Where are the formulas though?
Throwing an example list into this forum doesn't quite work for me.
Throwing an example list into this forum doesn't quite work for me.
forum
Best performances in the game
posted
Total Posts
70
Topic Starter
How much pp would you say is appropriate for Scarlet Rose SS? In my system it's 330pp, I'm not sure if this is too much or not.Gigo wrote:
Some scores, which I think are underrated as all hell:
https://osu.ppy.sh/s/13244
https://osu.ppy.sh/b/72585
https://osu.ppy.sh/b/66246
It needs a higher OD first.
330pp is definitely too much for a 5.2 star OD7 map.
Sigh why do people only look at star rating nowadays...Mahogany wrote:
330pp is definitely too much for a 5.2 star OD7 map.
The map itself is ridiculously hard to FC, regardless of its star rating.
In my non-expert opinion, I think it should definitely be in the upper 200's, maybe even the lower 300's.Ziggo wrote:
How much pp would you say is appropriate for Scarlet Rose SS? In my system it's 330pp, I'm not sure if this is too much or not.
Why don't you try to FC it then? You have quite a few scores in the 200's, so if you think Scarlet Rose deserves its 178 pp, then you should definitely be able to FC it. And since it's only OD7, you should get good accuracy too.Mahogany wrote:
~5*, OD7, not a particularly high amount of hit objects...sounds about right imho.
The difficulty to FC it doesn't come from accuracy, but because it's hard to read. In my opinion, the scoreboard of a map should be the best indicator of how much points a map should be worth. If you look at Scarlet Rose, the top 50 consists of quite a few non-FC's. So it must be quite difficult, right? After that, if we look at Daidai Genome, we see that it's OD8. But then there are quite a few of DT FC's on the scoreboard... and all of them are in the 99-100% accuracy. And yet it gives 370 pp for an HD+DT FC. Accuracy plays a bigger role here than it should.
But since the algorithm can't take into account how difficult a map is to read, I don't know how this can be fixed.
Then the star rating system needs an overhaul firstGoldenWolf wrote:
Sigh why do people only look at star rating nowadays...
The map itself is ridiculously hard to FC, regardless of its star rating.
My skillset isn't appropriate to FC Scarlet Rose.Gigo wrote:
Why don't you try to FC it then?
With the current PP system, Scarlet Rose's current PP reward makes sense. It's OD7, so it gives little acc PP. It has a lot of breaks, so fewer hitobjects for the marathon bonus, and IIRC there are a lot of sliders, which are underweighted.
Yep, it makes sense to me why Scarlet Rose is worth what it is.
Can you elaborate a little on what exactly your skillset is and why isn't it appropriate for Scarlet Rose? After all, the map is worth 54 pp less than your current top performance, so it should be quite easy for you to SS it.
Because, depending on how you look at it, in Scarlet Rose, there are either very fast singletaps or slow and very spaced streams. Neither of which I am any good at.
My skillset is based around accuracy and the 170-180BPM range.
My top rank does not have very fast singletaps or very spaced streams.
That's like saying Index_ should be able to HDDT -7- seven because it's worth 50pp less than his top rank
My skillset is based around accuracy and the 170-180BPM range.
My top rank does not have very fast singletaps or very spaced streams.
That's like saying Index_ should be able to HDDT -7- seven because it's worth 50pp less than his top rank
330 pp is too much for an OD7 map that's not like 5000x combo and based heavily on sliders.GoldenWolf wrote:
Sigh why do people only look at star rating nowadays...Mahogany wrote:
330pp is definitely too much for a 5.2 star OD7 map.
The map itself is ridiculously hard to FC, regardless of its star rating.
Give sliders more acc pp.
Another example of an underrated map:
I wonder if people are gonna find an excuse for this one and try to justify such low pp for an SS...
I wonder if people are gonna find an excuse for this one and try to justify such low pp for an SS...
Holy shit this map is only worth 310pp? Wtf...
Star rating does not determine difficulty. Not even physical difficulty.Mahogany wrote:
~5*
Sliders are under rated in aim and star rating, along with adding uncompensated reading difficulty and not giving any acc pp. Scarlet Rose SS should definitely give 330 pp.
