Site banner

Improving in Etterna

Table of Contents

  1. tl;dr
  2. Skillsets
  3. Technique
  4. Feedback
  5. Playing Consistently (Or, avoiding rust)
  6. Reading
  7. Various common pitfalls
    1. Overthinking/Overoptimizing
    2. "I'm 1X.XX MSD, how do I get chordjack/handstream rating?"
    3. The issue with "play more"
    4. Caring about only improvement or scores
    5. Giving prescriptive advice
  8. Addendum
    1. What about vibro?
    2. What about keyboards?
    3. What packs should I play?
  9. See Also

tl;dr

I think the best way to summarize this while avoiding the typical issues with the extreme simplification of "play more" is the following:

The rest of this post is either elaborating on those points, or commenting on certain common improvement questions or anti-patterns.

Skillsets

As stated in the tl;dr, playing charts with specific patterns will make you better at similar patterns in other charts, but may not be as useful for charts with significantly different patterning. Consequently, some players might seem like they're improving much faster than others because they're specialized in one or a handful of skillsets, and consequently they can get more impressive scores for the amount of time invested. Of course, if they played any other skillset instead the scores would be significantly less impressive, but this can be avoided by not posting scores on the skillsets neglected.

Conversely, a player's skill might be underappreciated because they're a generalist that spreads their playtime across many different skills, some being very niche, and consequently may seem to be of lower skill because none of their scores are particularly impressive on their own relative to their time spent playing, even if those that could do better would not be able to match them in some other skillsets.

One notable example of this is Foxfire, who along with having a lot of experience with Stepmania pre-2017, also played Etterna extremely often from 2017-2022, having three 2-3 hour sessions per week consistently for 5 years, making him have one of the highest playcounts on Etterna. Despite this, he ended up peaking at only 31 MSD, which is generally much lower than one would expect from playing that often for such a long period of time.

However, this is not because Foxfire is an underwhelming player but rather because he spread his gameplay over a much larger span of skills, as he played pretty much everything and also was often requested to play files on stream that targeted extremely niche skillsets, i.e. rolltech. This meant that while he wasn't the best player in any skillset (besides maybe rolltech, as it's a skillset almost nobody else has dedicated themselves to), there is likely no other player that can match his scores in every single skillset.

On the other hand, there have been plenty of cases of players that gain MSD extremely quickly, but this comes at the expense of every skillset that does not lead to gaining MSD.

It can be tempting to overspecialize for a few reasons. The aforementioned ease of setting good scores if you just one-trick is one thing, but also it can tend to feel bad or uncomfortable to play skillsets that you're bad at, especially if you're comparing to your best skillset without considering the difference in time spent practicing. Ultimately it is up to you how you want to play the game, and it's valid to specialize if that's what you like, but do realize that there may be a lot of very enjoyable chart experiences you will miss out on if you neglect certain skillsets, and it's particularly a shame if you're just doing it because you refuse to acknowledge that learning an underpracticed skillset will take effort.

Technique

This general idea also applies to technique. The same pattern can be hit in multiple different ways. For example, while chordjacks are generally considered a single skillset, the motions your hand will do to hit a chordjack greatly depend on the density. For very dense chordjacks the preferred technique is wristjacking, while with lighter CJ it's more optimal to use a mix of techniques depending on the length of anchors and other details of the patterning. Either way, you will get better at using whatever technique you use to hit the pattern.

Another example of this is pattern manipulation. If you jumptrill a rollwall, you will get better at jumptrilling rollwalls, but it's not going to help you with hitting them conventionally. In fact, it may make it harder due to having to consciously override the muscle memory you already have to jumptrill the pattern.

Feedback

What's also worth noting is that what generally causes the improvement in your muscle memory is the subconscious response to the feedback the game gives you. This is primarily through judgements, but it can also be through other means, like combo, column highlighting, or the error bar. The point of most feedback is to point out outliers, or around the 1-5% of worst hits I would personally say. Any more than that and your feedback is likely too sensitive and you won't be able to recognize every mistake made, and any less your feedback is probably too insensitive and you're likely being unaware of mistakes that could be acknowledged and adjusted to.

