forum
Performance Points feedback and suggestions (Standard)
posted
Total Posts
2,750
hd hr on that map kpielak wrote:
you guys obviously haven't seen this map https://osu.ppy.sh/s/42234
So what about this map? It's just a troll for those who somehow can't spin. But bear in mind that 90% of players can spin incredibly fast (enough to past augoeides or desire dream) and nobody apart from newbs finds it really impressive.
In overall, those kinds of maps are more like DRAIN hard, not SPINNER hard. We're talking about situation where player x has better score than player y because of the spinner. And according to Vile player x should gain more PP for that. I don't need to comment on this idea, I guess.
In overall, those kinds of maps are more like DRAIN hard, not SPINNER hard. We're talking about situation where player x has better score than player y because of the spinner. And according to Vile player x should gain more PP for that. I don't need to comment on this idea, I guess.
Drain is a dumb reason to fail a map under any circumstance. I was only saying that anything that affects difficulty negatively or positively should affect pp. If you say it makes it easier then maps should be worth less pp. I just don't feel like they should be considered like they are break time there when they clearly are more than that.
Also when I made the suggestion I thought that you could figure out the bonus that one could get from the map. The cons slightly outweigh the pros now that it absolutely must require new information to submitted instead of existing plays.
Also when I made the suggestion I thought that you could figure out the bonus that one could get from the map. The cons slightly outweigh the pros now that it absolutely must require new information to submitted instead of existing plays.
I don't like how I can lose pp by using a mod and getting higher score but lower accuracy. I wouldn't mind just not getting pp but I don't wanna use mods if I'll lose pp unless I work back up to the same accuracy as before.
This has already been discussed multiple times. There is currently no way around this issue.Jean-Christophe wrote:
I don't like how I can lose pp by using a mod and getting higher score but lower accuracy. I wouldn't mind just not getting pp but I don't wanna use mods if I'll lose pp unless I work back up to the same accuracy as before.
Make mod usage worth more. Easy fix.mcdoomfrag wrote:
This has already been discussed multiple times. There is currently no way around this issue.Jean-Christophe wrote:
I don't like how I can lose pp by using a mod and getting higher score but lower accuracy. I wouldn't mind just not getting pp but I don't wanna use mods if I'll lose pp unless I work back up to the same accuracy as before.
I would say that there's a limit to how much accuracy one should lose though. If you're getting double digit 100s then obviously this will affect your pp count greatly. There is no preventing a terrible performance with an added mod overwriting a good performance.
You know damn well that finding a good balance between not making mods useless without giving them too much reward is a rather complex and subjective solution. Even so, what I meant in that post was aimed more-so towards the notion of being able to disregard your score so that you don't lose PP, or something along those lines. The servers can only store your top score because of the way the system works and has always worked, so Tom94 can only work with that saved score.TheVileOne wrote:
Make mod usage worth more. Easy fix.
Topic Starter
Mod for at least HardRock and DoubleTime will likely never be adjusted manually. That is, because they are literally treated equally to nomod scores, just with the mods' effects applied.TheVileOne wrote:
Make mod usage worth more. Easy fix.
I would say that there's a limit to how much accuracy one should lose though. If you're getting double digit 100s then obviously this will affect your pp count greatly. There is no preventing a terrible performance with an added mod overwriting a good performance.
If anything is at fault, then the algorithm for scaling the wrong way.
Compare previous score's pp with new score's pp and use the higher value???mcdoomfrag wrote:
This has already been discussed multiple times. There is currently no way around this issue.Jean-Christophe wrote:
I don't like how I can lose pp by using a mod and getting higher score but lower accuracy. I wouldn't mind just not getting pp but I don't wanna use mods if I'll lose pp unless I work back up to the same accuracy as before.
If this is supposed to be an indication of skill then there's no reason to assume a player got worse for having his/her score overwritten with one of lower pp value...
