That cap is impossible to humanly obtain on most maps, though. But thanks for answering.

Rewben2 wrote:

Someone I know made an unlimitation play and got 177pp for this score: http://puu.sh/8J8SX.jpg. 89% acc, 1 miss.
I made this play: http://puu.sh/8J8VD.jpg. I have 2.5% better accuracy and another miss, yet the score isn't even on my top performances so is worth <108pp.
For frame of reference, a 97% accuracy fc is worth 217pp, http://puu.sh/8J907.jpg.

I doubt the insane pp difference is caused by 1 miss (the improved accuracy I got should just about account for the extra miss alone, shouldn't it?) so is my score worth so little because of low max combo? Other than accuracy and the amount of misses, the only thing that could be causing a loss of pp is max combo, right?

If so, I'm a bit surprised that a lower max combo can really affect a score from being 180~ to <108.

I have noticed this same thing happening with EVERY map, combo being the only variable. Unfortunately I do not have exact numbers, but here are the rough figures from memory: (on https://osu.ppy.sh/b/291996)
I had ~400/545 combo with 96% accuracy and was given about 40pp for it
Then I went back and got a full combo with a slightly higher accuracy (~96.5%) and was given 100pp for it

another example (exact numbers from https://osu.ppy.sh/p/beatmap?b=229180):
~96% accuracy with 539/617 combo: 79pp
SS: 131pp

third example for demonstration's sake (exact numbers from https://osu.ppy.sh/p/beatmap?b=93842):
~97% accuracy with 330/892 combo: 56pp
SS: 150pp

How the <edited explicit> is a full combo worth more than double a near-full combo with comparable accuracy? PPv2 weights combo so high that combo is pretty much the only deciding factor until you get a full combo (and then accuracy only shifts +/- 10% or so).

mastaa_p wrote:

Rewben2 wrote:

Someone I know made an unlimitation play and got 177pp for this score: http://puu.sh/8J8SX.jpg. 89% acc, 1 miss.
I made this play: http://puu.sh/8J8VD.jpg. I have 2.5% better accuracy and another miss, yet the score isn't even on my top performances so is worth <108pp.
For frame of reference, a 97% accuracy fc is worth 217pp, http://puu.sh/8J907.jpg.

I doubt the insane pp difference is caused by 1 miss (the improved accuracy I got should just about account for the extra miss alone, shouldn't it?) so is my score worth so little because of low max combo? Other than accuracy and the amount of misses, the only thing that could be causing a loss of pp is max combo, right?

If so, I'm a bit surprised that a lower max combo can really affect a score from being 180~ to <108.

I have noticed this same thing happening with EVERY map, combo being the only variable. Unfortunately I do not have exact numbers, but here are the rough figures from memory: (on https://osu.ppy.sh/b/291996)
I had ~400/545 combo with 96% accuracy and was given about 40pp for it
Then I went back and got a full combo with a slightly higher accuracy (~96.5%) and was given 100pp for it

another example (exact numbers from https://osu.ppy.sh/p/beatmap?b=229180):
~96% accuracy with 539/617 combo: 79pp
SS: 131pp

third example for demonstration's sake (exact numbers from https://osu.ppy.sh/p/beatmap?b=93842):
~97% accuracy with 330/892 combo: 56pp
SS: 150pp

How the <edited explicit> is a full combo worth more than double a near-full combo with comparable accuracy? PPv2 weights combo so high that combo is pretty much the only deciding factor until you get a full combo (and then accuracy only shifts +/- 10% or so).

Tom posted the formula for max combo. It is pretty harsh yeah, unless you're towards the full-combo mark your pp output is going to get shot down by quite a bit. If you get less than half of the maximum combo you can pretty much say goodbye to your pp. I'm not complaining, the intention is that the combo is the main way of getting score thus it should be a big factor in getting pp, which I understand.

Your first example is rather interesting; the difference is a tiny bit of accuracy and the first score was about 73%~ of the max combo, yet the pp output was about 2.5x higher for the second score. Are you sure those are the exact values you had for both before and after and the pp given? Because that doesn't seem normal. Did you have a few misses on the first score before you fc'd? That could explain it.

