(This is an unusually large amount of words to find on a music game forum. It's not a lot of words for an article somewhere but way more people will read the music game forum entry than the article, so I'll tolerate a few people complaining about how long this is because I know it'll reach more people anyway.)
If you've ever played a music game at higher levels you're probability familiar with the feeling of hitting a "wall," where you play and play but don't improve very much.
Most people don't bother to think about why they hit a wall; they just keep playing and after a bit of time, they manage to overcome it. I hit a wall in 6key/7key Stepmania improvement for a while and I didn't know why it was that I was at this wall. I took a step back and started thinking about getting better at Stepmania like lifting more weight: how do you do that? The way I've always done it is just by going in and playing the hardest songs I can until I get better. But the gym equivalent of that is going in and lifting something you can lift twice at most. And even then, hard songs aren't consistent -- there will be some hard parts and some easy parts.
It occurred to me that I had developed a way of thinking about skill increases in music games that probably wasn't reflective of skill gains in general, especially since music game gains are more dependent on bodily gains than, say, chess skill gains. So I reconstructed the way I thought about difficulty from the ground up, and developed a coherent way of thinking about improvement that can be applied to the vast majority of music games. This works the best for hand-based music games like Stepmania, IIDX, Guitar Hero, Rock Band and so on, but if the game sufficiently resembles those enough you can probably use this guide for it.
Now at this point some of you probably think this is silly. "I play the game for fun, not to get better." That's completely understandable. Some people, like me, find playing for skill the most satisfying way to play; something like this is fun to us. But if you're a casual player or just don't want to dedicate your game time to skill increases, no one's judging you. Besides, people like me are the weird ones to want to optimize skill increases on a music game.
That said, everything here will be about that, so if you don't find that interesting you might want to stop reading.
Skill Types
Virtually all music games across all platforms (arcade, console, PC) have skill types that can be divided into two categories:
1. Performance (can you hit the buttons/notes?)
2. Reading (you decipher the game's symbolic representation of the buttons/notes to hit?)
Most people do not regard #2 as a form of skill, or, if they do, they regard it as a poor representation of skill. This is because most music games use songs that have a fixed chart of button presses that go along with the song, so in theory you could just memorize that chart and the visual representation wouldn't make a difference -- you could play the song blindfolded, for all anyone cares. Some people will purposefully randomize the patterns and crank up the visual difficulty so that they would
have to be exercising superior reading ability, but you could still memorize the song's rhythms, which would narrow down your button presses from "where and when" to just "when".
So in other words, most people see #2 like if someone made a difficult math problem, but printed it in wingdings -- or if someone made an obstacle course, but you had to do it with kaleidoscope goggles on. They provide some additional level of difficulty, but a kind of difficulty that is irrelevant to the skills you get from playing the game itself. It'd be more difficult to complete a hard song if your uncle was yelling at you and a rat was gnawing on your toe, but most people wouldn't consider that "being good at a music game," they'd consider that "being good at resisting distractions."
Skill Subtypes
Skill type #1 ("performance") contains two categories (physical and cognitive), in which there's three major factors:
1. Physical ability (can you hit the buttons/notes)
2. Physical endurance (can you hit the buttons/notes for n period of time) -- this is usually exclusive to DDR-style games or music game songs that involve very fast button-pressing for long periods of time
3. Cognitive/timing (can you time the buttons/notes consistently)
I don't know for sure how to increase music game timing to very high levels, mostly because I've never placed much value on it but also partly because I've never been that great at it. However, I've observed that as physical ability increases, timing tends to increase along with it. This is because your ability to time songs well is partly influenced by the ease with which you can press the buttons for that song; if your hands are very nimble, you can tense your hands for greater precision. But if you have to tense your hands as it is to do the song period, you won't have much room to hit the buttons more accurately.
Some players exist who have very good timing but not very good physical ability, or the other way around, but "splitting" like this is rare. Most people who have high physical ability also have good timing. There are some weird cases, like if you are able to hit very difficult pitches for the vocal parts of Rock Band but can't rap fast or enunciate well at all -- but that's a particularly isolated case, and with hand-based games, this pattern largely holds.