If new pp value is smaller than current pp value for some map, don't change the pp. That's one line of code. Seems like common sense to me to make it work that way.
Only data you need is pp obtained and pp for the same map for the same player which looks like it's stored on server.
Maybe there can be a disconnect between pp given for a map and top score saved but that doesn't really matter.
I agree, i feel somewhat dissapointed at being unable to play HR, Hidden, DT or any combination of them, previously a 92% HR HD fc was worth way more than a 98% nomod FC - now it's not only worth less, but it will overwrite the score and take away PP in a way that prevents you from ever getting it back. Everyone that i know who plays the game at a decent level feels somewhat cheated by this - nobody likes to get 150pp and then overwrite it with a 130pp score and be unable to get their 150pp back - at the VERY least, if there is no possible workaround, let us delete the score and re-play to score again!
I think this is a pretty basic and important flaw in one of the key features of the game we all love, so i'm sad to see it unfixed for months
osu plz <3
I think this is a pretty basic and important flaw in one of the key features of the game we all love, so i'm sad to see it unfixed for months
osu plz <3
Topic Starter
No. Please don't just assume things without reading the discussion.Jean-Christophe wrote:
Compare previous score's pp with new score's pp and use the higher value???
If this is supposed to be an indication of skill then there's no reason to assume a player got worse for having his/her score overwritten with one of lower pp value...
If new pp value is smaller than current pp value for some map, don't change the pp. That's one line of code. Seems like common sense to me to make it work that way.
Only data you need is pp obtained and pp for the same map for the same player which looks like it's stored on server.
Maybe there can be a disconnect between pp given for a map and top score saved but that doesn't really matter.
The simple argument against your proposal is, that whenever the algorithm is changed all the scores' pp values need to be recomputed - which is impossible without actually storing all the most giving scores. Before you propose just storing the most-giving pp score per map, please read the discussion about it containing the arguments against it.
The most optimal solution would be storing mod-specific highscores. Until that happens there won't be a feasible fix.
Just gonna drop some feelings out here real quick: I think deathstreams are incredibly overrated. So many maps contain relatively easy long streams and provide insane amounts of pp from them. It'd be nice to see faster streams rewarded more and longer ones rewarded less (hopefully this will balance out for long fast streams). At the moment fast stream provide barely any amount of pp compated to longer streams(say a few 270bpm 12 note streams vs a 40 note 220bpm stream).
To add to this, I find large jumps to be rewarded too greatly while very fast short jumps are underrated. Half screen 300bpm jumps should be valued as much as full screen 230bpm jumps imo, especially when in rapid succession for relatively large periods of time.
It also seems that the difficulty increase/decrease from the size of hit circles is overrated. Large hit circles seem to help more than the current system thinks while smaller hit circles make the map a fair deal harder.
My final point is the scaling of pp with map length. The system currently seems to scale penalties for combo based on how far from a full combo you are and amount of misses in total, this is kinda frustrating because it means that full-combos on a 1200 combo map are worth as much as a slightly easier 500 combo map. There should be a large reward for comboing longer maps and less of a penalty for missing on them (it seems strange to me that a 900 combo on a map with 1200 max combo awards LESS pp than a map of equal difficulty and 500 max combo with an fc).
To add to this, I find large jumps to be rewarded too greatly while very fast short jumps are underrated. Half screen 300bpm jumps should be valued as much as full screen 230bpm jumps imo, especially when in rapid succession for relatively large periods of time.
It also seems that the difficulty increase/decrease from the size of hit circles is overrated. Large hit circles seem to help more than the current system thinks while smaller hit circles make the map a fair deal harder.
My final point is the scaling of pp with map length. The system currently seems to scale penalties for combo based on how far from a full combo you are and amount of misses in total, this is kinda frustrating because it means that full-combos on a 1200 combo map are worth as much as a slightly easier 500 combo map. There should be a large reward for comboing longer maps and less of a penalty for missing on them (it seems strange to me that a 900 combo on a map with 1200 max combo awards LESS pp than a map of equal difficulty and 500 max combo with an fc).
