much so. However, I still think there are lots of problems that need to be taken care before we can safely transfer over to it. Many of which have already been listed. Right now I think it is way too early still to be pushing this feature as hard as this thread seems to be wanting to. The system isn't even closed to being finished with the other modes, Seems like there is still some balancing issues to deal with, etc. If there is going to be such a big change as switching over to a different method of scoring, everything has to be completed and ready to go, doing bits and pieces at a time would make it a disaster, especially with how delicate the system is with reasons you (kyou) and tom have mentioned.
Ziggo
121 posts
Joined March 2012
Ziggo

Kyou-kun wrote:

Why should there be more ways to get a good rank besides going super fast or having godlike accuracy? Those are the skills that deserve high ranks, and currently, they often get low ones instead.
Well, guess we have different views about skill then. I think ranks like top HD or FL scores are well desevered.

Kyou-kun wrote:

As for the 100s at the start not mattering as much: Do you really think that makes sense? That a lower accuracy score with the same mods and combo often beats a higher accuracy score because of that?
Yes, I do. Because you can easily retry over 100s at the start, so they should be way less important than 100s later on.
dennischan
454 posts
Joined June 2013
dennischan
pp is far too inaccurate for anything like this.
it only considers max combo and acc, which are not sufficent for any case of score calculation.
there is no pattern recognization or anything that could determine the skill of a player accurately.

I think you are getting to hyped up over a system which is not operational.
Before per hitobject data can be provided,
it is simply nonsense to base scores on it.

maybe this should be implemented, but in the far future after pp is fully functional and gives accurate scoring.
then we can talk about this.

ps:FL is much harder than DT. DT is just high speed spamming. (excluding sayo's plays)
Drezi
217 posts
Joined January 2014
Drezi

dennischan wrote:

ps:FL is much harder than DT. DT is just high speed spamming. (excluding sayo's plays)
Not sure if serious.
Fudgy
47 posts
Joined December 2013
Fudgy
I really hope this will be the case for osu!mania in the future. For now, using DT is useless because it's extremely hard to beat a no-mod score when there is a score limit and DT does not boost score.

DT scores are harder to achieve and should be on top of the rankings and also boost performance.
jesse1412
1,457 posts
Joined January 2010
jesse1412

Ziggo wrote:

Kyou-kun wrote:

Can you offer reasons for why you wouldn't like it? Do you think the pp system is inaccurate, or perhaps the current score system is more accurate? If so, why?
The main reason is that I think the current scoring system is more fun. Some examples: There are more ways to get a good rank besides going super fast or having godlike accuracy. Also you don't need to restart a map every time you get some 100s at the start if you want to get into the rankings.

The second point is, that changes in the pp-system would change the scoring system every time. This would result in even less changes to the system. Also, I believe a consistent scoring system is better in general.
If you're not going super fast or producing scores with godlike accuracy/aim your score isn't good (relatively) and hence should not be rewarded with a spot in the leader boards.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami

dennischan wrote:

pp is far too inaccurate for anything like this.
it only considers max combo and acc, which are not sufficent for any case of score calculation.
there is no pattern recognization or anything that could determine the skill of a player accurately.

I think you are getting to hyped up over a system which is not operational.
Before per hitobject data can be provided,
it is simply nonsense to base scores on it.

maybe this should be implemented, but in the far future after pp is fully functional and gives accurate scoring.
then we can talk about this.

ps:FL is much harder than DT. DT is just high speed spamming. (excluding sayo's plays)
It sounds to me like you don't understand how pp works. It does consider every hit object in the map, and calculates the beatmap's difficulty based on the average and maximum strain of your aim, speed, and accuracy. Sometimes this can cause inaccurate calculations, but those are pretty much always only inaccurate for the overall ranking system, and don't really affect individual leaderboards. Yes, Koigokoro is overrated in terms of pp, but each individual score on the map can be placed very accurately. In fact, I think pp works better as a scoring system than as a ranking system.

As for saying "DT is just high speed spamming", that's your inner rank 30,000 speaking.

Ziggo wrote:

Well, guess we have different views about skill then. I think ranks like top HD or FL scores are well desevered.
What about getting 95% FC with Flashlight is more skillful than 90% FC (or even less than FC) with DT on FREEDOM DiVE? Do you really think those two scores are even comparable? One is doable by a few players if they put enough time into the game; the other is impossible for any player at the current time, no matter how much time they put into the game. pp already considers how much DT increases difficulty by recalculating, so DT gives larger bonuses to difficult maps (as it should), where Flashlight's bonus is static (as it should be).