The last two examples are rather expected. The second example has pretty bad accuracy and the last one is very far off a fc with also missing acc.

Rewben2 wrote:

mastaa_p wrote:

I have noticed this same thing happening with EVERY map, combo being the only variable. Unfortunately I do not have exact numbers, but here are the rough figures from memory: (on https://osu.ppy.sh/b/291996)
I had ~400/545 combo with 96% accuracy and was given about 40pp for it
Then I went back and got a full combo with a slightly higher accuracy (~96.5%) and was given 100pp for it

another example (exact numbers from https://osu.ppy.sh/p/beatmap?b=229180):
~96% accuracy with 539/617 combo: 79pp
SS: 131pp

third example for demonstration's sake (exact numbers from https://osu.ppy.sh/p/beatmap?b=93842):
~97% accuracy with 330/892 combo: 56pp
SS: 150pp

How the <edited explicit> is a full combo worth more than double a near-full combo with comparable accuracy? PPv2 weights combo so high that combo is pretty much the only deciding factor until you get a full combo (and then accuracy only shifts +/- 10% or so).

Tom posted the formula for max combo. It is pretty harsh yeah, unless you're towards the full-combo mark your pp output is going to get shot down by quite a bit. If you get less than half of the maximum combo you can pretty much say goodbye to your pp. I'm not complaining, the intention is that the combo is the main way of getting score thus it should be a big factor in getting pp, which I understand.

Your first example is rather interesting; the difference is a tiny bit of accuracy and the first score was about 73%~ of the max combo, yet the pp output was about 2.5x higher for the second score. Are you sure those are the exact values you had for both before and after and the pp given? Because that doesn't seem normal. Did you have a few misses on the first score before you fc'd? That could explain it.

The last two examples are rather expected. The second example has pretty bad accuracy and the last one is very far off a fc with also missing acc.

I probably had a miss or two, I was just trying to highlight how ridiculous the weight is on combo as opposed to accuracy or map difficulty (50% combo on a map twice as hard as another with a full combo is worth less than half the fc pp). As to addressing the first example: I can verify that the pp given was around 40 (I was annoyed by this at the time) and the exact figures from the score were 400/535 combo with ~97% acc. Somehow I managed that with 8 misses, no idea how I did that XD.

Tom94 wrote:

Let me repeat once again, the system rates the quality of your performance. It is not intended to give a lot of points to plays on hard maps which you can barely pass. Imagine playing the piano. You don't perform in a concert with a damn hard piece, that you can barely play with quite a lot of mistakes.

This doesn't apply perfectly, but the idea is the same; closer to perfect performances are rewarded, breaking combo kind of kills that.

hi, a friend of mine recently improved his score on With a Dance Number [0108 style]:

the supposed improvement was from a 700 combo 96% score to this. he actually lost pp for setting the score, somehow - and i can't think of any possible causes for this happening(not to mention, the map itself is a lot harder than it is credited for, and i'm blaming the difficulty calculation for that). could you look at it and try to find the cause for it? if it was a bug of sorts, it should probably be checked, and if not, i'm not sure what the system is doing, but it can't possibly be what it should do.

e: http://osu.ppy.sh/ss/1627393 this is the old score that was overwritten - 3 less 100s, but 11 misses as opposed to none, for reference.

I just want to ask. Does Hidden or Flashlight increase pp in any way? If so by how much?

-GN, that seems like a glitch. Less misses, more max combo, better accuracy, same mods. There's no way it could have possibly given less.

@Monstrata, yes, they do give bonuses. Refer to the wiki page about performance points for an explanation.

mastaa_p wrote:

I probably had a miss or two, I was just trying to highlight how ridiculous the weight is on combo as opposed to accuracy or map difficulty (50% combo on a map twice as hard as another with a full combo is worth less than half the fc pp). As to addressing the first example: I can verify that the pp given was around 40 (I was annoyed by this at the time) and the exact figures from the score were 400/535 combo with ~97% acc. Somehow I managed that with 8 misses, no idea how I did that XD.

There lies your problem. Eight misses, with or without combo and accuracy, will murder your PP completely. I wouldn't be surprised if that alone is what cut down half of the PP.