I don't know as much about the long streams vs. fast streams since I'm not much of a speed player, but I definitely think smaller hitcircles and longer maps need to be rewarded much more.
I understand how the pp system currently accounts for smaller hitcircles, and that increasing your area is just like playing larger hitcircles, etc. But I can say in practice, it definitely isn't working like that. With how large the screen is, you can only make your area so big, and playing smaller circles requires more steady hands and better aim than the system makes it out to need due to how players snap. The difficulty gap is much more evident in CS6/CS7 maps + HR where aim is super hard but the aim tp is not very high, but I still think CS5+HR needs to be rewarded much more too. As a HDHR player, I've grown to dismiss CS5 maps as pp maps whenever I see them because they're always so much harder than CS4 maps that give the same amount of pp. I'm not sure exactly how to go about it so that the algorithm for it "makes sense", but I definitely think the scaling for points should be higher than the scaling for the circles themselves.
And as for longer maps, I've found that they generally do give lower pp than they deserve because of how much longer it takes to full combo them if you have some given percentage chance to miss at any point in the map. I mean, if you have a 1/10 chance at FCing a 500 combo map, you would ideally have a 1/100 chance at FCing a 1000 combo map, which means it would take on average 10 times as long. Obviously, if the map is really easy to FC, then it will take on average less than twice as long, but those maps won't give much pp anyway. Nerves are also a much bigger factor on longer maps. I don't think the map length scaling needs to be massive, but I think it should be at least noticeable.
I understand how the pp system currently accounts for smaller hitcircles, and that increasing your area is just like playing larger hitcircles, etc. But I can say in practice, it definitely isn't working like that. With how large the screen is, you can only make your area so big, and playing smaller circles requires more steady hands and better aim than the system makes it out to need due to how players snap. The difficulty gap is much more evident in CS6/CS7 maps + HR where aim is super hard but the aim tp is not very high, but I still think CS5+HR needs to be rewarded much more too. As a HDHR player, I've grown to dismiss CS5 maps as pp maps whenever I see them because they're always so much harder than CS4 maps that give the same amount of pp. I'm not sure exactly how to go about it so that the algorithm for it "makes sense", but I definitely think the scaling for points should be higher than the scaling for the circles themselves.
And as for longer maps, I've found that they generally do give lower pp than they deserve because of how much longer it takes to full combo them if you have some given percentage chance to miss at any point in the map. I mean, if you have a 1/10 chance at FCing a 500 combo map, you would ideally have a 1/100 chance at FCing a 1000 combo map, which means it would take on average 10 times as long. Obviously, if the map is really easy to FC, then it will take on average less than twice as long, but those maps won't give much pp anyway. Nerves are also a much bigger factor on longer maps. I don't think the map length scaling needs to be massive, but I think it should be at least noticeable.
i dont think high cs is that underrated.
this map is worth 80+ aim tp with hd dt only because of the cs, the map itself is a piece of cake
this map is worth 80+ aim tp with hd dt only because of the cs, the map itself is a piece of cake
I agreejesus1412 wrote:
At the very least low cs is overrated.
One thing for sure: DT is clearly overated.
I mean, completely messed up.
I mean, completely messed up.
People who say this are very careless. In some environments it's overrated but in others it's severely underrated. Large jump maps can rewards up to 500pp while far faster maps can barely top 300pp. It really depends on the map, jumps formed from dt tend to be easier to follow which is why people would call dt an overrated mod while short fast streams and smaller faster jumps are rarely ever mentioned because they don't fit the "overrated" trend. It seems the problem is that balanced maps are underrated rather than extreme maps overrated, a map which is extremely fast and mildly jumpy seems to be penalized far too much for not having long streams or immense jumps despite maintaining huge difficulty throughout (I'm looking at maps DT'ed by dragonhuman and co.).Yales wrote:
One thing for sure: DT is clearly overated.