Ziggo wrote:

Kyou-kun wrote:

As for the 100s at the start not mattering as much: Do you really think that makes sense? That a lower accuracy score with the same mods and combo often beats a higher accuracy score because of that?
Yes, I do. Because you can easily retry over 100s at the start, so they should be way less important than 100s later on.
So because those 100s are easier to fix, you think it shouldn't matter as much? What's the reasoning behind that? All the hit circles have the same judgment, and ignoring variations in speed, it's equally easy to get a 100 on all of them.
Kaeru
302 posts
Joined February 2013
Kaeru
I wouldn't mind being able to sort plays by pp gains, but I don't want performance points to become the main scoring system. We already have enough people spamming DT on everything and this would just encourage that kind of play style further rather than teaching people to become accurate. If you can get a high accuracy with DT on a difficult song then you deserve to hold the top 50 on that map. I don't think it's fair to completely throw the other mods out the window. Ether go for pure pp gains or try to achieve high scores on maps, do you really need both at the same time?

In short, I don't really care if a score is deemed more impressive or not by the pp system. These biases will always exist, and removing score will just make this into an even bigger problem. Not to mention ruining all the current scores and making everything into a big mess. I think everything works fine how it is now. Being able to sort scores by pp is a much better idea in my opinion (and there are already a few topics that suggest that).

I agree that the current system is far from perfect, but I don't think the pp system is meant to be the main scoring system.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
Speed and aim are often better than accuracy. Accuracy is often better than speed and aim. pp calculates this and decides which score involved the most skill. I don't see how it's encouraging players to be less accurate. thelewa is known for having great accuracy, where his other skills aren't really comparable, and he's rank #5 overall by the pp system. How would this be possible if pp encouraged speed and aim but not accuracy?

The fact is that speed and aim are vastly underrated with the current system, whereas pp is balanced.
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko
DT best mod
make it give all the pp
and 1.13x thx
tu much hd hr pushing dt plays off the scoreboard when not as impressiv

better alternative to pp scoring
Amianki
85 posts
Joined January 2011
Amianki
I like this idea enough to match your stars.

Hope this gets included simply because it pissed me off how I turned an FC-5 into an FC but still lost about 10pp once because my accuracy was slightly lower.
Fudgy
47 posts
Joined December 2013
Fudgy
Can someone explain to me why the overall pp-based ranking is accepted and considered accurate by the majority of players but at the same time, the score-based map rankings are still considered to be the appropriate way of sorting performances on each map?

How can someone think that global rankings and map rankings should use a different method? How does that even make sense?

How come the overall rankings do not use score, if it's supposedly more "honorific" and "accurate" than pp on each individual map?

Is there something obvious I overlooked which makes the situation more complicated that I made it seem?
Kodehawa_old
Kodehawa_old

Ziggo wrote:

That'd be awful. I hope this will never happen.
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma

CalignoBot wrote:

I like this idea enough to match your stars.

Hope this gets included simply because it pissed me off how I turned an FC-5 into an FC but still lost about 10pp once because my accuracy was slightly lower.
That is not the point. I think that will be implemented regardless soon-ish (I linked it in an earlier page). So you don't have to lose pp for a 'better' score anymore. But that is not what the discussion is about here. :p

Im gonna remain neutral for the rest of this thread I think :B
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko

Fudgyking wrote:

Can someone explain to me why the overall pp-based ranking is accepted and considered accurate by the majority of players but at the same time, the score-based map rankings are still considered to be the appropriate way of sorting performances on each map?

How can someone think that global rankings and map rankings should use a different method? How does that even make sense?

How come the overall rankings do not use score, if it's supposedly more "honorific" and "accurate" than pp on each individual map?

Is there something obvious I overlooked which makes the situation more complicated that I made it seem?
Because Score is, in effect, a "perfect" way to compare scores and more precise [maps can get up to 1.5bil score, pp doesnt really go above 500. less room for error on the scoreboard this way.] since otherwise there will be like 5 sets of the same amount of pp on the scoreboard for your average full dthd/hdhr scoreboard.

Also there's no obfuscation. People will know exactly how they got on the scoreboards because the way score is calculated is perfectly common knowledge and makes for absolute accuracy and is better for comparison purposes.

And it's always been that way.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
But it's not perfect. It's flawed. You also don't seem to have read the third post in this thread.