Endaris wrote:
Where are the formulas though?
Throwing an example list into this forum doesn't quite work for me.
Topic Starter
Sorry, but it's not that easy. I'm using current pp calculation (star rating, etc...) combined with lots of statistical data from every score in the game.Endaris wrote:
Endaris wrote:
Where are the formulas though?
Throwing an example list into this forum doesn't quite work for me.
If you are interested in the basic idea I can tell you this: I'm basically trying to add the component of map complexity into the pp system. However map complexity in itself is a pretty vague term as that can refer to just about anything. For that reason I'm using the wide spread of statstical data to cover for as much possibilites as possible. As there are many parameters in my calculations, it's likely that certain types of maps might be op pp-wise. For that reason I'm asking for opinions on the results so I can refine the parameters more.
It's hard to give decent feedback though if I don't which parameters you use and how much influence they have.
Especially considering that we're talking about very hard maps most people here aren't even close to playing properly(I can't get a grasp at least).
I'd prefer to see a list of maps that are easier and popular so I have an easier time relating. Talking about current pp-maps like my topranks or some 08/15 Sword Art Online or Attack On Titan theme. A decently popular map like https://osu.ppy.sh/b/315614&m=0 and hard to read maps like https://osu.ppy.sh/b/191580 https://osu.ppy.sh/b/59130 https://osu.ppy.sh/b/54708 would also be very interesting to get a look at, same for EZ-mod-plays.
Especially considering that we're talking about very hard maps most people here aren't even close to playing properly(I can't get a grasp at least).
I'd prefer to see a list of maps that are easier and popular so I have an easier time relating. Talking about current pp-maps like my topranks or some 08/15 Sword Art Online or Attack On Titan theme. A decently popular map like https://osu.ppy.sh/b/315614&m=0 and hard to read maps like https://osu.ppy.sh/b/191580 https://osu.ppy.sh/b/59130 https://osu.ppy.sh/b/54708 would also be very interesting to get a look at, same for EZ-mod-plays.
E m i
If Sliders eould be considered in SR, Many maps get a huge jump (What will StrangeProgram, Lesjuh's tag be then O.o)
Topic Starter
I can see your point. The main reason I didn't post results for lower ranks as well was that I only have access to the top 50 scores of each map. A lot of scores would be missing that way.Endaris wrote:
It's hard to give decent feedback though if I don't which parameters you use and how much influence they have.
Especially considering that we're talking about very hard maps most people here aren't even close to playing properly(I can't get a grasp at least).
I'd prefer to see a list of maps that are easier and popular so I have an easier time relating. Talking about current pp-maps like my topranks or some 08/15 Sword Art Online or Attack On Titan theme. A decently popular map like https://osu.ppy.sh/b/315614&m=0 and hard to read maps like https://osu.ppy.sh/b/191580 https://osu.ppy.sh/b/59130 https://osu.ppy.sh/b/54708 would also be very interesting to get a look at, same for EZ-mod-plays.
I can check the top scores on the maps you posted though. I don't have access to my pc for the weekend but I will post them on Monday.
Yeah, maps with low fc counts get a small bonus. I'll still need to consider different mods more though. When I started out the bonus was way higher but this map https://osu.ppy.sh/s/28158 got ridiculous amounts of pp so I had to lower it.Momiji wrote:
scarlet rose 387, forgotten 631 B-]
Have you magnified the pp based on relative score from leaderboards or the amount of FCs regardless of mods used?? Seems to be it, seeing as maps like Miraizu that jump from 373 to 596 to 665 pp have been slightly nerfed.
I think pp should be as objective as possible and having bonuses for maps "that don't yet have a lot of fcs" just hinders this because then pp becomes a bit more dynamic with the status of the playing community.
Maybe objective isn't the best word and I mean to say "More static and not based around currently what people are doing/can do and/or the perception of what's worth more or less"
Maybe objective isn't the best word and I mean to say "More static and not based around currently what people are doing/can do and/or the perception of what's worth more or less"
Topic Starter
But isn't that how pp has been all along? I remember ar 10.3 giving bonus to pp for a long time, but because more and more people learned to play it Tom decided to remove that bonus.Khelly wrote:
I think pp should be as objective as possible and having bonuses for maps "that don't yet have a lot of fcs" just hinders this because then pp becomes a bit more dynamic with the status of the playing community.