Types of physical ability in music games
Some music games, like ITG, have specific sub-abilities like heel/toe prowess. Stepmania has one in particular which we know as "vibrajacking," (repeated arm vibrations on one key) though you could implement vibrajacks in other games if you really wanted to. All hand-based music games have at least two, but usually three kinds of ability:
1. key-to-key / button-to-button (abbreviated K2K): the ability to move your finger or hand from one key to another. This is the ability primarily used when someone plays the drums on Rock Band, or index on Stepmania. You could also think of this as "repositioning ability" or "repositioning speed", but I like this term because it's easy to remember.
2. finger-to-finger (abbreviated F2F): the ability to alternate between one finger or another. Guitar Hero, IIDX, and 6-key Stepmania are mostly this ability; 6-key Stepmania in particular is nearly entirely this ability, since your hands are stationary on every key. You could also think of this term as "finger nimbleness", but again, I like this term better because it's concise and easy to remember.
3. wrist ability. Since on a very basic level you're just pressing down buttons, your ability to do this at very fast speeds will depend on how fast your wrists can make downward motions.
If you play any instruments and find this easier to think about abilities in terms of instruments, try this: piano will be mostly F2F and some K2K, while drums will be mostly K2K and wrist ability.
Pre-music game skill factors
Some skills exist where you can start out on an equal playing field relative to everyone else. If you and I have never cooked a meal in our lives or chopped a single vegetable and we both sign up for a cooking class, chances are we're both going to look like idiots at an equal rate for a while. Maybe one of us will have a greater aptitude due to some other unrelated factor, but for the most part we'll start as equals.
Not everyone starts out as equals on music games. If I made an investment simulator that strongly mimicked the process of investing in a company, people who had taken finance courses prior to playing the game would be at a huge advantage. Music games are like this; they're unlike most other video games with a high skill cap in that you can import a great deal of skill that's external to video games.
If you've played piano before, you will start with an enormous advantage in F2F compared to someone who hasn't -- and this is especially true if you've played piano at a high level. Piano pieces are graded from 1-8,
here is one such list. If you can play at grade 5 starting, you're at an enormous F2F advantage compared to someone who hasn't played the game before. Potentially at least a year's worth of play, if not two. You can see this effect in action in certain expert players: LISU, one of the first players to play one-handed IIDX at a very high level, was a classical pianist prior to playing the game.
Likewise, if you've played drums prior to playing, say, index Stepmania or Pop'n Music or, more obviously, drums on Rock Band, you'll have an enormous K2K advantage as well as an enormous wrist advantage. Very fast snare rolls train wrist ability, and the "map" you'll be accustomed to using for a drum kit will make switching between keys/buttons on a video game seem a lot easier.
Skill for new players
Based on how much I've hyped playing an instrument, you'd think that a classical pianist could pick up Beatmania IIDX and start dominating. Virtually everyone who has played music games at a high level knows this doesn't happen, but there are deeper things going on. On a superficial level, it's easy to say "they're different things." A lot of people who play instruments in their spare time will pick up a game like Guitar Hero, suck at it, and say something like "hah! I can play [instrument] but can't play this" or, alternatively, "wow, this is hard. I can play [instrument] but I can't play this." This leads those players to believe that their skills on those instruments either doesn't help or, if it does, it's downplayed.
Sure, it's true that when you first start playing a music game, you will *suck* at virtually everything. This is beyond dispute. You will not be familiar with the way music games symbolize notes, and the controllers are nothing like a real instrument. But this is a lot like if someone who has spoken Italian their whole life (but is illiterate) and someone who has spoken Chinese their whole life take a Spanish class: although they both have no idea how to write down the sounds and they both find Spanish unfamiliar, the Italian speaker has way more of a familiarity with the mechanics of Spanish, since they both come from Latin. Once the Chinese-speaker and illiterate Italian-speaker have learned how to write and are getting in the motions of things, the illiterate Italian speaker will
obliterate the Chinese speaker in terms of progress.
Remember, there are two skill types: performance and reading. Someone who is
completely new to music games and instruments will start at nothing and be new to everything, performance-wise and reading-wise. Someone who has played instruments before will suck reading-wise and may
seem to suck performance-wise, but that's because of two reasons: they haven't learned how to read the notes on the game yet, and they haven't developed neural connections (a "mind map") for the game's controller and when the game tells them to press a note.
