This post is TL;DR. Read the section titles and only refer to some parts of it if not everything interests you.
Related to the list of top PP-giving maps:
In an ideal osu!, star ratings would show the true difficulty of maps.
In an ideal osu!, the hardest maps would give the highest rating for the rankings.
In an ideal osu!, competitive players would know which maps are the hardest, would be able to easily find them with search filters, and play them to rank the best.
Revealing which maps give the most pp obviously leads players to try them out, but that is not wrong - that's what SHOULD happen. Like BD said: the whole point of online ranking is the competitiveness, and a competitive player would naturally strive to play the most competed maps. While I think that those maps should be the hardest ones in their raw level, PP judges this by their "contestedness" level which can be more or less ok too - but that means showing which maps are the most contested should be public information, just like showing which maps are the hardest. Hiding the list of top-pp maps shows distrust in the quality of the pp system;
If you have trust in pp, there would should be no problem with revealing the top pp-giving maps, because it means that playing just those map specifically would not make players rate higher than usual - since the way they would rank on those maps would be consistent with their skill in the way they played until now, and thus stay as it always was. (i.e. they'd get lower ranks on top-pp listed maps, compared to lower maps on the list or maps outside the list where it would be easier to get higher ranks.)
Hiding the list means that you believe some maps on the top-pp list are actually getting higher pp worth than they actually deserve - which means the pp system is flawed if it really is the case.
I was really impressed with peppy first choosing to reveal the list, because it meant he really trusts the pp system, but now that it was hidden it seems like even the creator himself doesn't trust the accuracy of his system. Disappointing.
Putting aside lowdiffs that showed on the list, even stuff like Blue Night being higher than Chipscape looks like nonsense; Blue Night is obviously more popular because it's much much easier, and I'm sure I can FC with enough retries until I stop getting random miss on stack, while chipscape is hard to even just pass no matter how much you play it. Earthquake Super Shock 0108 is also a pretty easy map to FC and you could add Hidden on top of that no problem, but on since it's on a very popular mapset it rates higher than many other harder maps. And yet, I'm sure most of the people who played the 0108 of supershock didn't FC it often, which means the contestedness level of it is very high without really being very truly contested. The top 50 for it is nomod S with 99.33%... anyone can pass this rank with Hidden.
---
@ response to winber1, [CSGA]Ar3sgice, boat (from 2 pages ago):
Player B is worse than Player A, because he does less to try to rank well on the competition system. So he deserves a lower score (and we already have a scoring system). However PP is supposed to rate skill, not effort, right? While Player B may not be as dedicated as Player A, if they both played a new map that just came out, who do you think would do better on it? That's what I meant by "more skilled", which is different.
PP claims to rate players by skill, not by effort. This sort of rating can be useful; for example, if you're hosting a multiplayer room, you will play maps that require a certain skill, so knowing your own skill as well as others' is a good way to find a multiplayer room that will be fun for you to play in. But if PP rates skill incorrectly and instead rates your desire to have high PP then its existence is pretty much pointless isn't it?
"...to rank people by using one song..." - the thing is, I actually meant the opposite. This example was to show that players are going to put different effort levels into different maps. Player A might pour a lot of effort into one group of maps, which player B would pour a lot of effort into other groups of maps. They can both beat each other and lose to each other on the rankings based on how much effort they put into each map they played. However with the current system, the player who was lucky to play more popular maps would end up ranking higher - and this would be independent of skill, it would boil down to whether you were lucky to choose to play popular maps. This is the problem with a system that contests based on map popularity instead of map difficulty.
" I don't think someone who can only do 89% can magically get a SS even if playing for 3-5 hours" this is actually not uncommon in low levels of play :p high level players can adapt quickly, but low level players take longer to adapt. For example I usually play 170-190 bpm maps, if I'd play a 140bpm map now I'd get very poor accuracy on it. But if I'd spend 1-2 hours playing 140bpm maps, I'd probably be able to gold SS it. I'm sure a lot of you can find something similar to this on your plays (maybe jumps, spacing, reading patterns on a specific map, etc. retries on one map can specifically help you for that map without helping your general skill except for other maps similar to that one).
---
In response to concern of low-pp-rated maps getting less attention because of publication of the list:
This is actually still the same as what I said earlier. Players will still play whatever they like the most. And if a player likes just playing for fun, knowing the top-contested maps won't make him play just them if he doesn't like them.
Meanwhile, if a player only plays to get the highest rank the fastest, then right now he'll already check top ranks of top players, as well as go over maps and check their top50 ranks, and then intentionally only play the stuff where he knows it's the easiest for him to rank the highest for the most points.
So this changes nothing. Knowing the top pp-giving list won't make any maps be more popular or less popular, players will still play like they always have. It will only make PP-farming easier, but if PP really isn't farmable like it intends to be then this wouldn't be possible. If revealing the list actually allows people to farm PP easier, then this proves PP weightings are not functioning correctly. And wasn't this the point of revealing the list? To let players check the maps on it and report what is weighted higher or lower than it should be, to provide useful feedback and will help tune the PP system to be more accurate. I really don't understand why was this list hidden.
Are you worried that some players would suddenly get higher PP than they deserve due to revealing this? If this is even a concern, it means clear distrust in the quality of the PP system. Are you worried that players would complain about info on the list and call PP bad? afaik PP is considered to still be in beta stages of development and is not final, so just ignore those complaints and use the output from them as feedback to improve the system.
---
In renpose to bwross:
"Player C grinded a map on an offline machine, to get a good accuracy on it on his first try online. This is why amount of retries should not be taken into account".
My response to this is a little complicated, so read carefully:
1. If the map is easy enough to SS consistently once you practice it enough, then you will most likely be able to SS consistency every single map on that level on your first try once you learn to do it for that map. Keep reading.
2. Some maps are so hard that, even if you can SS them sometimes, you won't be able to do it every time. Even if you played that map 1000 times offline and had several SS ranks on it offline, you cannot guarantee that on your first try online you'd get an SS, and if you're not good enough to SS it reliably every time then it most likely won't happen.
3. Grinding a map until you get really good at it will make you about as good on all maps of that same level. Like I stated earlier with the 140bpm example, I really suck at 140bpm right now, but from past experience if I practice this specific range for a while it's possible to get good accuracy on it consistently on all maps that have that bpm.
Now lets combine 1 & 2 & 3. The concern is: players hiding their skill, and getting rewarded higher than they deserve for high accuracy on maps due to offline retries.
The system I suggested takes into account your accuracy on every try that you do. It doesn't just go "oh hey, this player got 99.5%, and it took him 20 tries.". It goes "the player played x, had 99%, give him +/- y points. the player played x (again), had 98.5%, give him +/- y points." etc, per retry. If you actually got good enough to play a certain level with reliable accuracy on every try, allowing you to get that level of accuracy on your first try online, then you're actually good enough to do it and you deserve the rating you get. Had you done those 1000 retries online, your rating would slowly go up over time until reaching your level, while by doing it offline your rating just suddenly spikes upwards, but you end up on the same spot.
