On August 18 2014 09:04 Cyro wrote: I restarted this ~500 times, which was only actually like 1-2 hours iirc, to get fc with 0x100 on this part just as forced practice
I think I would go insane long before reaching an hour of retrying on something lol
I made a video recently (of mania) that I dont think I've posted here (or have I?):
I only ever replay something if I'm trying to playtest a map or something, so this was literally first try. Actually, this was the first time I played that diff on this computer, and I'd only played it maybe 3-5 times total before that, I believe.
Also, this is some of the best unstable rate I've gotten lately, and this map is full of chords (many keys at once) which are notoriously hard to keep good accuracy on:
So I found some of the replays from when I first started playing and they are fucking hilarious. Hard to believe I could've been that bad lol
Also, beat a couple songs today that I've been working on for a while. I feel like my timing has gotten a lot better after talking to Yorbon about how to approach things. Scores are going up across the board on songs I've already beaten too
Glad to hear you're improving falcon! Although i thought it was hard to pinpoint issues, apparently you found advice to be useful. If you have further questions, don't hesitate to ask
@ bobbias/cyro: I personally think you're mixing up two things. I think Bobbias is talking about repeating stuff to gain rank, or in other words, repetition for very short term gain, while cyro's talking about repetition for long term gain (practice). For rank purposes, much repetition is pretty lame, in the sense that it doesnt add much except pp, without having value in the long run. In my view, it isnt for nothing that my best scores all needed max 3 tries, not meaning that it was the 3rd time i played the map, but i played for a good score max 3 times in a row. Bobbias also mentioned the sentiment of wanting a better score. I can understand that sentiment; i, too, have a tendency to work to a better score if i feel like i should be scoring higher than i have. The intention is always to be able to play x, instead of to gain y pp at map x. Cyro's talking more about repetition for practice purposes. Depending on what technical goals you're trying to achieve, repeating map tries may be worth it. Especially by combining certain practice maps with some kind of specialized practice, trying maps again may provide some indication of improvement. As opposed to bobbias' case (i'm assuming) i'm not talking about continuous repetition, but repetition in intervals. I do think repetition has it's pros, but i'm very ciritical of preferring to play a map more than 5 (or something around that number) times above trying to find maps to improve yourself on specific aspects of the game, whatever those may be. To be honest specifically, i think repeating a map 500 times to get it SS is not worth it, even if it's only for practice purposes.
I hope it's a bit clear what i mean, cause im drunk writing. I may edit next morning.
I may actually just stick to easier maps and refine my play until I can comfortably A most songs 4.5 and under. Still not sure what the best way to improve in this game is, so we'll see how it goes.
On August 20 2014 10:00 Yorbon wrote: @ bobbias/cyro: I personally think you're mixing up two things. I think Bobbias is talking about repeating stuff to gain rank, or in other words, repetition for very short term gain, while cyro's talking about repetition for long term gain (practice). For rank purposes, much repetition is pretty lame, in the sense that it doesnt add much except pp, without having value in the long run. In my view, it isnt for nothing that my best scores all needed max 3 tries, not meaning that it was the 3rd time i played the map, but i played for a good score max 3 times in a row. Bobbias also mentioned the sentiment of wanting a better score. I can understand that sentiment; i, too, have a tendency to work to a better score if i feel like i should be scoring higher than i have. The intention is always to be able to play x, instead of to gain y pp at map x. Cyro's talking more about repetition for practice purposes. Depending on what technical goals you're trying to achieve, repeating map tries may be worth it. Especially by combining certain practice maps with some kind of specialized practice, trying maps again may provide some indication of improvement. As opposed to bobbias' case (i'm assuming) i'm not talking about continuous repetition, but repetition in intervals. I do think repetition has it's pros, but i'm very ciritical of preferring to play a map more than 5 (or something around that number) times above trying to find maps to improve yourself on specific aspects of the game, whatever those may be. To be honest specifically, i think repeating a map 500 times to get it SS is not worth it, even if it's only for practice purposes.
I hope it's a bit clear what i mean, cause im drunk writing. I may edit next morning.
You actually are right that we're talking about slightly different things, but at the same time, I also don't particularly agree with practicing things with heavy repetition either.