On an insane map, 1x100 makes a difference with pp, so it's as precise as it needs to be.
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko

Kyou-kun wrote:

But it's not perfect. It's flawed. You also don't seem to have read the third post in this thread.

On an insane map, 1x100 makes a difference with pp, so it's as precise as it needs to be.

lets take into account you cant.. and will never see your PP on the play in game while playing... for many reasons such as it just feeling crappy mechanically, i.e the first 10% of the map or so you wont see any pay off for hitting those circles at all, and it's not every single note that will increase this 'in game pp score' so the gameplay feels disconnected from scoring, especially considering most plays arent going to be worth much pp anyway compared to the players "best" plays so it will just be demoralizing in game. And because the way PP is calculated.. and because unranked maps do exist.. etc..

That aside, if we keep the way the game works right now and have score used in game [which we will], just not for anything but measuring your individual circle clicks, you're playing the map and there's this arbitrary number on the top right called "Score" while on the left the scoreboard just has combo and PP. That kind of disconnect doesn't really promote competition but just obfuscates it. Why even show your play underneath the scoreboard in-game if it's just going to show your score/combo compared to their pp/combo, how would you even climb onto the scoreboard whilst playing since PP isnt even calculated til the end?

And on easy-normals its going to be rather silly to just see something like "38 pp" repeated 50 times on 2 different maps that might have a very different max combo like 300 vs 1000 but they were worth the same PP and everyone SS'd it so whose to say what's different about playing them? :roll:

If anything, just get rid of HD HR getting a decimal advantage over DT in score multiplier, and get rid of Half Time scores showing up entirely, and scoreboards are perfect the way they are imo. There are even less reasons why Peppy would disagree with that proposal but even so there's still a 0% chance of it happening because he hates touching the scoreboards.

On a final note, Peppy severed the association between PP and actual rank on maps with ppv2 came out, i dont think he'd link them again.

p.s wouldnt mind being able to see scores by PP rank with a filter like friends/country etc as a supporter though. As Global its just too much.
Drezi
217 posts
Joined January 2014
Drezi

Bassist Vinyl wrote:

lets take into account you cant.. and will never see your PP on the play in game while playing... for many reasons such as it just feeling crappy mechanically, i.e the first 10% of the map or so you wont see any pay off for hitting those circles at all, and it's not every single note that will increase this 'in game pp score' so the gameplay feels disconnected from scoring, especially considering most plays arent going to be worth much pp anyway compared to the players "best" plays so it will just be demoralizing in game.
Who looks at their score while playing?? It's just a meaningless large number.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
Where do I start...

Bassist Vinyl wrote:

lets take into account you cant.. and will never see your PP on the play in game while playing... for many reasons such as it just feeling crappy mechanically, i.e the first 10% of the map or so you wont see any pay off for hitting those circles at all, and it's not every single note that will increase this 'in game pp score' so the gameplay feels disconnected from scoring, especially considering most plays arent going to be worth much pp anyway compared to the players "best" plays so it will just be demoralizing in game. And because the way PP is calculated.. and because unranked maps do exist.. etc..
1. There are plenty of ways that pp could be shown in-game while playing. I think the most reasonable way to implement this would be to have a "descending" score method; that is, you start with the max pp possible for the map, and it goes down as you make mistakes. I already proposed this for accuracy a long time ago in this forum, but it never got anywhere.

2. Unranked maps don't affect this at all.

3. Why does it matter if every note gives a pp or not? You don't pay attention to your score the whole play, I'm sure. The only time your score matters is at the end, which is why what I proposed in point "1" of this post seems reasonable.

Bassist Vinyl wrote:

That aside, if we keep the way the game works right now and have score used in game [which we will], just not for anything but measuring your individual circle clicks, you're playing the map and there's this arbitrary number on the top right called "Score" while on the left the scoreboard just has combo and PP. That kind of disconnect doesn't really promote competition but just obfuscates it. Why even show your play underneath the scoreboard in-game if it's just going to show your score/combo compared to their pp/combo, how would you even climb onto the scoreboard whilst playing since PP isnt even calculated til the end?
4. Why would the number at the top right have to continue to exist in the game? In my opinion, it just adds clutter. Even if score were to remain the scoring method, it would still be reasonable if that number were removed.

5. A method to calculate pp dynamically, mid-play can be created.

Bassist Vinyl wrote:

And on easy-normals its going to be rather silly to just see something like "38 pp" repeated 50 times on 2 different maps that might have a very different max combo like 300 vs 1000 but they were worth the same PP and everyone SS'd it so whose to say what's different about playing them? :roll:
5. I don't get what you mean by the "different max combo like 300 vs 1000", as a combo difference that drastic would clearly result in a difference in pp. I already told you that the highest score will get the higher rank in situations where pp is equal.