This isn't a problem with the system, missing is punished harshly, as it should be. I, too, have a ton of scores with 1 miss, 80~% combo, and 98~% accuracy. Just get over it. Think of it this way, the harder it is to get maximum PP, the more satisfying it is when you get it with a good play.

-GN wrote:

hi, a friend of mine recently improved his score on With a Dance Number [0108 style]:

pic

the supposed improvement was from a 700 combo 96% score to this. he actually lost pp for setting the score, somehow - and i can't think of any possible causes for this happening(not to mention, the map itself is a lot harder than it is credited for, and i'm blaming the difficulty calculation for that). could you look at it and try to find the cause for it? if it was a bug of sorts, it should probably be checked, and if not, i'm not sure what the system is doing, but it can't possibly be what it should do.

e: http://osu.ppy.sh/ss/1627393 this is the old score that was overwritten - 3 less 100s, but 11 misses as opposed to none, for reference.

There is no way, that the newer score is giving less pp than the earlier one. I suspect, that your friend lost ranks due to others passing him, while the pp gain from With a Dance Number was just too little to have any effect on his total pp. With a Dance Number not even being in his top100 pp scores confirms that theory.

I do acknowledge, that With a Dance Number should be worth substantially more pp than it currently gives - which is a fault of the difficulty algorithm. Maps with very low spacing and fast singles are currently severely underrated, but I have not found a way to properly fix this without breaking other things far too much.

Suggestion: display performance points along with the other measures of performance after clearing a map.

why does FL HD give more PP than HD HR?



the play was inferior in all respects
i suppose memorizing takes more skill than playing with smaller circles and od10 on a map with fullscreen jumps though so perhaps this is intendedright..

It's exactly because the map has fullscreen jumps, they are kinda overrated and on top of that you have the flat bonus from HD + FL given to the aim value

Altough I'd say the HD FL looks more impressive to me than the HD HR score.

I think people forget that when you play with FL, you still have to make those relatively hard jumps. I think FL should give more pp on jumper maps like Airman. Seriously if you could do FL on Airman that's way better than a HD HR play.

the FL bonus alone is like 50 aim on that map, and the stat that gives the most bonus on hr (acc) was really low on the hr play, so it makes sense that the fl score gives more.

still seems wrong

Bassist Vinyl wrote:

still seems wrong

just your opinion

GoldenWolf wrote:

Bassist Vinyl wrote:

still seems wrong

just your opinion

feedback

The acc on those scores were so low it doesn't even matter (in fact, it wouldn't matter if it was higher because the map i od9 without hr) and FL>HR in terms of aim (i do agree with that and if you don't, it's just your opinion).

Rewben2 wrote:

@Monstrata, yes, they do give bonuses. Refer to the wiki page about performance points for an explanation.

Thanks. I've read the wiki page and now I'm wondering how much more pp will you get from playing on Hidden and Flashlight? Flashlight is clearly more difficult so i'd expect a much higher pp multiplier but is it to the extent of DoubleTime?

Also asking this because I am now playing Hidden to increase some of my scores but i notice my accuracy tends to drop by ~2%. I just want to know if getting a higher score with hidden can still make me lose pp due to a lower accuracy.

monstrata wrote:

Also asking this because I am now playing Hidden to increase some of my scores but i notice my accuracy tends to drop by ~2%. I just want to know if getting a higher score with hidden can still make me lose pp due to a lower accuracy.

It's the easiest way to lose pp. When in doubt, don't submit lower accuracy HD scores, especially if you're replacing a 99% with a 97%, or a 1x100 with 3x100.

monstrata wrote:

Also asking this because I am now playing Hidden to increase some of my scores but i notice my accuracy tends to drop by ~2%. I just want to know if getting a higher score with hidden can still make me lose pp due to a lower accuracy.

This is PPv1 style, since a 96% HD is MUCH easier than an SS no mods and gives more score. You should avoid this now, because HD isn't worth nearly as much PP-wise as it is score-wise.
It's pretty much also the reason I stopped playing HD and started having accuracy OCD.

monstrata wrote:

Rewben2 wrote:

@Monstrata, yes, they do give bonuses. Refer to the wiki page about performance points for an explanation.

