forum

Rank system. [invalid]

posted
Total Posts
12
This is a feature request. Feature requests can be voted up by supporters.
Current Priority: +0
Topic Starter
[deleted user]
Well, first I have to say that currently rankexpres bad the players who are good. The reason is that the most scores are ranked better. But the ranking score not determines the best person playing. If not the person who has won more points, which can be done just by playing many songs.

I found an algorithm that could give the possibility to see just played really, without actually using the point system.

The algorithm works as follows:

When someone plays a song, have two sets of points: The rank score and Total score. The rank only collects the best scores of a song played, of all the difficulties. The total score contains all the points made in the game, including tries.

So we can say that the total score is much more accurate than the rank score.

On the other hand we have the plays. The show plays as we played. Given that each retry also scores points we can say that the more points in fewer attempts you play better.

Following we have the accuary. This marks the momentum you have and the certainty at the time of her notes. There are people who is better at the streams and there are others who can not see them. That scores.

Thinking is actually a bit easier.

Score per try scores. Accuary too. So?

Total Score / Plays accuary * / 100 000

The 100,000 is because the number is very large cause of the score.

So for example, many already know H1ko and Cyclone, one of the best players in Osu.

H1ko would have 1467 points.
Cyclone would have 1901 points.

Does this mean that cyclone is better than h1ko?

I do not know. That would require to test.

Me , for example, regardless of my rank, I'm very bad xD.

I have a 227 points. And other people for example with less rank than me have 400. This is because they are better.
So giving this idea, and this algorithm, perhaps to implement something new, like Player Point system or something. No removing the actually system but it will can be an implement.

I would like to give your opinion. Thanks for reading my bible xD.
Shohei Ohtani
Wait what the hell.

Your system pretty much is saying that people who play in multi and attempt to pass a map like, 50 times, is better than someone who generally plays in solitude and passes maps generally on the first try.

;~;.

There is really no way to tell who's the best until you actually know the person and see them play, and even then, it's all based on opinion.
Topic Starter
[deleted user]
I didnt said that. Read good.
Roga-don
So ... My Taiko score would be 2,30 ?

Wait, there is a problem with taiko mode and your Rank System :lol:

ED:No .... what's play accuracy for you ?
Playcount or Hit accuracy ? :lol:

ED2: Ok .. It's 204.4 for me on taiko ...

The formule is Total score / Playcount*Accuracy / 100 000

And it's not fair for the "not really good accuracy" people :?
Arusha Shuna

Roganis wrote:

And it's not fair for the "not really good accuracy" people :?
maybe OP thinks the one that has better acc is a better player
Roga-don
Ok ... So players that play only beginner maps are better players than players that play only Insane/Hard maps ? :lol:
Topic Starter
[deleted user]

Roganis wrote:

Ok ... So players that play only beginner maps are better players than players that play only Insane/Hard maps ? :lol:
Nop because more difficult more points xD, normally.

And the accuary is the accuary in game, rythm accuary.

Taiko have a different score system xD.
anonymous_old
Let me point out the fundamental flaw in your system:

Darkylight wrote:

Well, first I have to say that currently rankexpres bad the players who are good. ... [T]he ranking score not determines the best person playing.
Because skill is widely varied (some people are good at jumps, some people at streams, etc.), it is difficult to compare all players and to put them into a single ranking system ordered by "goodness" or "skill" or whatever. I'll go out and say it's impossible unless you feel like shoving in massive amounts of computer cycles to crunch the numbers of perfect learning formulae every time someone finishes a map.

There's also a flaw in your implementation. Using your system, ShaggoN comes out at 1060 points. ShaggoN is unarguably one of the best players in the game. That's just one example of why you shouldn't rely on metrics such as play count.
Sir Minelli
viewtopic.php?f=30&t=11138&st=0&sk=t&sd=a&hilit=rank

Posted that long time ago~

(Not sure why it went to completed requests hmm..)
Saten
Play Count doesn't favour anything in skills afaik


Also there's a new Rank System

Performance Points
http://osu.ppy.sh/wiki/Performance_Points
Emaal

Saten wrote:

Play Count doesn't favour anything in skills afaik


Also there's a new Rank System

Performance Points
http://osu.ppy.sh/wiki/Performance_Points
>2 years old

Go Saten!
deadbeat

Saten wrote:

Play Count doesn't favour anything in skills afaik


Also there's a new Rank System

Performance Points
http://osu.ppy.sh/wiki/Performance_Points
very true. and i think it works a lot better than w/e the op was suggesting. i'm sure no one will mind if i deny this~
Please sign in to reply.

New reply