Bassist Vinyl wrote:

If anything, just get rid of HD HR getting a decimal advantage over DT in score multiplier, and get rid of Half Time scores showing up entirely, and scoreboards are perfect the way they are imo.
6. Even with DT being worth a little bit more, it's still imbalanced in my opinion. You don't even need close to a full combo on some maps to get really high pp on them with DT because of how the pp system recalculates map difficulty based on mods. DT doesn't affect slow maps as much as it does fast ones, so a flat bonus could never work well for a scoring system.

Bassist Vinyl wrote:

On a final note, Peppy severed the association between PP and actual rank on maps with ppv2 came out, i dont think he'd link them again.
7. He severed the association between pp and score rank in specific, because he recognized how little score matters when considering actual skill. If the leaderboards did represent actual skill, I don't think there would be any problem with connecting them to the ranking system.]
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko

Kyou-kun wrote:

5. I don't get what you mean by the "different max combo like 300 vs 1000", as a combo difference that drastic would clearly result in a difference in pp. I already told you that the highest score will get the higher rank in situations where pp is equal.
well except it wouldnt unless they had the exact same star rating.
Ohrami
685 posts
Joined March 2010
Topic Starter
Ohrami
So you're talking about different maps? I still don't get your point. Why would it matter if there were multiple different maps where several scores are worth 38 pp?
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko
and what about calculating pp dynamically while playing? you're rarely ever going to be coming close to your best performances so it's just going to be demoralizing unlike Score. It just doesn't seem like it will play well and will rarely be useful while playing. Maybe for replays.

Kyou-kun wrote:

So you're talking about different maps? I still don't get your point. Why would it matter if there were multiple different maps where several scores are worth 38 pp?
diff. max combos gave diff. score which gave a better idea of the map, showing only pp says nothing about the map really.
Drezi
217 posts
Joined January 2014
Drezi
Well, max combo could be easily displayed in addition to map length, number of hit circles, sliders, etc. Actually it should be either way, cause I don't like having to check global rankings (or in editor) to see it.
CodeS
1,142 posts
Joined September 2011
CodeS

Ziggo wrote:

That'd be awful. I hope this will never happen.
[-Cloud-]
438 posts
Joined June 2013
[-Cloud-]

Drezi wrote:

I don't understand why people are against this idea. It's just a matter of disliking change I guess, being attached to legacy methods.
All those FL scores dissapearing out of the Ranking. I think a lot of tears would fall.
Personally, I love this Idea, since I think HT and FL ist just lame. Killer Song is probably the best example for this.
haha5957
1,433 posts
Joined July 2010
haha5957
If you can't accept the fact that pp is accurate, at least compared to score system, you are dumb. I can confidently say that.

Bassist Vinyl wrote:

Fudgyking wrote:

Can someone explain to me why the overall pp-based ranking is accepted and considered accurate by the majority of players but at the same time, the score-based map rankings are still considered to be the appropriate way of sorting performances on each map?

How can someone think that global rankings and map rankings should use a different method? How does that even make sense?

How come the overall rankings do not use score, if it's supposedly more "honorific" and "accurate" than pp on each individual map?

Is there something obvious I overlooked which makes the situation more complicated that I made it seem?
Because Score is, in effect, a "perfect" way to compare scores and more precise [maps can get up to 1.5bil score, pp doesnt really go above 500. less room for error on the scoreboard this way.] since otherwise there will be like 5 sets of the same amount of pp on the scoreboard for your average full dthd/hdhr scoreboard.

Also there's no obfuscation. People will know exactly how they got on the scoreboards because the way score is calculated is perfectly common knowledge and makes for absolute accuracy and is better for comparison purposes.

And it's always been that way.
How is having a same rank or point a problem when they performed the same?
Seems like
(score of) 10 100s on very begining of the map >>> 1 100 on the last note
makes sense for you. There is no word that I can properly describe you but I'm pretty much confident that you don't make sense at all

the scoreboard "looks" like they performed different because this score system has a huge flaw, making the first 10 100s virtually useless and the last 100 really matter much. If you think this score system is perfect for comparison, you got a problem there.