Maybe objective isn't the best word and I mean to say "More static and not based around currently what people are doing/can do and/or the perception of what's worth more or less"
Ok, I am back with yet another set of, what I think are, underrated maps:
Rainbow Dash should definitely give more for an SS than it currently does (419pp). Yes, 419 may seem like a lot of points (and they are), but I think most of us can agree that these points are not enough for the difficulty of this map. It's really hard to read, has insane jumps and OD8. In my opinion, it should be in the upper 400s (480-500).
Tout Petit Moineau should also give more pp for an SS (293 currently). It's a very intense map with short and medium streams, high object density and OD 8. In my opinion - somewhere in the upper 300s (360-380).
Unpleasant Sonata currently gives 288 for an SS. It is the same type of map as Tout Petit Moineau, so the same reasoning applies here, as well as the same proposed pp gain (360-380).
He Has No Mittens should give more than 253 pp for an SS. It's true that the maps is short and it only has one hard part, but that one part is REALLY hard. I'm thinking somewhere around 340pp would be satisfactory.
Chipscape is another candidate for an underrated map with only 388pp for an SS. Intense and streamy map, requires some outstanding stamina and is almost 4 minutes long. I'm thinking it should give around 450-470pp.
This is only my opinion, don't jump on me (unless you're a cute girl) if you don't share the same view on the matter or if you think my proposed pp gains are unrealistically inflated.
Rainbow Dash should definitely give more for an SS than it currently does (419pp). Yes, 419 may seem like a lot of points (and they are), but I think most of us can agree that these points are not enough for the difficulty of this map. It's really hard to read, has insane jumps and OD8. In my opinion, it should be in the upper 400s (480-500).
Tout Petit Moineau should also give more pp for an SS (293 currently). It's a very intense map with short and medium streams, high object density and OD 8. In my opinion - somewhere in the upper 300s (360-380).
Unpleasant Sonata currently gives 288 for an SS. It is the same type of map as Tout Petit Moineau, so the same reasoning applies here, as well as the same proposed pp gain (360-380).
He Has No Mittens should give more than 253 pp for an SS. It's true that the maps is short and it only has one hard part, but that one part is REALLY hard. I'm thinking somewhere around 340pp would be satisfactory.
Chipscape is another candidate for an underrated map with only 388pp for an SS. Intense and streamy map, requires some outstanding stamina and is almost 4 minutes long. I'm thinking it should give around 450-470pp.
This is only my opinion, don't jump on me (unless you're a cute girl) if you don't share the same view on the matter or if you think my proposed pp gains are unrealistically inflated.
E m i
rrtyui andGigo wrote:
Rainbow Dash should be in the upper 400s (480-500).
a few people could probably 99.8% it but they can't even FC it, it should give 580 imo B-]Gigo wrote:
Chipscape should give around 450-470pp.
Gigo wrote:
don't jump on me
Ora ora ora ora is the new mittens, sort of
Uh, yes it does? The system might have a few failings, but it very, very much does determine difficulty.B1rd wrote:
Star rating does not determine difficulty. Not even physical difficulty.Mahogany wrote:
~5*
If we're arguing that Scarlet Rose is far tougher than 5 stars, then that's a problem with the star rating system, not with the PP system.
stars determine difficulty to FC, not necessarily physical difficulty. i.e an 8 star map with a tiny section of like 10 notes that gives it all its difficulty to fc but can be bullshitted through to survive, vs a map that is 8 stars because of consistent difficulty, it's not even close which is actually physically harder.
But the 8 stars bit is still 8 stars. You can't rate it less because there's a bit that actually is 8 stars worth of difficulty and anything less would be incorrect imho
Yes, the system has problems, and there will always be exceptions to a rule, but I don't think there are enough examples of this happening just yet. I'm of the opinion that star rating calculation is alright right now.
Yes, the system has problems, and there will always be exceptions to a rule, but I don't think there are enough examples of this happening just yet. I'm of the opinion that star rating calculation is alright right now.
you can't rate it less in 'difficulty to fc' or whatever, but you can absolutely rate it lower in terms of raw physical difficulty like how hard it is to pass. Stamina, consistency, and reading all come less into play.