The main case where this typically applies for Etterna is using higher judges, like judge 7. If you aren't able to consistently get near quads, or in other words hit very consistently in the J4 Marvelous window, then having a timing window twice as tight will not help you because you arent going to be able to consistently hit in that tighter window, so being informed when you're not hitting that precise window is noise.

That isn't to say that there aren't other valid ways of using higher judgements even if you can't do J4 MFCs, however. Some players push on J5 because they mainly focus on their great count, and they think J4's great window is too lenient. I push on J7 because I don't use judgements and instead use CB column highlighting to identify outlier hits, including greats, so J7's ability to make greats count as CBs is valuable.

One last note: The visibility of the feedback matters too. One reason I personally don't like using combo for feedback is because it's so big and obvious that it tends to cause players to prioritize not dropping cbs to keep the number going up, even if doing so results in worse accuracy. This isn't as much of an issue in Etterna due to goods/bads also breaking combo, but it's still worth noting.

Playing consistently (Or, avoiding rust)

When you're playing Etterna, you always want to be improving on your peak skill. This is because your performance, while not always at your peak, will be somewhere close to it. Even if you do not play the game for a year, if you come back you will likely be able to play at least 5 MSD below your peak skill, so if you're 30 MSD you can expect to at the very least be able to play 25 MSD files within your first session coming back after a year. There's also the more vain reason that playing at your peak produces more impressive scores, but this shouldn't be your main motivator in my opinion.

So how do you stay at close to your peak skill level? By avoiding rust.

Rust is the term for when you stop playing the game for an extended period of time, and consequently you lose your muscle memory and must spend time playing the game to get it back. You want to avoid this because time spent derusting is time that could have been spent playing improving your peak skill if you hadn't rusted in the first place.

That being said, it's not the end of the world if you do end up rusting; getting back to your peak skill from rust will take much less time than it took to get to that skill level in the first place, generally around 1-3 weeks. This also doesn't mean "You should force yourself to play constantly even if you don't feel like it", but if you're really serious about improvement you should play at the very least one session per week to avoid time being spent inefficiently derusting.

This mainly applies for people that tend to play the game for say, around a month or less at a time and then don't touch the game for another month before coming back to it. At that point, the majority of your time is spend derusting rather than at your peak, so it's no surprise when your peak skill ends up not improving at all.

It's also worth noting that rust can also apply for individual skillsets. If you happen to really focus on improving one skillset for a few weeks, don't be surprised if you fall behind a bit on your other skills. Whether this is acceptable or not in order to focus on improving one skill more is up to you.

Reading

Generally speaking, your settings should stay fairly consistent. Adjusting your CMod (scroll speed) as you start off is fine, but you want to avoid bad habits like having different CMod for different patterns, as that screws with your reading in multi-skillset files and also just in general. If you hit late/early on certain types of patterns that will go away with practice over time, don't force it with microadjustments. Another reason you don't want to do this is because, as mentioned later, performance is very unpredictable in the short-term and it's easy to mistake random variance for improving due to a change in settings.

Also on the topic of reading, don't focus too hard on the game but also not too little. Basically, you want to have some sort of focal point at around the middle of the playfield and look at around that area the whole time. Focusing too hard will make your eyes burn and focusing too little your eyes will drift too far away from the focal point and your reading will be messed up.

Also, don't think too hard about the specifics of what your hands are doing during the chart. You really don't need to and chances are it's just going to make them lock up or do something unexpected. This is a subconscious game, let it do the work.

Various common pitfalls

Overthinking/Overoptimizing

While a lot has been said in detail in this post, it all can be summarized pretty simply, and most players intuitively do the right thing because it isn't actually complicated.

Looking for further complexity where it doesn't exist however can just lead to frustration when playing. Namely, while improvement at a skillset is strongly correlated with playing more, it is not absolute. Sometimes you will have off-days, and sometimes you will be playing unusually well. Consequently, if you try to aggressively optimize your gameplay by doing session-to-session comparisons, or even score-to-score within the same session, you will end up just frustrating yourself with a bunch of inconsistent results due to the inherent unpredictability in performance.