In other words, the initial difficulty is a facade. Everyone sucks at music games first because music games have an idiosyncratic way of representing notes. But after you're comfortable with the way music games represent notes, real noticeable differences will emerge in how fast you improve. This is important to consider because it may at times seem like someone just has an innate ability for the game, and rarely is that the case.
Improving skill: the naive approach
There are two ways to approach improvements in, well, anything: with a method and without a method. A friend of mine calls methodless improvement in the gym "fuckarounditis", because when people go work out without a method they tend to be completely unaware of what it is they're doing to improve.
With respect to music games, the equivalent is playing songs at whim without any sort of structure or goal. Certainly, you can improve this way. In fact, I wouldn't be surprised if some of the best players had done mostly this. But even the best players, unconsciously, play songs that challenge them or help them improve to an extent. The urge to get better hits someone even if they don't have structure.
What songs you play matter. Eventually you will get to a point where a range of difficulty exists that will not help you improve at all. This will be anything below 60% of the maximum of whatever it is you're capable of doing. So on a scale of 1-100, if the hardest song level you can do is a 70, anything below 42 will be worthless in terms of skill improvement; in fact, it will probably put you to sleep. This makes sense, if you think about it: that would mean that the lowest a DDR player (DDR goes from 1-10) who maxes out at 7s should go is 4, and the lowest a person who maxes out at 10 should go is 6. The exception to this is if you're playing a game that gives you access to a very strict timing window, in which case reverting to easier songs might help you. But for increasing physical skill, playing significantly below your ability range will not help you at all.
Most really good players, I've noticed, tend to gravitate toward songs that are about 70-80% of their maximum skill level without consciously realizing that's what they're doing. Interestingly, this is also the ideal % range for building muscle when you're lifting weights. But this comes naturally because after a certain point, very easy songs become very boring. This is also why there was such a controversy in DDR about the addition of hard songs: expert players wanted harder songs to be added because they were at such a point of proficiency (13s on In The Groove) that most of the game was far below 70% of their skill level, and to play at 75% would require them to play hard 9s or easy 10s. To play at 80% would require them to play a difficulty range that didn't exist, nevermind above that. Less advanced players who could pass 10s, but just barely, felt threatened by this and didn't see the need for a change: they were technically at the highest level of the food chain (10 was the highest rating), but playing at 70% of their maximum was still very possible (7s) and many songs fit within their comfort range.
The need for a method
The naive approach works for games like DDR and index Stepmania for two reasons:
1. There are only four keys
2. These games involve endurance, K2K and wrist ability more than F2F ability. At the higher levels it's almost entirely endurance/wrist ability.
If you haven't played a music game at all, your first few months of playing any music game will be completely mental. You simply will not have neural connections for how the game works. Your brain has yet to map out an intuitive idea of where each key is and develop connections for the music and the game's visual note system. This is also true with lifting weights: your brain has not adjusted to the movements of weightlifting yet, and during this time you'll see enormous gains whether you're doing it right or not. So it's very likely that
any play will increase your skill level, because the act of playing is the act of getting your brain acquainted with the game's layout.
And up to a certain point, this might work. But there will be a point where it stops working as well as you'd like it to. Or alternatively, there will be a point where it doesn't work at all -- you'll hit a wall.
And
in all cases, you'd have had faster gains in skill
had you used a method. Which is to say,
A: no method
B: method
Assuming B's method is optimal, B will beat A every time, given equal starting ability (equal or no familiarity with instruments) and equal playtime. This is true with piano, true with test prep, true with weightlifting... true with skill increases in pretty much everything. All other things being equal, method > no method. (Do not take this as an endorsement of method acting.)
Even something as simple as socialization or playing video games has productive and unproductive uses. Your parents may have told you to get outside and socialize, but hopefully you didn't say "OK mom, I went outside and socialized. My gang initiation's tomorrow." And to give you a personal example, I played 60 hours of Tropico 4 a ton last year, which taught me a lot about scale and how a small financial decision (in the game's case, variation in worker wages) can have huge impacts across large-scale operations. But I also played about that much of Elder Scrolls IV: Oblivion and didn't really get that much out of it, except possibly that I'm super jealous of the layout/architecture of the Arcane University and I wish a learning environment like that existed in real life.