Originally I was just meaning to comment on how much of the osu community is obsessed over rank and will do things that don't make much sense to me. Specifically, I mean both repeating the same map over and over just to get a slightly better score as well as grinding out scores on maps that you don't particularly like just for the points. And I hope nobody here thinks I was trying to attack them if they happen to do these things. What right do I have to dictate how someone ought to enjoy a game anyway?
In any case, I personally don't enjoy direct repetition for any reason, practice or otherwise, and I'm also of the opinion that in general, training focusing on repetition of highly specific elements in isolated practice is not the right way to train for long term goals, based on some reading I've done.
I feel that the best way to practice any sort of pattern that gives you trouble is to practice it mixed in with everything else you play and in it's own context. What I mean by that is not intentionally playing 'practice maps' containing only that pattern, nor replaying the same map, but by finding multiple examples of the pattern in different maps and intentionally interleaving those into your standard play session slightly more often.
While it doesn't result in as immediate a result, the data seems to suggest that this is much better for overall retention. I'd say the phenomena is strikingly similar to cramming for a test. Sure, you can learn a lot in a short period of time when you cram, but you forget it all pretty quickly after, whereas things learned in a varied context tend to take longer to really learn, but stick much deeper in memory.
In terms of what difficulty practice should be focused on, I'd be interested to see if anyone here has seen any literature on something like that. I'd like to explain my approach to practice in terms of picking what difficulty of things I play. Before I start though I just want to point out that I don't pay attention to star difficulty. When I think of a map's difficulty, I think of it in terms of what kind of score I can get on it currently.
My approach has been to intentionally practice on things far above my ability to do well on. Many players seem to feel most comfortable playing songs they can A or S from what I've seen, but in my experience, while this may improve your ability to do well on that difficulty of maps, it will do comparably little to help you play harder maps. In my experience, once you can reliably A a map, you are generally capable of 'playing' that map, and improvement beyond that level is generally just refining your play, and refining your play doesn't lead to the same sort of improvements in skill that learning to deal with completely new challenges do.
I tend to practice on everything from maps I can S all the way to maps that I can barely reach 60% accuracy on (mind you, mania is different from standard in this regard). If you begin playing maps that you can't do well on (anything below an A really) at first it will probably feel intimidating because it's outside your comfort zone, but if you continue to practice on those things, you'll find you get used to it relatively quickly and the intimidation factor of playing something 'too hard' will lessen quite a bit.
I am tempted to use this sens, but i feel the added limits from being on lower sens (900 height instead of 800) and i should be able to learn more than enough accuracy on 800
So, I checked to see what my PP is at. turns out I'm above 3k in mania. 3,556 pp and relatively capped out in terms of hard mania specific maps lol. But it's a hell of a lot easier to break 3k pp in mania by the looks of your videos.
@ bobbias: I'll quote some parts of your wall of text and give my take on them. I have to say i didnt read the articles, though. Most probably due to my bad english, i think we're closer on practice views than our posts suggest. Throughout the post i'll assume we're talking about playing for practice purposes.
"I'm also of the opinion that in general, training focusing on repetition of highly specific elements in isolated practice is not the right way to train for long term goals" Completely agree. This is one of the reason i prefer maps with a constant high level above maps with level 'peaks' for a short amount of time.
"What I mean by that is not intentionally playing 'practice maps' containing only that pattern, nor replaying the same map, but by finding multiple examples of the pattern in different maps and intentionally interleaving those into your standard play session slightly more often." I'm not a 100% sure what is meant by 'containing only that pattern'. Personally, what i'd call a practice map is a map containing a high density of 1 certain aspect/pattern. Important in this is variation is that certain aspect. Repeating the same jump in a jump practice map does not make sense. Maps like "Blue stahli-let's go" do, because the jumps are varied enough. I don't mean one should play that map alone, I'm saying that incorporating it in your play is worth it, if you want to practice jumps. Within that play, imo one can repeat that map several times. Playing it too much causes memorization, which diminishes a lot of use for practice. Depending on map length and person, the amount varies. As I'd like to analyse my play and correct myself on it, most of the time ill play a map around 3 times, focussing on the most blatent errors.