I mean, completely messed up.
I'm not talking about maps DT>160BPM, the amount of pp is clearly deserved in this case.jesus1412 wrote:
People who say this are very careless. In some environments it's overrated but in others it's severely underrated. Large jump maps can rewards up to 500pp while far faster maps can barely top 300pp. It really depends on the map, jumps formed from dt tend to be easier to follow which is why people would call dt an overrated mod while short fast streams and smaller faster jumps are rarely ever mentioned because they don't fit the "overrated" trend. It seems the problem is that balanced maps are underrated rather than extreme maps overrated, a map which is extremely fast and mildly jumpy seems to be penalized far too much for not having long streams or immense jumps despite maintaining huge difficulty throughout (I'm looking at maps DT'ed by dragonhuman and co.).Yales wrote:
One thing for sure: DT is clearly overated.
I mean, completely messed up.
But those 130bpm maps that got more than 50 players with SS DTHD. (no offense, I know a SS is a hard thing).
I remember I did a random hddt play one time on a 130bpm map (200~ with dt), got quite a bad accuracy, 1 miss or 2 at the middle of the map, the full combo would be at 700... And I got the exact same amout of pp than a FC hrhd 1300 combo and quite good accuracy with the same bpm (200~).
This is a totally random example to explain that there's A LOT of maps overated like this. Especially those DT <160 BPM that are even easier with DT. Nobody can deny it.
It looks like you're comparing two different maps' pp values with different mods, which isn't a good comparison. The two maps could have vastly different mapping patterns, which would skew the pp values.Yales wrote:
I remember I did a random hddt play one time on a 130bpm map (200~ with dt), got quite a bad accuracy, 1 miss or 2 at the middle of the map, the full combo would be at 700... And I got the exact same amout of pp than a FC hrhd 1300 combo and quite good accuracy with the same bpm (200~).
This is a totally random example to explain that there's A LOT of maps overated like this. Especially those DT <160 BPM that are even easier with DT. Nobody can deny it.
(Assuming you're talking about two different maps since you're saying that you DT'd something at "~200" BPM and comparing the pp value gained from that to someone with an "FC hdhr 1300 combo... with the same bpm (200~)")
I don't know any single map that becomes easier with DT, mind showing some examples?Yales wrote:
This is a totally random example to explain that there's A LOT of maps overated like this. Especially those DT <160 BPM that are even easier with DT. Nobody can deny it.
The only maps that become easier with DT are maps from 2007-2008 that have badly chosen difficulty settings.
http://osu.ppy.sh/s/31408GoldenWolf wrote:
I don't know any single map that becomes easier with DT, mind showing some examples?Yales wrote:
This is a totally random example to explain that there's A LOT of maps overated like this. Especially those DT <160 BPM that are even easier with DT. Nobody can deny it.
http://osu.ppy.sh/b/195039 <- I was talking about this map btw.
http://osu.ppy.sh/b/237649
http://osu.ppy.sh/b/61684
http://osu.ppy.sh/s/170775
https://osu.ppy.sh/s/95148 <- Perfect example too
etc.etc.etc.
Basically.
I'm comparing what I find easier/harder to play against the result.TMoI wrote:
It looks like you're comparing two different maps' pp values with different mods, which isn't a good comparison.
Assuming you're talking about two different maps
Let's take this one as example. https://osu.ppy.sh/s/95148 The BPM might be slow. It's way harder to clear/fc with HR than DT (BECAUSE this is a slow bpm). But at the end, you'll have more PP with DT. Something like this.TMoI wrote:
Assuming you're talking about two different maps
I wasn't talking about the same map. But this works too.
To make it clearer. Whether dt is overrated whether hr is underrated but in any case there's a problem here. At a point that I didn't even find necessary to explain this in my first post. I thought that was really obvious.
