Old thread msg:

Surprisingly the thread got very big

I'm almost 3200pp now

magicali wrote:

I have a question about this PP stuff, i play oftenly and i have nearly 2,000 pp on my account but my girlfriend plays almost as good as I but she has only 24 pp thats ridiculous, i think it's bugged since the other day when we came back to osu from like 1 year or more break our PP from 20s jumped to 300s 600s and but on my girlfriend it dropped back to 20s again, is this normal? i tried playing some songs on her account, the ones i think they have a big influence on the PP but they didn't do anything at all.

What can it be? It is outdated or what?

Her username is: itazuranakiss

I will post with her account later

I checked itazuranakiss account and she has 1,447pp not 24 o_O

I've been thinking a lot about the PP system lately, and people's reactions to it, so I am going to dump a little blog here. This is all my opinion and analysis, and in no way represents the opinions of Peppy, or the community. It is only what I see happening.

The community has a double standard about Farming vs Skill. On the one hand they believe that a person can manipulate a ranking system to generate better numbers, and thus appear to be a good player when they truly are not. On the other hand they think that a person's Skill should be able to be proven tangibly by doing Skillful things.

So we have a problem here... Any system that measures a player's performance in some way will be vulnerable to attack, because it will be claimed that people are Farming it (manipulating the numbers). However, we still need a system, because we want to be able to show how skilled we are, and "win" at the game, in the sense of succeeding in some measurable way.

So what do we do!? Well, Peppy gave us the answer... He made a system that is so complex that it is impossible for us to completely understand it. To further obfuscate things, he has not released all of the information on how the system works. This is an attempt to appease both sides of the community. On the one hand, the system is able to measure us, so we can use it to express our skill. On the other hand, it's so impossible to understand that it cannot easily be manipulated, and thus it is difficult to claim that Farming is spoiling the system.

It sounds like he's solved the problem right? Unfortunately no... there is still a problem. The problem now is that we have a system which we can't exactly "win" at, because we don't really know how. Farming is stunted by making it unclear how to get better numbers, but "winning", which is very closely related to farming because it deals with manipulating a system in your better interest, is also stunted by being called into doubt and made hard to understand. This means that we can't be clear on how to gain Skill, and we can't be clear about what is really Skill and what is possibly a new kind of Farming.

That said, the new PP system does seem to deliver a decent value for measuring Skill, despite its difficulty as a guideline for gaining Skill. It may be that your inactivity for so long has caused the PP system to have to do calculations based on very little information, and thus you have risen far above your equally skilled friend. The system is based on placements in individual maps, so it may be that you have some placements that happened to look better than her placements, or she just hasn't placed as many times as you have recently since beginning to play again. He old scores won't make any difference, so if she isn't making new ones, then it won't show in her PP. At least that's how I understand it.

Liutprando wrote:

magicali wrote:

I have a question about this PP stuff, i play oftenly and i have nearly 2,000 pp on my account but my girlfriend plays almost as good as I but she has only 24 pp thats ridiculous, i think it's bugged since the other day when we came back to osu from like 1 year or more break our PP from 20s jumped to 300s 600s and but on my girlfriend it dropped back to 20s again, is this normal? i tried playing some songs on her account, the ones i think they have a big influence on the PP but they didn't do anything at all.

What can it be? It is outdated or what?

Her username is: itazuranakiss

I will post with her account later

I checked itazuranakiss account and she has 1,447pp not 24 o_O

I would like you to re-check that, because her profile auto shows Taiko instead of osu (She played alot of Taiko and she is pretty good at it) but the thing is maybe the performance points are not updating at all with the Taiko being the default profile for her, i don't know.

Remember im talking about osu here.

And for the other post, Yeah i believe PP system is a great way to show the "skill" of each players.

So once again if you guys can check her profile again and take a look at a couple of recently played songs you will realize that the songs that she is able to do are much harder that what a 20s PP player can do.
Actually i did some research and the player just 1 rank above in one song had 1100 PP, for the song USATEI also the guy one rank higher (for like 25 points in the song) had also 1210PP.