"My approach has been to intentionally practice on things far above my ability to do well on. Many players seem to feel most comfortable playing songs they can A or S from what I've seen, but in my experience, while this may improve your ability to do well on that difficulty of maps, it will do comparably little to help you play harder maps. In my experience, once you can reliably A a map, you are generally capable of 'playing' that map, and improvement beyond that level is generally just refining your play, and refining your play doesn't lead to the same sort of improvements in skill that learning to deal with completely new challenges do." I do definitely agree with the first part. Getting used to much harder patterns generally lets you improve a lot, regardless of the difficulty you're playing at at that moment. However, i'd like to note something on practicing beyond A level. Once you get an A on a map, you're generally capable of playing a map. Your view that everything beyond A is 'just' refining your play seems risky to me. I feel refining one's play is a very important aspect of improving. In my view, extending your limits and after that refining your play leads to 'consolidation' of improvement. You're right that refining your play is no substitute for pushing you limit. But think it's important to note the other way around is true true as well: pushing your limits will not refine play per se.
And last, I play anything from just below 70% to close to ss-ing, in that order. Especially SS-ing and fcing have very little priority as the only thing you gain from doing it is consistency. An example of a map i think is worth practicing beyond A is for example Umetora - Ifuudoudou with dt. I have a few A's on it, but i'm not comfortable at all on it. There are some very nasty patterns, and it's ar10.3, making it worth it for me. It'll probably faze out when I can consistently get 95% though. By that time I'll probably be too comfortable playing ar 10.3 to play the easier 10.3 maps. And of course this map is not the only map i play 10.3, nor is it the only one with harder patterns i practice. I just want to be able to see, read and tap it without needing too much concentration, meaning i'm truly past the map.
Repetition (the original point) imo only make sense in this context. My mistake in the drunk post was calling the repeating of a map ánd playing the map again at another time, both repetition. I don't think the word repetition is used in the second sense, is it? But well, i hope not too many people have wasted their time on this, and i hope things are a bit more clear now.
The word repetition is appropriate for both cases, tbh. Also, you really should read the paper I posted. It's a peer reviewed study on learning.
I don't play standard any more, so I'm not familiar with the map you mentioned, but I think I get what you mean anyway.
I do want to point out that I can now regularly achieve scores in the high 90's (96/97/98) on quite a few of the easier mania maps that I used to be unable to score so well on before, and that is essentially refinement of skill resulting from pushing the top end of my skill and comfort zone. From what I've gathered refinement does actually happen while pushing your skill. However, the refinement is generally limited to the low end of difficulty where you could already do rather well, and doesn't generally bring your scores as high as focused training would.
EDIT: Lol better score than the one I just posted:
I read the first paper you posted, and it was interesting indeed. I don't think i read anything that makes me change my mind significantly, but it was good to read some feedback on how to improve practice at some specific parts.
Tbh, that refinement occurs when pushing your limits isn't a surprise at all. What's a refining skill for one map, is an essential one for another. There are however other factors (bpm and ar come to mind for example) that come into play, that correlate with the skill needed to play maps (at least in osu standard, i don't know how it works in mania). At some point your ability to play maps will decrease when you're pushing your limits further. And that's when i think playing for refinement is worth it for practice purposes. Your ability to play maps should be independent of bpm/ar (provided you're fast enough or are able to read good enough), which i think is impossible by just pushing your limits. As an example, there are quite some people who are unable to read ar9/8/7, because it's too slow. In my view, that's because of a lack of refinement practice. I'm not saying refinement is absolutely necessary. I just think it has some value.
I'm sorry i was using maps you didnt know. I thought they were old enough for you to know them, but apparently not. ):
should i just keep playing ar10 stuff? I'm at about ~88% accuracy atm on Stan SB - Dead +hr without actually messing up much stuff due to OD, just constant misreads on streams and patterns and that's way better than i used to get (i couldn't survive) but it seems laughable compared to basically SS on nomod ar9
It's stuff like on a triple for example, i open up hitting it with finger 1, then i hit with finger 1 again at half beat spacing - but it's actually a quarter beat triple, so end result looks something like:
On August 21 2014 22:07 Cyro wrote: should i just keep playing ar10 stuff? I'm at about ~88% accuracy atm on Stan SB - Dead +hr without actually messing up much stuff due to OD, just constant misreads on streams and patterns and that's way better than i used to get (i couldn't survive) but it seems laughable compared to basically SS on nomod ar9
It's stuff like on a triple for example, i open up hitting it with finger 1, then i hit with finger 1 again at half beat spacing - but it's actually a quarter beat triple, so end result looks something like:
1--12 or 1--1*miss*
instead of 1-2-3
I don't know if it's similar, but i have/had a problem similar to it. At ar10/10.3+ i couldnt play rythms right. For standard od, that didnt make much of a difference, but with hr (in the case of ar10 ofc), it had much more impact. I saw exactly what happened, but my fingers just didnt react fast enough. I talked to some guy about it (he's way better than i am, around rank 500, but it's lower than his skill due to shoulder problems, long story short), and he said that practicing ar10.3 is the way to get better at it. I know you like to practice relax, and there's nothing wrong with that. But consequence seems to be that you'll see much much more than you can tap. Since i've been practicing 10.3, tapping comfort has gone way up at ar10. I feel i can now play almost any nonextra map with hr. https://osu.ppy.sh/s/29290 is an excellent map to start, due to less predictable rythms. https://osu.ppy.sh/s/13019 is very hard, but excellent practice Or you can try to edit easier maps into ar9 to practice.