Malcolm Gladwell has a seriously overhyped book, Outliers, where he mentions a "10,000 hour rule" for obtaining mastery in anything. The gist of this is that if you want to master something, you need to invest at least 10,000 hours into it. To a certain extent he's right: mastery requires a ton of practice. But what he overlooks is
how you're practicing. 10,000 hours might be the average, but assuming ideal and non-ideal practice, it could be as low as 5,000 or as high as 15,000. Hell, maybe even the variation could be greater.
The Isolation Method
The basic principle of the isolation method is this: if you want to increase your skill in something music-game related, you need to find out what it is you're not able to do and isolate for that thing.
This does not mean "play the songs I am bad at a lot until I get better at those songs," because it could be that you're only bad at parts of those songs. If you're playing in an arcade, this may be the only option. However, if you think of increasing skill on the song level ("which song am I bad at"), you could be
extremely under-utilizing your time. If a song is 90 seconds and only 10 of that is the part you're bad at, 87% of your time is utilized in a way that doesn't improve your skill level.
So the solution is to take the part you're bad at, and isolate for that. Sometimes this is possible, and sometimes it isn't. But in all cases, you should be aiming for doing about 70-80% of what you're capable of. So if you're aiming to increase the speed with which you can hit I O P O (repeating) in order, the best thing to do would be to slow it down to about 70-80% of what you're capable of and do that, over and over.
Isolating for K2K
Since K2K is primarily about your neural map for the game, you will want to pick parts that seem wonky or that make you feel tripped up. So you should try to deliberately confuse yourself: if you're playing music games with an editor: go into the editor and make patterns that are very confusing and do those.
For example, index Stepmania has some very easy K2K patterns and some very obscure ones that almost no one uses. Easy patterns would be something like left-down-up-right, or 4286 on the number pad. Very difficult patterns would be something like 4864286426, or left-up-right-left-down-up-right-left-down-right, which is extremely awkward because you're "facing" the wrong direction as you need to pivot your hands. If you find these kinds of patterns awkward, deliberately confusing your hands and doing this thing over again on slower speeds will help you until you get it.
Generally, though, any time you feel tripped up or not coordinated enough to hit a pattern is when you should isolate for K2K skill.
Isolating for F2F
F2F does not have the same obviousness that K2K does. Which is to say, in some games it may be as simple as playing the game to raise K2K. But for F2F, you may have to get more determined about the way you isolate.
For increasing F2F speed, you can do the same as you would for K2K: find part of a song that has, say, a trill at 80% of your maximum speed, and do that a lot. Once you feel your hand/arm burning, you'll probably get some kind of increase.
But sometimes when you play a game that's very F2F-based, like 6key Stepmania or IIDX, you'll encounter the feeling of facing a "wall" of notes. This is a way of saying that there's a blur of notes that you should be able to read under other circumstances, but can't for some reason. Almost always, this translates to some physical difficulty with what the patterns are asking you to do. Occasionally that's not true, but most music games have enough speed modifier options now that you never have to worry about this.
It would seem like the solution is to slow down the wall of notes and just play that section. But sometimes, not even that works. If you're playing the hardest part of a song in F2F, it may not even be
the part of the song that's giving you trouble. Which is to say, you're using two hands: it could be that your right hand is what's struggling and not your left. That's not unheard of; people who just start to bench press may find that one arm is much stronger than the other. Usually the hand with greater F2F will be your dominant hand, but sometimes this might be the other way around.
The solution, if the game lets you, is to isolate for a particular
hand and just do that. If the game lets you (Stepmania for example), this will be in the form of chopping off an entire hand's worth of notes, and just playing the song with your weak hand. But if you can't do that, a solution might be to play one-handed with your non-dominant hand, and use practice mode to replay parts of songs that most resemble the patterns in the "wall" that you can't do.