PP only gets recorded if you get at least rank 500 on a song.
looking at her recent plays and the scores she got on them, they are most likely not within the 500 ranks.

The bulk of your 1928 pp comes from a single map (given the huge jump from 300 pp to 1900), which means that you managed to do well (top 500 with high accuracy) on a reasonably hard map. Apparently, your girlfriend hasn't yet done this.

pp is skewed in favor of your best plays, which means that a few (or one, in your case) high ranking scores will have a huge impact on your pp. See the infographic (http://osu.ppy.sh/wiki/images/d/d6/Perf ... points.png) for more info.

Thanks that last answer was what i was looking for, btw how do i get to know "which song" was the one that boosted my PP

nvm i found it.

Btw how does someone "drops" on PP?

It appears that you lose PP when people start out performing your score on songs that you have played. However, it's unclear exactly how much you lose for what, or how badly they have to beat you to create a change.

Any song that you rank above 500 on should affect your PP, but it is unclear how much, and sometimes it looks like you will lose PP when the system recalculates, if you're not performing better than everyone else is.

Recent scores have the most influence on PP, this is why the amount is gradually reduced over time.

G0r's answer is right too, but I'm pretty sure that the changes on the PP are very minor, unless you rank on a very recent map with an average score (then you get quickly pushed back to 500+).

getting a score knocked out of top 500 is a pretty big hit on pp, depending how much you have! losing a #1 hurts a lot too but i don't think you have many of those xD

There's no doubting this is a good system, my only small complaint is that only top 500 scores are taken into account, therefore encouraging people to play less than insane diffs with mods to get the pp, they get very little, but it adds up to a lot.

^That needs to be removed.

Ahhhh and that. Oh my I forgot.

....................

JappyBabes wrote:

^That needs to be removed.

That might be fine, it's so damn easy to get VERY high ranks on VERY old maps... like try any super old map on DT, you'll get top 6 easily and get tons of pp because most scores there are damn old so getting a high rank is easy, plus it's not likely that many players will go outrank you there since it's a map which gets something like a few plays per week.

It's not talking about date of map. It's talking about how long ago you got your SCORE on that map which is why it should be removed. There are some seriously skilled players hanging in the #100-200s just because they got some amazing scores pre-July '11.

I think the reason it is there is to keep scores relevant. ie. an old score doesn't exactly represent your current skill. I honestly don't care whether it's there or not, just stating that's probably the reason and it's a good enough reason for me ^^

So am I correct in understanding that cookie gets 0 peppy points for his #1 on Cry For Eternity (just as an example)?

If how it's explained in the picture is how it actually works and he doesn't improve any scores nor plays until 2013 he won't have a rank.

The inactivity drain won't go further down than half your original. It was stated in a pp thread by peppy somewhere and I think it's on the wiki too (Wow I'm lazy). So Cookizie won't get 'unranked' but he will start losing pp soon enough to make place for the current pros.

JappyBabes, look at bikko his last score would have put him at no "giving" pp last month. He's still top 100. He will lose pp, but the scores still remain to some degree i suppose.

There's probably thresholds, relative to your current pp, at which scores "give" or "don't give" pp. Bikko's probably lost half his pp on each individual score but the threshold would mean they still count in numbers and kick his score up to the number it is now. Otherwise White Wolf would be much higher than he is in pp score.

pp gain is logarithmic, it's harder and harder to gain as you have more... i'm thinking 7000 is probably the max, at least until there's more plays in the system xD

MillhioreF wrote:

losing a #1 hurts a lot too but i don't think you have many of those xD

For my case:
- Initial pp was around 4,700pp (forgot what it was, exactly)
- Played Klaus Badelt - He's a Pirate [Hard] and scored rank #1 with only HD+FL, Performance jumped up to 5,093pp
- Played some random maps until having Performance around 5,105pp
- rockleejkooo took my #1 on the said map above, Performance dropped heavily down to 5,009pp D: (you can see a noticeable jump-down spike at my performance graph lol)
- I re-took #1 again on the map with HD+HR+FL, and Performance raised up to 5,134pp

@Wishy22: That beatmap I mentioned above was ranked since 2008.