Thanks. I've read the wiki page and now I'm wondering how much more pp will you get from playing on Hidden and Flashlight? Flashlight is clearly more difficult so i'd expect a much higher pp multiplier but is it to the extent of DoubleTime?

Also asking this because I am now playing Hidden to increase some of my scores but i notice my accuracy tends to drop by ~2%. I just want to know if getting a higher score with hidden can still make me lose pp due to a lower accuracy.

HD will only compensate for a 1%~ accuracy loss from what I've read, so a 97% HD score = 98% nomod score pp wise. There's more to it of course but I think that's an alright guideline.

Rewben2 wrote:

a 1%~ accuracy loss from what I've read, so a 97% HD score = 98% nomod score pp wise. There's more to it of course but I think that's an alright guideline.

It depends a lot on how hard the aim in the map is and how low the OD is. If you use HD on Airman for instance, then you will most likely gain pp, even if you lose a couple % of accuracy.

I still feel like I can "game" this pp system because some maps just give so much ranking for a reason I don't really understand (I have not read this thread, so please don't pitchfork me if it's been explained).

There are some scores/full combos with decent accuracy on my page that I'm proud of (Remote Control is the only one really depicted in my top ranks), but I can go FC some very easy TV Sizes/short 1:30-2 minute songs that I see have granted my friends a lot of pp and gain just as much. I know I'm at a lower pp rating than most posting about the system in this thread, but it just feels so silly at times.

A lot of these songs seem to have lots of slow streams/lots of short sliders, as opposed to more jumps and single taps. I don't know if these factors are weighting their score a lot, but that's honestly what it feels like.

I don't mean to be complaining or anything of the sort, just posting my current experience with the system.

Also, are you the Tom94 that released a program to protect against a vulnerability in the first set of Flyff private server files a good six or so years ago? If so, fancy seeing you again here working on this great game.

Cinagro wrote:

I still feel like I can "game" this pp system because some maps just give so much ranking for a reason I don't really understand (I have not read this thread, so please don't pitchfork me if it's been explained).

There are some scores/full combos with decent accuracy on my page that I'm proud of (Remote Control is the only one really depicted in my top ranks), but I can go FC some very easy TV Sizes/short 1:30-2 minute songs that I see have granted my friends a lot of pp and gain just as much. I know I'm at a lower pp rating than most posting about the system in this thread, but it just feels so silly at times.

A lot of these songs seem to have lots of slow streams/lots of short sliders, as opposed to more jumps and single taps. I don't know if these factors are weighting their score a lot, but that's honestly what it feels like.

I don't mean to be complaining or anything of the sort, just posting my current experience with the system.

Would be cool if you could provide examples of scores (user + beatmap) you believe are overrated / underrated. I (and others reading this thread) can than have a look.

Cinagro wrote:

Also, are you the Tom94 that released a program to protect against a vulnerability in the first set of Flyff private server files a good six or so years ago? If so, fancy seeing you again here working on this great game.

Yup, that's me.

Tom94 wrote:

Cinagro wrote:

I still feel like I can "game" this pp system because some maps just give so much ranking for a reason I don't really understand (I have not read this thread, so please don't pitchfork me if it's been explained).

There are some scores/full combos with decent accuracy on my page that I'm proud of (Remote Control is the only one really depicted in my top ranks), but I can go FC some very easy TV Sizes/short 1:30-2 minute songs that I see have granted my friends a lot of pp and gain just as much. I know I'm at a lower pp rating than most posting about the system in this thread, but it just feels so silly at times.

A lot of these songs seem to have lots of slow streams/lots of short sliders, as opposed to more jumps and single taps. I don't know if these factors are weighting their score a lot, but that's honestly what it feels like.

I don't mean to be complaining or anything of the sort, just posting my current experience with the system.

Would be cool if you could provide examples of scores (user + beatmap) you believe are overrated / underrated. I (and others reading this thread) can than have a look.