I tried to explain it in a general way but if you want my own example to explain this. It will be even more obvious.
To me it just seems like you're confusing "what I think is hard" with "What is hard".
I suck at HR. There are probably a bunch of maps I can FC easily with DT but not as easily with HR. Does that mean that HR should be worth more pp? No. it simply means I'm bad at HR. Nothing more.
HR starts getting you a lot of pp if you use it on maps that are hard.
I suck at HR. There are probably a bunch of maps I can FC easily with DT but not as easily with HR. Does that mean that HR should be worth more pp? No. it simply means I'm bad at HR. Nothing more.
HR starts getting you a lot of pp if you use it on maps that are hard.
Which is not the case for DT which gives a good amount of pp no matter what. Glad you understood.PlasticSmoothie wrote:
HR starts getting you a lot of pp if you use it on maps that are hard.
It depends on your individual skills. You get more accuracy with high bpm than low bpm (apparently even with the tighter timing windows when applying DT to slow maps). Personally, I find the DT version of the maps harder (specially this map https://osu.ppy.sh/s/95148, which is too fast for me currently; I get more accuracy with HR than with DT even though the OD is 7).Yales wrote:
This is a totally random example to explain that there's A LOT of maps overated like this. Especially those DT <160 BPM that are even easier with DT. Nobody can deny it.
http://osu.ppy.sh/s/31408
http://osu.ppy.sh/b/195039 <- I was talking about this map btw.
http://osu.ppy.sh/b/237649
http://osu.ppy.sh/b/61684
http://osu.ppy.sh/s/170775
https://osu.ppy.sh/s/95148 <- Perfect example too
etc.etc.etc.
Basically.
Let's take this one as example. https://osu.ppy.sh/s/95148 The BPM might be slow. It's way harder to clear/fc with HR than DT (BECAUSE this is a slow bpm). But at the end, you'll have more PP with DT. Something like this.
I wasn't talking about the same map. But this works too.
To make it clearer. Whether dt is overrated whether hr is underrated but in any case there's a problem here. At a point that I didn't even find necessary to explain this in my first post. I thought that was really obvious.
I tried to explain it in a general way but if you want my own example to explain this. It will be even more obvious.
I think that considering the faster version of the maps harder makes more sense, since they are technically harder. A player that knows how to stream slow or is fast enough to singletap the streams in the low bpm maps would consider them very easy.
You may have a point. Thank you for your answer.Full Tablet wrote:
It depends on your individual skills. You get more accuracy with high bpm than low bpm (apparently even with the tighter timing windows when applying DT to slow maps). Personally, I find the DT version of the maps harder (specially this map https://osu.ppy.sh/s/95148, which is too fast for me currently; I get more accuracy with HR than with DT even though the OD is 7).
I think that considering the faster version of the maps harder makes more sense, since they are technically harder. A player that knows how to stream slow or is fast enough to singletap the streams in the low bpm maps would consider them very easy.
But I don't think you're totally right. the new pp system is supposed to work on individual skill. Why I can't gain pp for trying to be accurate with hr then? I gain PP only by playing dt. More or less. At least I gain pp easier this way. -> It means there's a problem on the calculation of my skill then? My accuracy sucks, there's no way I reached a limit xD.
-> dt > hr (in my case even I find it easier, for the map I quoted at least).
Because yes, my topranks is full of random bad dt plays. while my osutp is full of hdhr for which I spent time, and it was hard to reach this result (FC, acc, etc.) Almost none of those maps are in my toprank PP though.. And I'm talking about insanes maps only, of course.
Here's a draw I made of what I think the system calculation looks like. (ofc, every map is different but osutp made this difference pretty well so... go check it).
This is supposed to be a pyramid.
Here's the version as it is now: http://puu.sh/8GYjs.jpg
And here's my improved version : http://puu.sh/8GYom.jpg
Fair enough I guess.
None of these maps were any easier with DT to me..