It's not that MAPS stop giving pp after a year, it's that scores that you beat won't give you any pp unless it was set less than a year ago.

If I go #1 rank a really old map with a decent score, I get no pp, because the records I beat were old.
If someone else beats my score and takes #1, he gets pp because the record he beat was set recently.
In order for me to benefit from this, I have to keep my #1 by beating the new score, getting even more pp than the guy who beat my old score.

pp is gained by maintaining high ranking spots instead of getting as many as possible. The more ranks you sustain, the better your pp will be, but if you lose a rank on one of your maps, your pp goes down.

It's also interesting to note that there is currently a limited number of pp to go around because each ranked map has 40 spots for pp, and there is technically a limited number of maps to rank on.

At least this is what I think about it. ._.

the biggest flaw i find in pp is that when someone gets a rank #1 with lets say 96% acc DT, and someone else gets 100% DT except for 1 sliderbreak in the middle (leaving him out of top 500) the first person will probably get a ton of pp while the 2nd wont get a single point, therefor still making score an influence on pp.

Score does influence pp, but it does not influence the rankings as much as it did before. If you score higher than other people in maps, you get points. It is no longer "Who played the most maps wins". With the old scoring system, people would rather get a #100 on every ranked map in the game than have #1 on a reasonable amount of them simply because working for the extra 10k score in a map was not worth it compared to doing new maps and getting 30m points and boosting your rank a ton.

Winshley wrote:

MillhioreF wrote:

losing a #1 hurts a lot too but i don't think you have many of those xD

For my case:
- Initial pp was around 4,700pp (forgot what it was, exactly)
- Played Klaus Badelt - He's a Pirate [Hard] and scored rank #1 with only HD+FL, Performance jumped up to 5,093pp
- Played some random maps until having Performance around 5,105pp
- rockleejkooo took my #1 on the said map above, Performance dropped heavily down to 5,009pp D: (you can see a noticeable jump-down spike at my performance graph lol)
- I re-took #1 again on the map with HD+HR+FL, and Performance raised up to 5,134pp

@Wishy22: That beatmap I mentioned above was ranked since 2008.

lol that map

fuck freedom dive, klaus badelt he's a pirate is the shit!

Nice pp spike there, silmarilen

EDIT: I think I must be prepared to play the map again, because...

have fun getting SS with dt

but dat OD.

I can SS OD5+HR+DT (as noted on some of my Top Ranks), the hardest being this. So don't be surprised if I managed to pull this off~

On another note, JappyBabes killed my 107pp.

If only I could play FL. I'd destroy you all! ;_;

kriers wrote:

If only I could play FL. I'd destroy you all! ;_;

Go learn from azuraer sensei~

kriers wrote:

If only I could play FL. I'd destroy you all! ;_;

i cant FL either, but this map is just incredibly easy

ok I got no pp from http://osu.ppy.sh/b/699

what

the

fuck

Jordan wrote:

ok I got no pp from http://osu.ppy.sh/b/699

what

the

fuck

Shut up noob.

Soly wrote:

Jordan wrote:

ok I got no pp from http://osu.ppy.sh/b/699

what

the

fuck

Shut up noob.

k

now seriously i don't understand that system I got a lot of pp from easy insanes and no pp from these things ;_;

If you've already given it time to update, then it might be that you didn't outdo anyone significant? It won't give you a lot of PP, if all the good players are above you in score on the map. I was thinking that this might explain why some people report getting a lot of PP from top 40ing Hards. It is simply easier to top 40 a hard, so you have a better chance of outdoing some really skilled player who didn't care enough to get perfect accuracy on the map, because it was incredibly easy for him. That's just a thought.