Even myself could think of how you can change this score system into pp-related, and friendly to users. I'm pretty sure Tom or peppy already has similar or better idea of "how pp can be directly converted to score". Please don't say we'll have tons of "#1 - 149pp" thinks on our scoreboard. they are going to be scaled to big number for sure when it comes to score. Although there might be lots of same score since there actually are many 100~99% accs with same mod and same number of 100-50-x in top #50 records, this is how it is supposed to be (and there is another thing called spinner)

I see literally no valid reason on why pp(or at least that reflects pp) score shouldn't be implemented in this entire post. Why? current scores? Cause it always has been? Yeah, of course the slavery and kingdom cuz it always has been for people in history? lmao.

Also the old scores can be converted pretty easily for sure if they decide to replace current retarded score system into pp related one. (only troublesome stuff I see is spinner)
Amianki
85 posts
Joined January 2011
Amianki

xasuma wrote:

CalignoBot wrote:

I like this idea enough to match your stars.

Hope this gets included simply because it pissed me off how I turned an FC-5 into an FC but still lost about 10pp once because my accuracy was slightly lower.
That is not the point. I think that will be implemented regardless soon-ish (I linked it in an earlier page). So you don't have to lose pp for a 'better' score anymore. But that is not what the discussion is about here. :p

Im gonna remain neutral for the rest of this thread I think :B
Where did I indicate that I was responding to anything in this thread outside the overall idea of it?
Lanturn
osu! Alumni
3,912 posts
Joined March 2012
Lanturn
So something like this?

Tommy Yang
50 posts
Joined October 2011
Tommy Yang
Posting on behalf of Kyou-kun:
"That's what I think would be best, but I don't think the replays save the spinner bonus, so there'd have to be some other way to show the score instead, perhaps only for plays where some other player is tied."
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma
If this were implemented HD mod would probably cease to be used for a lot of people. As of now hd is used in conjuction with other mods to give a score boost , however the pp gain from it is close to none (might even be none idk).
This could be argued with HR as well, although not as much as hd. Same for FL. And it would in a way monopolize DT (being the only mod that truly boosts your pp gain by a lot).

Having said that, the balance between accuracy, aim and speed would be broken. Favoring speed over the other two.

Screw staying neutral in this thread :D
Drezi
217 posts
Joined January 2014
Drezi

xasuma wrote:

Having said that, the balance between accuracy, aim and speed would be broken. Favoring speed over the other two.
So you're saying the current global player ranking system is broken? Well, both HD and HR are very rewarding if you can play them with comparable acc, maybe the acc slope is a bit too steep but that is a different issue to debate, not one connected to this idea.
Vuelo Eluko
5,484 posts
Joined October 2013
Vuelo Eluko
:cry:

Drezi wrote:

xasuma wrote:

Having said that, the balance between accuracy, aim and speed would be broken. Favoring speed over the other two.
So you're saying the current global player ranking system is broken? Well, both HD and HR are very rewarding if you can play them with comparable acc, maybe the acc slope is a bit too steep but that is a different issue to debate, not one connected to this idea.
HD HR isn't even in the same plane of existence as DT in terms of PP. xasuma is absolutely right.



the HD HR play has higher acc and yet is not even able to come to within half of the PP gain on the other play.

Anyone able to get at least a half combo on a song with DT is going to completely overshadow all HD HR full combos. I don't personally have a problem with the game basically becoming a DT only playground, but some [many] will.

In fact, I'd go so far as to say you'll be FORCIBLY YANKING rank 1's and spot on top 50 scoreboards away from players on like 90%+ of maps.

What it boils down to, does jesus deserve rank 1 on cataclysm for that DT FC? Absolutely. Do people who mash their way through tv size maps with DT and happen to get a decent combo deserve to completely replace people with HD HR 99%+ fcs? I'd say not...
Last edited by Vuelo Eluko , edited 3 times in total.
xasuma
The First Completionist
1,011 posts
Joined August 2013
xasuma

Drezi wrote:

xasuma wrote:

Having said that, the balance between accuracy, aim and speed would be broken. Favoring speed over the other two.
So you're saying the current global player ranking system is broken? Well, both HD and HR are very rewarding if you can play them with comparable acc, maybe the acc slope is a bit too steep but that is a different issue to debate, not one connected to this idea.
No, you are actually wrong. I have tested them. HD doesn't bother me , therefore I can achieve a lot of scores with the same accuracy as no mod. And the pp gain is none. OK maybe 0.xx but that could be something else for all we know .

And the pp gain of hr exists yes, but that is a steep curve to keep your acc. And even when you do, it's nothing compared to DT.
show more
Please sign in to reply.

New reply

0
1 / 318