What you find easier/harder might be harder/easier for someone else, you won't neccesarily get rewarded for a score you struggled to set, and you might get rewarded a lot for a score you got on the first try, whatever is considered hard or not by the system
That doesn't mean the system is perfect of course, but I think not everyone can be totally pleased because we all have different opinion on what is and isn't hard
The actual balance is fine in my opinion, I would only like to see a way to take reading difficulty into account though
What you find easier/harder might be harder/easier for someone else, you won't neccesarily get rewarded for a score you struggled to set, and you might get rewarded a lot for a score you got on the first try, whatever is considered hard or not by the system
That doesn't mean the system is perfect of course, but I think not everyone can be totally pleased because we all have different opinion on what is and isn't hard
The actual balance is fine in my opinion, I would only like to see a way to take reading difficulty into account though
I'm not saying it's bad! I mean don't get me wrong, when I don't play just to chill I play for tp instead of PP (personal preference) but it's always a bit sad to get some tp for some scores I found really hard and the pp system doesn't acknowledge me this. While it does for the 2 dt maps I play per week. You're saying it's just a coincidence that it's dt maps, i don't know. anyway.GoldenWolf wrote:
None of these maps were any easier with DT to me..
What you find easier/harder might be harder/easier for someone else, you won't neccesarily get rewarded for a score you struggled to set, and you might get rewarded a lot for a score you got on the first try, whatever is considered hard or not by the system
That doesn't mean the system is perfect of course, but I think not everyone can be totally pleased because we all have different opinion on what is and isn't hard
The actual balance is fine in my opinion, I would only like to see a way to take reading difficulty into account though
That's it.
" I would only like to see a way to take reading difficulty into account though"
Yep.
But tp is pp.Yales wrote:
I'm not saying it's bad! I mean don't get me wrong, when I don't play just to chill I play for tp instead of PP (personal preference) but it's always a bit sad to get some tp for some scores I found really hard and the pp system doesn't acknowledge me this. While it does for the 2 dt maps I play per week. You're saying it's just a coincidence that it's dt maps, i don't know. anyway.GoldenWolf wrote:
None of these maps were any easier with DT to me..
What you find easier/harder might be harder/easier for someone else, you won't neccesarily get rewarded for a score you struggled to set, and you might get rewarded a lot for a score you got on the first try, whatever is considered hard or not by the system
That doesn't mean the system is perfect of course, but I think not everyone can be totally pleased because we all have different opinion on what is and isn't hard
The actual balance is fine in my opinion, I would only like to see a way to take reading difficulty into account though
That's it.
" I would only like to see a way to take reading difficulty into account though"
Yep.
EDIT: outdated pp*
DT should be nerfed.
end of story.
end of story.
DeletedUser_613592
You can't nerf DT without nerfing maps of similar BPM and difficulty settings. The system treats DT, HT, HR as if they are a map with those difficulty settings.
Or you should be buffed. DT as a whole should not be nerfed, if anything buffed in some areas. Small notes should be buffed and large notes should be nerfed. Large jumps should be nerfed and fast jumps should be buffed.dennischan wrote:
DT should be nerfed.
end of story.
END OF STORY NO DEBATE ALLOWED I AM SUPREME LEADER.
uhhm well i have a question.
even though i keep on getting PP my rank decrease like when i had 1000~~ i been 38 000 or so and now when i have 1057 i am 40 000 and seem like i can't increase in rank. any tips how to get my rank higher? i usually play hards/insanes in which i get [a/s in hards] and [b/c/a in insanes]
ps: if i wrote in wrong section dont hate me please~~
even though i keep on getting PP my rank decrease like when i had 1000~~ i been 38 000 or so and now when i have 1057 i am 40 000 and seem like i can't increase in rank. any tips how to get my rank higher? i usually play hards/insanes in which i get [a/s in hards] and [b/c/a in insanes]
ps: if i wrote in wrong section dont hate me please~~