[added] Aim Accuracy Indicator - For fun~ · forum

Kitokofox

osu! Alumni

Joined August 2012

Topic Starter

Kitokofox 2013-01-09T10:14:56+00:00

I thought of this one day, when I tried challenging myself to aim inside the hit circles as good as I could

For Standard: The accuracy for each hit is defined by how close in the middle of the hit circle you hit. If you hit it in the dead center, you get 100%. If you miss the hit circle, you get 0%. If you hit it half a radius from the center, you get 50%.

For Catch The Beat: The accuracy for each his is determined by how close you caught the fruit in the middle of it. Catching it directly in the middle gets you 100%, missing a fruit gives 0%, while catching it halfway between missing the fruit and catching the fruit in the middle earns 50%.

Taiko and osu!mania can't have this feature because no aiming is involved in both of these modes.

All the aim accuracies for each note is then averaged for your aim score.

To enable this feature would be turning it on in the Fun Spoilers setting. It has no impact on the game, just meant to be fun or help you practice your aim.

RaneFire

1,181 posts

Joined May 2011

RaneFire 2013-01-09T12:40:29+00:00

Aiming is pretty important in osu! standard, sometimes I wonder myself just how accurate I "think" I am with regards to aiming my cursor. I try to give myself a mediocre indication by just saving difficulties anew as CS-7.

Although the game is first and foremost about rhythm, aim is becoming an evermore important part of the game (seen in the latest "challenge" maps) that separates the best from the rest. It would be very interesting to see some kind of "cursor aim indicator" even though just hitting the circle is all that matters.

But I admit I am also interested in seeing what other people's statistics would be as well.

If this were implemented I would suggest using the respective circle sizes as a template for cursor aim accuracy calculation (even up to and including CS-10, although no longer mappable, I'm sure osu! still knows what they are). Scoring would show number of 10's, 9's, 8's, 7's, 6's, 5's, 4's, 3's (largest iirc), and create an accuracy index from these, since the circle sizes are what is relevant and not the "exact" distance from the center of the hit circle.

chaotic_iak

287 posts

Joined February 2012

chaotic_iak 2013-01-09T13:38:19+00:00

Just saying, CTB's accuracy for sufficiently hard maps will be at most 80% due to edge cases (catch a fruit at the far left edge of the plate in order to be able to catch the next fruit at the far right edge of the plate, otherwise you'll miss some of them).

But for osu!std, I think it's good. Uploading them as a new stat might be messy though (what about the existing records?), so probably a just-for-fun stat.

Or otherwise I think "average distance from center" is pretty good too, so you know that if this value is high then you must practice aiming in order to play high-CS maps.

JFRN

233 posts

Joined May 2011

JFRN 2013-01-09T15:47:34+00:00

maybe, but this is hard to implement, i'm indiferent with this.

jemhuntr

933 posts

Joined February 2012

jemhuntr 2013-01-09T15:58:20+00:00

lol actually it's fun aiming at the edge of the circles without missing it. this would be nice. although it's not necessary. and after I use it for like, a week, I'll never use it again.

TheVileOne

osu! Alumni

8,431 posts

Joined March 2010

TheVileOne 2013-01-09T16:56:05+00:00

It is not hard to implement.

I support this.

theowest

osu! Alumni

6,891 posts

Joined January 2009

theowest 2013-01-09T17:56:10+00:00

Aim accuracy. I like it.
I don't think this should be ranked at all, but if so it would be a nice profile stat.

Saten

Global Moderator

3,866 posts

Joined August 2010

Saten 2013-01-09T20:29:11+00:00

iPanic

Kitokofox

osu! Alumni

455 posts

Joined August 2012

Topic Starter

Kitokofox 2013-01-13T07:42:37+00:00

theowest wrote:
Aim accuracy. I like it.
I don't think this should be ranked at all, but if so it would be a nice profile stat.

Profile stat? Nnnnyyyeeehh... Sketchy. The main point of the game is following the rhythm, but knowing how well you click on those circles is a fun stat to have by your side too. It's not hard to implement, and it's not too hard not to play with too. But yes, it'll remain unranked mostly.

chaotic_iak wrote:
Just saying, CTB's accuracy for sufficiently hard maps will be at most 80% due to edge cases (catch a fruit at the far left edge of the plate in order to be able to catch the next fruit at the far right edge of the plate, otherwise you'll miss some of them).

That, I don't think many players would mind. The point of this accuracy thing is to try and maximize the aim accuracy you can get, is it not?

Would make a lot of fun practice to tackle hardrock maps, too, on both modes.

Heibel

812 posts

Joined July 2011

Heibel 2013-01-14T10:54:47+00:00

As long as it is not going to affect the rank one might get, and it is only for fun to see how in the middle hardcore accurate you are, then i am all for it.

ieb

165 posts

Joined August 2012

ieb 2013-01-14T11:11:47+00:00

That's nice.

Wishy

2,974 posts

Joined September 2010

Wishy 2013-01-15T00:13:25+00:00

This would be quite fun to have.

Primula_old

Primula_old 2013-01-15T04:16:18+00:00

This would make for a pretty fun multiplayer mode.

Ephemeral

Inland Empire

3,945 posts

Joined April 2009

Ephemeral 2013-01-15T14:22:04+00:00

TheVileOne wrote:
It is not hard to implement.

I support this.

I don't see how you could possibly know, please refrain from making grand assumptions about the difficulty of implementing a particular feature - some require a lot more effort than you'd think.