Or maybe I'm misunderstanding, in that case ignore my post.
On August 21 2014 06:54 Yorbon wrote: I read the first paper you posted, and it was interesting indeed. I don't think i read anything that makes me change my mind significantly, but it was good to read some feedback on how to improve practice at some specific parts.
Tbh, that refinement occurs when pushing your limits isn't a surprise at all. What's a refining skill for one map, is an essential one for another. There are however other factors (bpm and ar come to mind for example) that come into play, that correlate with the skill needed to play maps (at least in osu standard, i don't know how it works in mania). At some point your ability to play maps will decrease when you're pushing your limits further. And that's when i think playing for refinement is worth it for practice purposes. Your ability to play maps should be independent of bpm/ar (provided you're fast enough or are able to read good enough), which i think is impossible by just pushing your limits. As an example, there are quite some people who are unable to read ar9/8/7, because it's too slow. In my view, that's because of a lack of refinement practice. I'm not saying refinement is absolutely necessary. I just think it has some value.
I'm sorry i was using maps you didnt know. I thought they were old enough for you to know them, but apparently not. ):
Actually, I think the problem with low AR stems from which information you prefer to use when determining when to hit a note in osu. I find that once I reached AR8 or so, I stopped watching the approach circles very much and began watching for when a circle actually appears. Since the time between when the circle appears and when you must hit it is tied to AR, and becomes low enough to act as a useful way to time things only at higher ARs, it makes sense to me that this would be a likely change for people to make.
However, another thing to consider is variance in individual player's skills. Not everyone is good at the same things, and I feel that low AR is another one of those extremes where a player's natural talents or lack thereof can come into play. Once I got used to high AR, like many people, I could no longer play low AR. Thing is, this kind of pattern has happened in every music game I've played. Once I get used to faster paced stuff, I get bad at slow stuff.
High speeds demand fast reaction speed, meaning you must look for cues in the game that can be used to determine the timing of a note as quickly as possible. By their nature, approach circles are not a good candidate for fast reaction because they all start at the same size, and the only way to determine correct timing is to watch them over a period of time. Watching when a note appears requires only noting when the note appeared and waiting a predetermined amount of time, so the timing of a note can be known from the instant it appears. This also breaks down at lower AR because low AR means longer delays meaning you need to hold the information in your head for longer and it also means more objects onscreen at once to keep track of (another potential reason why people struggle at lower ARs).
In mania, approach rate is replaced by the scroll speed, but unlike standard, you can actually control the scroll speed. However, the mapper can still cause the scroll speed to change by using SV multipliers which are applied on top of whatever speed you're using. for example, if you use a scroll speed of 10 on a song, and an SV multiplier of 1.5x is applied, the song will now scroll at speed 15 for the duration of that timing section.
EDIT: to clarify, speed calculation in mania takes into account BPM, SV multipliers and the scroll speed you specify. As far as I know, they are all linear in the sense that 1.5x SV = 1.5x scroll speed = 1.5x bpm.
I'm sorry, but i don't feel your post in any way refutes what i was saying. Talent and your ciritisms on using circles as timing seem irrelevant (there's a reason i say approach rate, and not approach circle), and your personal situation as well as the reason people struggle with lower ar, are exactly what i mean. Osu has never been about reaction speed, always about pattern recognition. Keep in mind we were talking about refining play by repetition. In my view, if you once were able to play a map, but no longer, you've only become faster, not better.