One way people like to describe the unpredictability of improvement is that improvement follows a sine wave. This gets the idea across that you can play worse than before even after a good session. Though to be pedantic, it would probably be better to describe improvement as being strongly correllated with time spent practicing, with some random variance. A graph is provided below, think of each dot as a session practicing a certain skillset.

A graph showing the idea of skill and time being strongly, but not completely, correlated

"I'm 1X.XX MSD, how do I get chordjack/handstream rating?"

This is much more specific to Etterna and the current culture of the game than most of my other points, but this gets asked often enough that I figured it was worth covering anyway.

The calculator is not, and has never been, targeted towards new players. It was designed by Mina, a top player, for other high level players back when Etterna was just Stepmania but with all of the features that the hardcore keyboard community wanted.

This means that the difficulty calculator makes some assumptions about difficulty that do not hold, or even make sense at lower difficulty levels.

When you are new to the game, you will find it very hard to find handstream/chordjack charts because the calculator does not believe it is possible for handstream/chordjack to exist at such a slow speed, and will probably classify it as another skillset. If you want to practice denser chords at a low level, look for charts that have a lot of hands in the chart's stats. Alternatively, play the few packs that exist that are dedicated to these patterns at low levels, like Oxicleaned and Neon Club Handstream 1/2. These charts might not be classified correctly for the aforementioned reasons, but they will help your reading. Examples of where to find the hand stats on the two most popular themes is provided below:

Rebirth hands display Til' Death hands display

It's also worth noting that as a result of this, pretty much every new player will have imbalanced stats, and that's fine when starting off even if you're concerned about having a balanced skillset later. If you want to keep your skills balanced, you should probably ignore MSD until you're around 20-24 MSD. Even then, realize that it's not a perfect indicator, so also consider the opinions of other players and your own intuition rather than trying to perfectly balance the calculator's numbers.

The issue with "play more"

The issue with just saying "play more" is that not only does it feel like an unsatisfying hand-wavey answer to whoever's asking for advice, but it also isn't correct either. If the game really just came down to "play more", then we could just develop a 100% accurate leaderboard by just sorting profiles by playcount. In reality however, the players with the highest playcounts, while very good, are far from the best. The main purpose of writing this article is to fill in the gaps of just saying "play more" while still keeping the advice generally applicable regardless of what a player wants to improve on.

That's not to say that "play more" is entirely bad advice either. Play count is still generally correlated to player skill, and giving general advice like this implicitly discourages the player from overcomplicating things.

Caring about only improvement or scores

If the only reason you are playing this game is to get better at it, you are bound to disappoint yourself when your expectations do not match what you can get, and it will be frustrating because as mentioned before, sometimes there isn't a good reason why you're doing worse than usual in the short term.

To avoid this, having reasons why you play the game that don't have to do with getting better will make the game more enjoyable, motivating you to play more and indirectly leading you to improving more.

For example, for a very long time from 2017-2019, I was stuck at around 15-17 MSD. I was stuck in a cycle of picking up the game again with the sole intent of getting better, and then playing for less than a month before I tilted myself enough to not want to play anymore, and then coming back fully rusted. Obviously, in these conditions, I was basically never improving. Eventually, I decided to instead come back to the game and just play through some old packs I was interested in, and that was when I started improving again.

This doesn't mean that wanting to improve is an inherently bad mindset, though. It should just be tempered with the knowledge that you will not always be performing well, and that you really shouldn't place too much weight into just getting better. Having other reasons to play the game helps with that.

There's also plenty of other ways to enjoy the game besides completing packs as well. It can be as simple as just being in a voice call with friends while playing, or just listening to a video in the background or something and treating Etterna like an exercise you're doing to multitask since the game doesn't require much active thought. Playing to give feedback on charts can also help if you happen to be a charter/playtester.

Giving prescriptive advice

This is particularly directed at those that are giving advice. I see fairly often people giving prescriptive specific advice on what to play or what skillsets to practice or even what judgement or settings to use.