G0r wrote:

If you've already given it time to update, then it might be that you didn't outdo anyone significant? It won't give you a lot of PP, if all the good players are above you in score on the map. I was thinking that this might explain why some people report getting a lot of PP from top 40ing Hards. It is simply easier to top 40 a hard, so you have a better chance of outdoing some really skilled player who didn't care enough to get perfect accuracy on the map, because it was incredibly easy for him. That's just a thought.

he got rank 3, topping players like cookiezi, doomsday, reisen udongein or shaggon.

G0r wrote:

If you've already given it time to update, then it might be that you didn't outdo anyone significant? It won't give you a lot of PP, if all the good players are above you in score on the map. I was thinking that this might explain why some people report getting a lot of PP from top 40ing Hards. It is simply easier to top 40 a hard, so you have a better chance of outdoing some really skilled player who didn't care enough to get perfect accuracy on the map, because it was incredibly easy for him. That's just a thought.

So you're basically saying the pp system is ridiculous. You gain more for playing easier what a shame.

I think the huge contrast between Jordan's hhr score and WW's hhr score might have something to do with it.

If it had been the only hhr score, maybe he'd earn some pp. I'm only speculating.

You're forgetting the fact that other people's playing has an impact on your pp

for example if you have a #4 and someone else takes that #4, you will lose pp

Also the fact that it's 95% acc... js... at least for me, I get 0pp no almost any map I get <97% acc on

Edit: I'm completely fine with this btw.

I got like 1-2 pp for all of this together: (Well... solid state squad wasn't surprising though. DT FC with 89% shouldn't be worth anything anyways.)


Well... happened often. Stopped questioning the system after all... some ranks also give incredibly huge boosts of pp. Just play and you will eventually stumble upon them. :>
Xlasher not giving anything is strange though.


In my opinion pp shouldn't take accuracy or mods into account at all. We already got a scoring system which scales with them, so just count scores as they are. There is absolutely no need to double dip modifiers. Especially not with them being hidden from us. Rank 5 shouldn't be worth more than rank 3 for instance, even if rank 5 got a higher accuracy.

Tom94 wrote:

In my opinion pp shouldn't take accuracy or mods into account at all. We already got a scoring system which scales with them, so just count scores as they are. There is absolutely no need to double dip modifiers. Especially not with them being hidden from us. Rank 5 shouldn't be worth more than rank 3 for instance, even if rank 5 got a higher accuracy.

That's the major problem with PP. It's completely hidden from us in favor of not being farmed. If it's really all that good at judging skill... then why do we need to hide how it does this to keep people from farming it? It shouldn't be farmable in the first place. Why the cloak and dagger?

^I hadn't thought of it like that before, that's an interesting point. Peppy should be able to release all the information on the pp system if it truly isn't farmable. This would also shut a lot of people up about it's ways (finally).

Edit: I disagree with Tom94's point about accuracy though. A #5 DT HD 98% is < #3 DT HD 97% ... I don't think so. just no. Accuracy > all imo.

YodaSnipe wrote:

^I hadn't thought of it like that before, that's an interesting point. Peppy should be able to release all the information on the pp system if it truly isn't farmable. This would also shut a lot of people up about it's ways (finally).

Edit: I disagree with Tom94's point about accuracy though. A #5 DT HD 98% is < #3 DT HD 97% ... I don't think so. just no. Accuracy > all imo.

I agree on your point that a #5 DT HD 98% is < #3 DT HD 97% is just wrong.
However this is a problem with the scoring system. If PP would take only scores into account, fixing the scoring system would also fix PP in consequence.

Also: What if the 100s were right at the beginning and one didn't want to sacrifice another retry? I often got higher ranks with lower accuracy just because I was too lazy to restart after getting 100s on the first 3 hitcircles due to wrong guessing the first beat with HD.
In those cases I still would value the 97% run more.

Hey guys I'm #10 or #11, that should be enough proof that this system takes accuracy into account a bit too much