TheVileOne

osu! Alumni

8,431 posts

Joined March 2010

TheVileOne 2013-01-15T15:48:49+00:00

Ephemeral wrote:
TheVileOne wrote:
It is not hard to implement.

I support this.
I don't see how you could possibly know, please refrain from making grand assumptions about the difficulty of implementing a particular feature - some require a lot more effort than you'd think.

I guess it depends on your frame of reference and approach. The code is already implemented for fixed objects (taiko and mania), it would be more complex to code it for dynamic objects, and multiple angles, and so forth. And it would be more or less harder to do in an efficient manner. It wouldn't be unreasonably difficult to compare screen/ grid positions in real time. Notes are already given coordinates, and the game can already track coordinate movement of the cursor.

Upon hit, compare cursorPosition to notePosition in buffer. Compare X position of cursor to notePosition in the buffer. Compare Y position in the buffer. From this information, calculate pixel difference from center of note. Next part would be a bit trickier. If you want complete 1/100th degree of accuracy, you would take note's diameter based on circle size, and compare it to the pixel difference.

Lets say the note is 20 pixels wide. The pixel difference is 10 pixels. 10/20 = 0.5 = 50% accuracy. I guess the note's actual width is unimportant compared to the size of the hit window.

It's probably not very efficient to do it the simple maths way. Bring on complex bit comparisons.... It can be as complex or as easy as one wants it to be.

Oh and there will be bugs. On a complexity scale, this seems to me to be a 6 where 10 would be impossible to implement. 4 if it's inefficient and buggy.

Ephemeral

Inland Empire

3,945 posts

Joined April 2009

Ephemeral 2013-01-19T14:11:28+00:00

i reiterate, unless you have access to the source code, i don't see how you could possibly know. extrapolating in such circumstances is a waste of your time.

i've had an idea for an unranked "flourish" mod that would work on a similar premise, so I may look at implementing something similar at some point in the future - don't wait for it, though.

eat_now

1 post

Joined November 2009

eat_now 2013-03-07T05:11:16+00:00

To be clear, "accuracy" and "precision" below refers to the distance between the cursor and the hitcircle when the key/mousebutton is clicked, and not timing.

As far as I am aware, osu! replays aren't terribly informative about how precisely or accurately a player is clicking. We get to know if we are in the circle, on the edge of the circle or not in the circle at all, but that's about it(Correct me if I am wrong). I enjoy jump maps the most, so I felt it would be nice to have such feedback, which would answer such questions:

-How precise am I when I click? (variance of clicks)
-How does my precision change when jumps increase in distance and speed?
-When I change my mouse or tablet, is my precision increasing or decreasing? (Even if I am getting similar scores)
-Does my precision get affected by sleepiness?
-In what other ways can a player's performance be graphically represented?

The closest thing I've found to this is this feature request, but I have a feeling it won't be implemented in the near future.
http://osu.ppy.sh/forum/t/113594&start=0

Statistics can also be obtained outside of game via simple click-logging and comparing the output to the original beatmaps. In fact I am sufficiently interested in this to have started on a script, but it occurred to me that someone else could already have a working utility that does this. Just asking before I continue to reinvent the wheel.

enik

276 posts

Joined December 2011

enik 2013-03-07T06:57:14+00:00

eat_now wrote:
The closest thing I've found to this is this feature request, but I have a feeling it won't be implemented in the near future.
t/113594&start=0

I don't think it would be any fun nor useful. Full combo is already enough to say how precise you are, for more challenge we have small circles and HR (which in fact many people hate). It would exclude any flow/dance cursor actions because you must aim straight to the circle centre. Just another reason to get frustrated and depressed because you suck in this game.

Saten

Global Moderator

3,866 posts

Joined August 2010

Saten 2013-03-07T12:41:00+00:00

That is something I don't really wanna know. I mean I don't necessarily aim for the center of a hitcircle when I play.
Just my thoughts~

MillhioreF

Pro Tester

5,004 posts

Joined July 2011

MillhioreF 2013-03-07T15:36:33+00:00

peppy actually plans to add something like this sometime in the future, after redesigning the results screen.

Also, this is a feature request, so moved accordingly.

TheVileOne

osu! Alumni

8,431 posts

Joined March 2010

TheVileOne 2013-03-07T17:44:17+00:00

Merged this topic, because it is the exact same thing.

Winshley

osu! Alumni

2,384 posts

Joined May 2011

Winshley 2013-03-08T07:50:34+00:00

If the statistics is implemented in similar way as Hit Error and Unstable Rate, then the Replay data glitch (where you managed to barely hit the note at the edge of the circle during gameplay and showing as a miss at Replay data) might need to be fixed.

Ookami

19 posts

Joined January 2011

Ookami 2013-12-22T22:05:48+00:00

I've seen this suggested many times, and I'm not entirely sure why it's been shot down. I believe this would be a good idea. I personally think that it needs no score multiplier as it would simply be a tool to increase one's overall performance at Osu! I don't see how it's implementation could harm anything, Hope this gets somewhere. It's somewhat sad how this was suggested 5 years ago and completely shot down. >_>

Pengua

41 posts

Joined April 2013

Pengua 2013-12-23T23:20:42+00:00

I was just thinking of this, and instead of giving an accuracy number, make it a heatmap (see example image), including misses (to a certain degree). Of course, you can always add in a number for accuracy.