This isn't necessarily bad, but should only be done with a good understanding of the player's skill level and what exactly they want to improve on. For example, if someone is fairly new, say 13 MSD, and asks for how to get better, don't just say "go for AAAs", even if their highest accuracy scores are only 96%. New players are bad at accuracy in general, and this is fine. AAAing anything at all is likely out of their skill level, and will be until they get a fair amount of experience with the game. If they listen to your advice, they'll either have to play an extremely boring 4 MSD file to get it, or fail to get it at all and be discouraged from improving their accuracy because they seem completely incapable. It would be better to just recommend them to play files that they can FC in general, to avoid setting a standard for MA/PA that they can't achieve.

However, this also assumes that the player in question even wanted to improve their accuracy in the first place! I know certain players have strong feelings about proper technique, but it's best not to force this on others, especially if they're still just figuring the game out in general.

That being said, it's also fairly possible that new players will ask how to get better without having anything specific in mind. While "play more" works, especially at this skill level, if you want to be more specific while being general enough to apply regardless of what they want to improve in, you can provide them with something like the tl;dr of this post, or just linking them this post in the first place.

Addendum

What about vibro?

Vibro/vibrating/vibrajacking is a technique where you tense your arms and it causes them to shake, letting you do single column inputs very quickly at the expense of very little control. Some people also use it to refer to really long, fast but simple jacks that are easily hit with the technique.

Some people can do it and some people can't. I can hardly do it but those who can say that improving stamina and control is doable but speed isn't.

Due to being not generally accessible and the Etterna difficulty calculator not factoring it in, it's generally not taken seriously by most players and using it on the leaderboards is technically cheating but unenforceable (How to distinguish between vibro and just really fast elbow/shoulderjacking, especially with only replay data?). Feel free to do it if you want but realize that it's generally treated as separate from playing the game otherwise, and almost entirely does not translate to skill in other aspects of the game.

What about keyboards?

tl;dr: If you're not going for quads or you're not 30+ MSD, it probably doesn't matter. A $15 membrane keyboard will work fine.

If you're going for quads, you're going to want to find yourself a 1000 Hz polling rate keyboard. Most keyboards poll at 125 Hz, which is fine until you're trying to consistently hit in the marvelous window and a decent amount of your inputs, while they may be timed correctly, could be processed outside the marvelous window due to waiting on the next 8 millisecond interval the keyboard is scanned for inputs. Refer to this sheet for 1000 Hz keyboards; while some manufacturers state their polling rates, it can be possibly incorrect because polling rate doesn't matter for basically anything but rhythm games, and most other games/software don't have a reliable means of showing polling rate visually like Etterna.

If you're playing really fast (30+ MSD) stuff, you also probably want a mechanical keyboard with a low actuation force so you can reduce the force you need to exert on each individual keystroke.

Rapid trigger keyboards like those made by Wooting make it easier to vibro and play minedodge, but otherwise have no significant benefit. The more valuable aspect of hall effect keyboards is that due to not being contact-based like traditional mechanical keyboards, there is practically no electrical wear, making the only concern of keyboard damage the stems themselves breaking. This makes hall effect keyboards in practice significantly more durable than most other mechanical keyboards.

What packs should I play?

Depends how good you are, what skillsets you want to play, and whether you're fine with downrating or not.

The usual recommendations for new players are Stepping Stones 2, 3, and 4, and the FFR Novice Files Compilation. These packs have a wide variety of charts with multiple difficulties, making them ideal for getting a balanced experience with a wide variety of skillsets.

If you're really new, it can also be worthwhile to download some packs made for playing on a dance pad, like the BemaniBeats and Anime Extravaganza series. Do realize however that because these packs are made for pad, you will not encounter some patterns that are common in keyboard-specific charts.

You can also sort the pack listings page on EtternaOnline by overall rating in ascending order to see the easiest packs ranked, although the quality may vary.

Once you get a bit more experienced and want to find packs that specialize in certain skillsets, this list by FirstMaple is a decent starting point, though do note it hasn't been updated since 2020.

If you have any recommendations on other packs to list, or even better have made some sort of table or spreadsheet, do let me know so I can expand this section and make it more useful.

See also