AR11

posted
Total Posts
59
show more
cheezstik

Narrill wrote:

buny wrote:

I heard 52% of statistics are made on the spot.
I am the 48%. I double checked, and average human reaction time is actually closer to 215ms, so 75ms to interpret patterns. Still, that's not a lot of time.
As impossible as it might seem, there's that guy that can chop a bb gun pellet in half with a sword as it is flying at him, crazy shit like that makes me think that eventually, years from now, someone will be able to read ar11. Maybe it will even become how ar10 is now, and the new setting to compete over will be cs :O.
Dexus
The problem here is people think they have to react to high approach rates.

To expand on what I'm saying, your eyes are slow and there's latency in computers. If your audio settings are synced properly you can simply predict notes. If a mapper has done a good job on the map then the player can hit all the notes consistently. Problem is lots of mappers put stray shit everywhere and don't know what the hell they're doing. There's lots of issues with consistency in mapping. Players/mappers in this generation favor random spurt patterns though because the focus is on it being a challenge.

Yeah, it's about memorizing if you want to do super low object density.
RaneFire

Narrill wrote:

buny wrote:

I heard 52% of statistics are made on the spot.
I am the 48%. I double checked, and average human reaction time is actually closer to 215ms, so 75ms to interpret patterns. Still, that's not a lot of time.
I agree with that average amongst gamers tbh. When humanbenchmark.com was only known to a smaller population nearer 7 years ago, the average was 215ms for quite a long time. It's only as more users (read: untrained noobs) have started "testing" their reflexes, that the average has risen over the years. They still list the median as 215, even though clearly 270ms has more hits.

But a reaction test doesn't necessarily mean you have X amount of time to spare for reading either. You don't read with your fingers, you read with your eyes and brain, and those are much closer together than your brain is to your fingers. Sending the message is all that is done, whereas a reaction test requires the recognition aspect as well as sending the signal. Reading actually fits in the middle of all that, not the end or beginning.

With training you can drop your reaction times quite significantly, since upon first playing osu!, I tested at 230ms, but over the course of playing my reflexes have dropped down to 190ms average (170 low) and stayed there for the past 2 years.

A rule of thumb, I like to apply, is that you can eventually, with practice, read as fast as double your trained average reaction time.

It's unfair to compare an unpracticed reaction test vs a game so fervently practiced, where your reactions, attention and focus are truly at their best. However, the processing of information (recognising objects) is what increases that value a lot. Reacting to a huge coloured area the size of your screen is quite easy by comparison.

Dexus wrote:

The problem here is people think they have to react to high approach rates.

To expand on what I'm saying, your eyes are slow and there's latency in computers. If your audio settings are synced properly you can simply predict notes. If a mapper has done a good job on the map then the player can hit all the notes consistently. Problem is lots of mappers put stray shit everywhere and don't know what the hell they're doing. There's lots of issues with consistency in mapping. Players/mappers in this generation favor random spurt patterns though because the focus is on it being a challenge.

Yeah, it's about memorizing if you want to do super low object density.
Sure, but reaction speed sets the baseline.

We're also talking about reading, not prediction. Prediction is the main reason that easier maps get chosen for the majority of AR11 scores. Tricky parts are memorised, the easy parts are sight-read. Very few people enjoy complete memorisation, akin to FL.
Dexus
That is right but if you're playing osu! and don't suck your reaction speed is no slower than 4 tenths of a second. Usually you'd be sitting at 3 tenths or lower if you've properly trained on AR10, 10.3, 10.8

Here's an interesting article if anyone wants to read http://www.higher-faster-sports.com/reactiontime.html
RaneFire
Unfortunately, reaction time is highly genetic and typically can only improve 10 to 20%. Two things that help improve it are stimulants and overspeed training.
Well my reactions improved 20% already, so I guess that's as good as I'm going to get.
B1rd
that article has a whole lot of depressing stuff about natural born genetic advantages.
Dexus
One way to apply overspeed training is by watching replays of AR10 and hitting the 2x speed button (speed modify it first to be slower so it's played at a realistic pace). You would essentially be playing relax without the feedback but the speed is faster than ar11. Doing that several hours a day on and off could possibly yield results if anyone is willing to try it.

I personally have not tried to train my reaction times, but when I'm lazy I get 300ms, when I focus I get 220 to 230ms, and then when I drink coffee I get 185 to 190ms reaction times.
ZenithPhantasm
Like everything in life some people are just better at certain things. Nothing to get depressed about.
B1rd

ZenithPhantasm wrote:

Like everything in life some people are just better at certain things. Nothing to get depressed about.
it is.

Dexus wrote:

One way to apply overspeed training is by watching replays of AR10 and hitting the 2x speed button (speed modify it first to be slower so it's played at a realistic pace). You would essentially be playing relax without the feedback but the speed is faster than ar11. Doing that several hours a day on and off could possibly yield results if anyone is willing to try it.

I personally have not tried to train my reaction times, but when I'm lazy I get 300ms, when I focus I get 220 to 230ms, and then when I drink coffee I get 185 to 190ms reaction times.
I don't think you have to go to these extremes, training 10.3 or even 10.1 would be enough to do this 'overspeed training' I think.
FlyingKebab
I hate coffee with all my heart and soul, TEA FOR LIFE MEN. However, auto mode sped up replays are interesting but this would require a lot of patience to pull off.
Topic Starter
Pengua
Well, there's been a lot of discussion about AR11 in general but besides raw memorization, it seems like the only thing I can do is train reaction time.
ZenithPhantasm

FlyingKebab wrote:

I hate coffee with all my heart and soul, TEA FOR LIFE MEN. However, auto mode sped up replays are interesting but this would require a lot of patience to pull off.
Im addicted to the boba milk tea. It too good >_<
nrl

RaneFire wrote:

I agree with that average amongst gamers tbh. When humanbenchmark.com was only known to a smaller population nearer 7 years ago, the average was 215ms for quite a long time. It's only as more users (read: untrained noobs) have started "testing" their reflexes, that the average has risen over the years. They still list the median as 215, even though clearly 270ms has more hits.