Remember to think about the maximum of what you can do in terms of speed, not in terms of the game's difficulty. That is, think about it in terms of absolute rhythms per second. This is opposed to notes per second, which counts multi-hits. So 16ths at 120bpm are 8 RPS. It's best to think about it in this way because this will most measure what your fingers are physically capable of, as opposed to what the game rates you as being capable of.
Isolating for wrist speed and endurance
There are a few approaches to this.
One is the 70-80% principle that I've been talking about already. The other involves something that I'm not sure actually works, but I'll bring up anyway.
In fitness there is a cardio technique called high intensity interval training, or HIIT. This involves working out very lightly for about a minute or two, and then very intensely for about 20 to 30 seconds. (essentially, a 2:1 ratio of very intense to kind-of.)
This works very well for calorie burning, but I have no idea how effective it would be for wrist ability. I've also come very close to maxing how fast I can go for wrist speed, so I couldn't test its effectiveness even if I wanted to.
Isolating for timing
To my knowledge, there is really only one way to do this.
It involves playing Stepmania at very high judge difficulty settings. Now, remember that the rule is 70-80% of your maximum, so if you're comfortable on judge difficulty 4, try to do something like judge 6 or 7. Doing something like "justice" is a bad idea, because you will simply not be able to time that well consistently. Play songs that you can consistently get an A on, and try to get an AA or AAA on those. If you're getting a B or C, chances are that's too difficult to increase your timing reliably. If you're getting a B on everything, lower the judge difficulty by one notch.
Effectiveness and why this is a good idea
When I had hit my wall, I would play the same songs over and over again for 20-30 minutes a day and saw little to no score increase. Which is to say, I would get about 99% of my record every time, and if I broke that record it would be just barely.
When I started isolations, I did this for 60 minutes total across a few days and immediately saw improvements. All of my scores went up by at least 5%, and some as much as 10%, which I know because Stepmania 5 logs scores in terms of percentages. Obviously this rate of improvement will not hold every day, but it's remarkable that the improvement was this extreme.
When reflecting on it, it's obvious why this happened: most of my weakness was in my left hand, so when I would play parts that were hard for my left hand, I would treat the hard part as a unit and not something difficult for just my left hand. When I isolated for just the right hand by making a second notechart that had all of the right-hand parts in it, I could do the chart fine. But when I isolated for the left-hand, the problem was obvious: I
sucked at the left hand part, and I wasn't thinking of the notes in terms of "this is the hand I'm bad at," but rather "this is the song I'm bad at" or "this is the part I'm bad at."
Do the math: if you're playing a song that's 90 seconds long, and only 10 seconds of that song is a part you find difficult, then only 10 seconds of that song is improvement time. If you isolate, as much as 80-90% of your total playing time could be improvement time. So if normally you only get 20-30 seconds of a hard part per 2 minute song, and you play for 30 minutes a session (15 songs), only ~6 minutes of that is going to be improvement time. If you isolate, you could get practically the full 30 minutes of improvement. That's ~500% of what you were at before. Over the course of a year, you could improve as much as you could have in five years otherwise -- in theory, anyway.
You may think that this is obvious or silly, but you've probably spent a good deal of time playing the game. Would you like to be two or three times better than you are now? Probably. And in all cases where you've seen someone rapidly improve at the game, chances are they've done something like this.
Summary / tl;dr
- There are two types of skill at music games: performance and reading.
- There are three subtypes of skill: physical skill, physical endurance, and cognitive skill (timing).
- Of physical skill, there is mostly: k2k (key-to-key), f2f (finger-to-finger), and wrist ability.
- You will probably increase k2k and definitely f2f at a faster rate over the long term if you've played an instrument before, especially so if at a high level
- Increasing skill in general involves doing stuff at 70-80% of what you're capable of doing
- Isolate for
parts of the songs you're bad at, not the song itself
- For K2K, intentionally try to "confuse" your hands
- For F2F, isolate for the particular hand that is underperforming at particular parts of songs, since one hand might be way ahead of the other
- For timing, play on a higher judge (but not the highest judge) and play songs you can get an A on until you can get AA or AAA.
- Skill increases will probably be 2x or 3x your usual rate, could be as much as 400-500% in theory.