I was under the impression that Star Rating is meant to be difficulty to FC?
it is
but difficulty/physical difficulty is a much broader term, often not properly determined by stars in my experience..
but difficulty/physical difficulty is a much broader term, often not properly determined by stars in my experience..
Well if you're arguing that Star Rating should measure other things instead of difficulty to FC I can't say I agree with you honestly but I see your reasoning
I don't think that
Scarlet rose has a ridiculously low OD as well as small spacing. Forgotten also has ridiculously small spacing. Unfortunately we in a world where spacing dominates difficulty and pp calculations. :p
I feel like each component of pp/difficulty calculations needs to bliow up exponentially at some point the way acc does (ie. between 99% and 100% on high OD). Otherwise maps that are ridiculous in one or two components but a joke in the rest aren't overshadowed by maps that are merely above average in all components.
I feel like each component of pp/difficulty calculations needs to bliow up exponentially at some point the way acc does (ie. between 99% and 100% on high OD). Otherwise maps that are ridiculous in one or two components but a joke in the rest aren't overshadowed by maps that are merely above average in all components.
E m i
this, no map takes its difficulty from 400bpm star jumps where the circles still touch (Streams are already considered by the star rating to be exactly this) or 100bpm cs7 fullscreen jumps.chainpullz wrote:
Scarlet rose has a ridiculously low OD as well as small spacing. Forgotten also has ridiculously small spacing. Unfortunately we in a world where spacing dominates difficulty and pp calculations. :p
I feel like each component of pp/difficulty calculations needs to bliow up exponentially at some point the way acc does (ie. between 99% and 100% on high OD). Otherwise maps that are ridiculous in one or two components but a joke in the rest aren't overshadowed by maps that are merely above average in all components.
on a side note, no one will ever SS forgotten because of how easy it is to get a few 100s everywhere, you can't avoid all of them.
maybe mizuki- or keigoclear autopilot 99.5%
Alright, compare which of these is more physically difficulty and then tell me there is nothing wrong with the star rating system. https://osu.ppy.sh/b/129285 https://osu.ppy.sh/b/255342Mahogany wrote:
Uh, yes it does? The system might have a few failings, but it very, very much does determine difficulty.B1rd wrote:
Star rating does not determine difficulty. Not even physical difficulty.
If we're arguing that Scarlet Rose is far tougher than 5 stars, then that's a problem with the star rating system, not with the PP system.
Star rating and pp are connected. Sliders should give more star rating and in turn more pp, and getting 100's on slider ends or not should give an impact on the aim pp you get from it. There also needs to be a small reading pp bonus for sliders, as well as lower AR.
E m i
U can actually just not care abt pp and enjoy game like rrtyui
U dont need to be ranked 1 in the rankings to be widely regarded as the best player by the community
U dont need to be ranked 1 in the rankings to be widely regarded as the best player by the community
Can't be bothered.B1rd wrote:
Alright, compare which of these is more physically difficulty and then tell me there is nothing wrong with the star rating system. https://osu.ppy.sh/b/129285 https://osu.ppy.sh/b/255342
Besides, I've never said there's nothing wrong with the star system.
So, as I said, the problem is with star rating calculation, not with the PP system, as far as I can tell.B1rd wrote:
Sliders should give more star rating and in turn more pp, and getting 100's on slider ends or not should give an impact on the aim pp you get from it. There also needs to be a small reading pp bonus for sliders, as well as lower AR.
I said that star rating does not determine physical difficulty. And you contradicted me. So that mean you must have thought that the star rating does correctly determine physical difficulty.Mahogany wrote:
Can't be bothered.B1rd wrote:
Alright, compare which of these is more physically difficulty and then tell me there is nothing wrong with the star rating system. https://osu.ppy.sh/b/129285 https://osu.ppy.sh/b/255342
Besides, I've never said there's nothing wrong with the star system.
So, as I said, the problem is with star rating calculation, not with the PP system, as far as I can tell.
And I already said, the star rating is linked to the pp system, so they both need to be fixed.