But a reaction test doesn't necessarily mean you have X amount of time to spare for reading either. You don't read with your fingers, you read with your eyes and brain, and those are much closer together than your brain is to your fingers. Sending the message is all that is done, whereas a reaction test requires the recognition aspect as well as sending the signal. Reading actually fits in the middle of all that, not the end or beginning.

With training you can drop your reaction times quite significantly, since upon first playing osu!, I tested at 230ms, but over the course of playing my reflexes have dropped down to 190ms average (170 low) and stayed there for the past 2 years.

A rule of thumb, I like to apply, is that you can eventually, with practice, read as fast as double your trained average reaction time.

It's unfair to compare an unpracticed reaction test vs a game so fervently practiced, where your reactions, attention and focus are truly at their best. However, the processing of information (recognising objects) is what increases that value a lot. Reacting to a huge coloured area the size of your screen is quite easy by comparison.
I'm aware of all that, the 75ms thing is just a rudimentary estimate I throw around to remind people that we're literally talking about hundredths of a second here. The best of the best (meaning less than the top 1% of the playerbase, give or take) can read 10.3, and 10.3 is a full 100ms slower than AR11. That's a monumental difference, and while it may be misguided to question whether it's at all possible to sight-read at such a speed, we absolutely should question whether it's feasible for a player with near-average reaction times to reach a level where they could sight-read a map of average complexity at ar11. That's the tone this conversation should have.
pooptartsonas
The thing is, AR10.3 is 400ms whereas AR11 is 300ms. That's a pretty big difference, and assuming your reaction time is roughly 200ms, then on AR11 you have 100ms to react to the note and 200ms to react on AR0.3. That's a MASSIVE difference. So when people think "hey, AR10 used to be fast to the general population, doesn't that mean eventually AR11 will be the same?" It just doesn't work. Imagine if AR12 was a thing. It would literally be impossible to sightread it, because you can't react until AFTER THE NOTE HAS PASSED. Therefore, there has to be some line where sightreading simply becomes unfeasable for those of us without god-tier reaction times, and I assert that line is somewhere between AR10.3 and AR11. Processing the note and moving to it within 100ms is just too hard without knowing the map quite well, even on maps without long jumps. Yes, followpoints do slightly help this 100ms figure, but it has no effect on new combos, and is still super fast regardless.
Dexus

B1rd wrote:

I don't think you have to go to these extremes, training 10.3 or even 10.1 would be enough to do this 'overspeed training' I think.
... but I can already read up to AR10.6.

seeing the fastest possible approach rate with the lowest note density is prime training for reaction, but as poop said there's a physical limit to this stuff.
buny

pooptartsonas wrote:

Yes, followpoints do slightly help this 100ms figure, but it has no effect on new combos, and is still super fast regardless.
That's where you're wrong.
Follow points appear a lot sooner than you think, and it's not hard to react on that one beat on a new combo.
All I said was that the population before said that AR10 was beyond human. And before that, it was AR9. Of course the population that could read the standard AR dropped at each step.

And why are you guys making such a fuss over reaction times? Do you play osu simply by reaction to every circle? There's more to a map than just "OMG NEXT CIRCLE THER".

Reaction times faster than 200ms is much more common in the competitive gaming community than the general population, so you guys are just basing your claims over irrelevant info, and what you should assume is that the player has faster reaction time than just the "average" because this improves with more training.

edit: the point of the topic is not that AR11 would become the standard or not, but if AR11 is sightreadable. The physical limit is not between ar10.3-11 and that's a fact.
nrl

buny wrote:

And why are you guys making such a fuss over reaction times? Do you play osu simply by reaction to every circle? There's more to a map than just "OMG NEXT CIRCLE THER".
You've got this backwards. It would actually be much easier if osu! were purely reactionary.

buny wrote:

edit: the point of the topic is not that AR11 would become the standard or not, but if AR11 is sightreadable. The physical limit is not between ar10.3-11 and that's a fact.
I don't think this was ever in question. As I said, the point of the topic should be whether it's reasonable to expect ar11 sight-reading to occur with regularity outside the top .01% of the playerbase at any point in the future. The question isn't whether the absolute physical limit lies below ar11, it's whether that limit lies low enough for a lower caliber of player to reach.
buny
Well, a lot of posts like

B1rd wrote:

who can? is there anyone who can sight read FC a decently hard song at AR11?

winber1 wrote:

pretty sure the supposed ar11 have even said themselves that they pretty much memorize the map. thus no one can actually read it, but many people are good at it. personally i would not say i cannot read ar 10.3 at all, but i can still get fc's with it.

TUOPlays wrote:

ya know if you play a map a lot/memorize it that's when higher ARs become doable. I literally learned AR8 DT by playing a map i knew really well constantly, and it made the AR seem slower than it actually is. It requires a lot of memorization, but not as much as FL for example
imply that it isn't possible for anybody to sight read AR11, and I'm simply disagreeing with it.

Will it become the standard for top tier players? Probably not, since it is unnecessary to go anything above 10 really.
Is it possible for a single person to read AR11? Yes.
chainpullz
I love how buny is fighting a one man war over here lol. With that said, coffee is love, coffee is life. I hate the taste by the way.
Please sign in to reply.

New reply