Cinagro wrote:

Also, are you the Tom94 that released a program to protect against a vulnerability in the first set of Flyff private server files a good six or so years ago? If so, fancy seeing you again here working on this great game.

Yup, that's me.

Holy shit that's ridiculous! I used to go by Organic on RZ, but I really doubt you remember me.

Well, I don't know as much if its that I still have room to grow, but these maps just felt really overrated compared to everything else that was at the top of my top ranks. I don't know if length is factored into the score at all, but maybe it should. It would probably be difficult to balance it between very long marathons and TV sizes, but it's probably something you've already considered.

https://osu.ppy.sh/b/217253?m=0
https://osu.ppy.sh/b/323726?m=0
https://osu.ppy.sh/b/245816?m=0
https://osu.ppy.sh/b/398341?m=0

What's the exact effect of hidden on the calculations? Currently it seems pretty ambiguous. I see it affects aim but to what extent and how? The wiki just says "Hidden and/or Flashlight is active."

Dexus wrote:

What's the exact effect of hidden on the calculations? Currently it seems pretty ambiguous. I see it affects aim but to what extent and how? The wiki just says "Hidden and/or Flashlight is active."

If those values haven't changed in a while:
HD: 18% bonus Aim value, 2% bonus Acc value.
FL:36% 50% bonus Aim, 2% bonus Acc.

Tom94 wrote:

Also, FlashLight is giving a 50% aim bonus since the last update (you can find it in the changelog somewhere, it's been a week or two), not a 36% one anymore.

i dont think hd got changed tho

Not sure if you care about the easier maps, but https://osu.ppy.sh/b/327673?m=0 seems overrated. Without a doubt the easiest Hard I've so far played (very simple & took 2 tries to SS), yet it's my top score.

Can anyone explain this?

Hi Tom, could you check this please?
https://osu.ppy.sh/forum/t/213478

If Full Tablet's explanation is true, it needs to be fixed to prevent confusion..
Thanks.

The new pp system is based on the skill that player has. Which is good. But i still miss ppv1 system (That pp system was based on the rank that the player achieved by playing). Still, it's good. (PPV1 seems much more logical to me...) Anyway, good job introducing ppv2

I really prefer ppv2 over ppv1 from what I've seen because it actually feels valid to say that if you have a marginally higher PP than the opposing player you are probably more skilled, rather than just saying someone played more or grinded more ranks.

From all the replays and everything I've seen in top players this is what I think needs to be buffed/nerfed/changed. I'm not a very good player or anything but I guess I still have an opinion

1. Length (in hit objects) need to have more of a value. Players with high stamina in maps should be rewarded more, with some of the hard long maps/marathons actually being pp farmable. If you look at the best scores in osu!tp top 50 scores, there are 11 TV Size DT Scores (13 if you count Killer Song), that seriously popped out of nowhere, and didn't seem to be as hard (didn't take as much retries or effort, from what I've seen in spectating) as some of the longer maps like Freedom Dive or Bikko's HDHR of Mago De Oz.

I decided to think of a couple marathons that would be fucking sick to DT, and what came to my mind was: https://osu.ppy.sh/s/108296
This map is a 12 min long marathon that is p popular now, and was the map I personally used to practice AR9.

With that, I used osu!tp to estimate the amount of PP a HD,DT play of this map would be worth (with acc around the lower 99%, because its simply more convinient) and (although I know DT recalculates speed and aim difficulty) DT tends to buff the aim and speed values by about x3, and HD gives a 18% bonus to aim, DT/HD gives a x2.05 buff to Acc from raw estimate, this gives God Only Knows DT a surprisingly low tp value of 379 estimate (probably off, but aroudn that area)

That is less than a DT HD Play of this map: https://osu.ppy.sh/b/408128 -- a one and a half minute song which gave sayo 435tp after doing it. The stamina required to do all the jumps and rather high BPM streams (specifically at the mars part) seems to be more than the DT sayo did on the previously mentioned map.

The only problems I see are maps like The Unforgiving being worth a jesus load of pp lol.