There's usually an AR for standard that "works best" in terms of how easy it is to read patterns - too low and you can have pretty catastrophic results unless you are great at reading, too high and everything just becomes single circles instead of patterns
you obviously have to be able to read what you set, so you need to be fast to a certain degree but also able to handle lower AR's if it's better for the map or it's set that way. It's really important to be able to read a map, which means having an AR tolerance from say 7-10 instead of "i can only play ar9"
It's not really AR that's important for the map feel in standard, it depends a lot on the patterns, how circles are used and how many are on the screen. For example, Drop - Granat and some other maps like that 230bpm ar9 one are ar9 - but they actually feel very low. My performance on them is extremely strongly correlated to my ability to other "SLOW" maps, like 160bpm ar7's.
Those maps directly become easier when you edit the AR from 9 to 10 unless you're great at reading - with the effect of them being EASIER to play with hard rock than without in some cases, because ar9 is "slow" for the type of patterns they throw at you and amount of circles that you have to process at once
Another reason it's important is that we have a wall in terms of reaction speed - IIRC, ar9 is 600ms, ar10 is ~400ms.
As you increase pace, the "optimal" AR, for easiest to read in terms of patterns and how many circles at once increases. By like 180bpm that's already 10 on some maps. You can't continue raising it to make it faster and faster so that higher BPM or more dense maps sustain that ratio, because we simply can't process etc ar10.5. The only way to actually play them is to play them on an AR that is comparatively, too low to the point of making them harder - and you get those skills from playing like random 170bpm maps on ar7, or 230bpm ar9 etc.
[11:32:39] Cyro: " Saiya - Remote Control [Insane] (97.36%) less than a minute from now
153pp " [11:32:39] Cyro: fuck [11:32:45] Cyro: i broke combo on the second jump set [11:32:48] Cyro: not because it was too hard [11:32:53] Cyro: but because my hands were shaking so much [11:32:57] Cyro: my heartrate was way over 100bpm [11:34:38] Cyro: i reached them with 98.5 acc [11:39:55] Cyro: Saiya - Remote Control [Insane] (96.52%) about a minute from now
170pp weighted 77% (132pp) [11:39:57] Cyro: fc (: [11:40:03] Cyro: can get more pp outta that one (: [11:40:20] Cyro: it's od9 so enabling hidden would MURDER me [11:40:25] Cyro: i mean maybe it gives +15pp [11:40:35] Cyro: but having 95% acc instead of ss loses like 70 pp lmao [11:41:24] Cyro: sad part is the jumps really are the easy part [11:41:33] Cyro: well they're pretty hard to reliably combo [11:41:38] Cyro: without relax [11:41:43] Cyro: but [11:41:49] Cyro: i'm more comfortable doing the jumps than the other parts [11:41:51] Cyro: with high acc [11:46:11] Cyro: " Saiya - Remote Control [Insane] (98.19%) about a minute from now
192pp weighted 95% (182pp) " [11:46:12] Cyro: ok aeryn [11:46:14] Cyro: not bad, not bad [11:46:31] Cyro: ABOUT A MINUTE FROM NOW? LOL I'M REMOTE CONTROLLING IN THE FUTURE
ranting to myself as usual but yay, remote control FC! My second highest rank now.
Also i saw some people over 3k pp without any scores of even 160pp. With the way my top scores looks now i should be able to pad out my ranking a lot just by adding a ton of easier fc's, i think i broke out of plateau now
ohh god the rankings plz stahp
SMiLE.dk - Golden Sky [Insane] +DT (98.79%) 2 days ago
195pp weighted 100% (195pp)
Saiya - Remote Control [Insane] (98.19%) about an hour ago
192pp weighted 95% (182pp)
Drop - Granat [Extra] (100.00%) 2 days ago
187pp weighted 90% (169pp)
Lifetheory - Daisy [Blossom] (100.00%) about 20 hours ago
157pp weighted 60% (94pp)
And now this:
Blue Stahli - Shotgun Senorita (Zardonic Remix) [Insane] (99.61%) 8 minutes ago
170pp weighted 74% (125pp)
Drop - Granat [Insane] +HD,HR (95.81%) less than a minute ago
168pp
BAM
Drop - Granat [Insane] +HD,HR (97.82%) less than a minute from now
194pp
*sits back*
My work here is done
od10 is a harsh mistress.. dat part at ~0:24-0:27 - Also, which sadist made an od10 map that gives 160pp for 95% accuracy, but 250pp for SS? I can't imagine how much hair i would have ripped out if i was trying to get that~~
EPT