Well I guess i'll just drop some maps that in my opinion wouldn't be as undervalued if length were to be given a bulk

https://osu.ppy.sh/s/25403 - Various Artists - osu! Stream Compilation
https://osu.ppy.sh/b/343741 - Hanatan, yuikonnu & Mitani Nana - Songs Compilation [Rabbit Face]
Every DragonForce map
https://osu.ppy.sh/b/252238 - Tatsh - IMAGE -MATERIAL- <Version 0>

ok wall of text 1 is complete, ill try to make no.2 a bit shorter

2. Single Tap maps high BPM whatever

Although Scarlet Rose isn't the hardest map, I have no idea why WWW's HD/HR FC with only 6 100s is rated so low, not even worth 300tp on osu!tp, on a map that where the top 50 aren't even all FCs, or why With a Dance Number gives so little pp. It was mentioned previously in this thread that it should be buffed somehow, but really, val0108's maps really need to be worth more.

I did a similar test to the God Only Knows play, but this time looked at a play that has already been done

https://osu.ppy.sh/b/99094 - Megpoid GUMI - Carnival

Cookiezi HD,DT'd this map (I have no fucking clue how, but he did) Comparing the score on osu!tp of the player, furikake on carnival, which has a similar accuracy to what cookiezi got on the map, and applying the DT estimates of x3 Aim,x3 speed and x2 acc, carnival would be worth an estimate of 353tp. For one of the hardest DTs Cookiezi ever did, that is really little, in comparison to a DT like everyone doing Magic Girl!!, which is worth more tp than Carnival, as Magic girl is more focused on DT Jumps than single tapping like jesus. (mentioning this map because it has a similar length) fast singles definitely need a buff.

3. Just a quick note, FL needs to be buffed on longer maps, as being able to memorize a longer map is much harder than a short map. It's easier to remember a short song with jumps than Flashlight on a marathon.

TUOPlays wrote:

2. Single Tap maps high BPM whatever

Although Scarlet Rose isn't the hardest map, I have no idea why WWW's HD/HR FC with only 6 100s is rated so low, not even worth 300tp on osu!tp, on a map that where the top 50 aren't even all FCs, or why With a Dance Number gives so little pp. It was mentioned previously in this thread that it should be buffed somehow, but really, val0108's maps really need to be worth more.

From single tapping alone, how would you ever deter the player's preference to either alternate or single tap apart from looking at their replays? I believe val maps don't need a buff, they sit perfectly where they currently sit. There are a lot of other maps that are total BM compared to val's.

Clearly scarlet rose should atleast be counted as an extra diff. Still, I think Tom already knows most of the faults in calculating difficulty. It's just really hard to balance some things out without breaking others.

A question about ranks - My friend noticed that he's a higher rank than someone else yet having more pp, here are the profiles: https://osu.ppy.sh/u/1559215 and https://osu.ppy.sh/u/3693177

My friend is 4 pp higher yet 200 ranks higher. At first I thought it may have been an issue where the players ranks weren't updating due to not having played, but both these players have set scores as recent as 2 hours ago.

Rewben2 wrote:

A question about ranks - My friend noticed that he's a higher rank than someone else yet having more pp, here are the profiles: https://osu.ppy.sh/u/1559215 and https://osu.ppy.sh/u/3693177

My friend is 4 pp higher yet 200 ranks higher. At first I thought it may have been an issue where the players ranks weren't updating due to not having played, but both these players have set scores as recent as 2 hours ago.

iirc, updates do take some time to be updated to the correct values, it happened to me and a lot of other people. 2 hours is too quick for it to update, it happens where there is a lot of shifting going on.

Melt3dCheeze wrote:

Rewben2 wrote:

A question about ranks - My friend noticed that he's a higher rank than someone else yet having more pp, here are the profiles: https://osu.ppy.sh/u/1559215 and https://osu.ppy.sh/u/3693177

My friend is 4 pp higher yet 200 ranks higher. At first I thought it may have been an issue where the players ranks weren't updating due to not having played, but both these players have set scores as recent as 2 hours ago.

iirc, updates do take some time to be updated to the correct values, it happened to me and a lot of other people. 2 hours is too quick for it to update, it happens where there is a lot of shifting going on.

This is correct. The rank shown on the profile doesn't update in realtime.