The idea of Inferential Distance² is that it's a way to describe how many "steps away" you are from understanding something. I think a good example to demonstrate what exactly I mean here is to just point to 3Blue1Brown who has an almost unnatural ability to do this for math concepts: if you want to understand how to solve a problem on the putnam exam, which is a very difficult mathematics competition, you might start by assuming the reader knows some basic linear algebra and then start your explanation from there. If your audience doesn't know linear algebra, they will be completely lost from step 1, and no matter how beautiful your explanation is they will not understand. If you start with points, lines, and coin flips, you could probably explain the solution to this problem to any high schooler.

Knowing how many things need to be explained before something is understood is called inferential distance, and not thinking about it at all is why communication in STEM is so, so horrible in the professional world³.

Fiction, among the melee teachers, has what I consider to be the strongest ability to figure out the skill level of his students as fast as possible, and then use this information in order to pinpoint how best to explain the concept that he thinks is the biggest flaw(s) in the student's play. Something that you'll hear pretty often if you watch enough Fiction lessons is something to the effect of "this isn't the whole matchup, but this will work on pretty much everyone you play".

Some criticism I see of Fiction's lessons are, in my view, reflections of him adjusting his teaching style around the skill level of his student, and he will occasionally provide a somewhat oversimplified description of a situation to give the student a starting point. The idea is that the simple description will get them 80% of the way there⁴, and the time spent giving additional nuance is better spent giving more structure in other places instead.

My favorite examples of this are Fiction's two lessons with Zuppy and Veronica on the Fox vs Ice Climbers matchup. The former laid out a fairly detailed, overall gameplan against the Ice Climbers. In the latter, the student seemed to have beautifully implemented all of Fiction's advice from the first lesson and still lost, which made for a very instructional lesson on the "next level" of that matchup, with some explanation on when the lead climber is your opponent and when you should go after Nana. Rather than aiming to make you as good as possible at the matchup, it's a lot more oriented to make this matchup not-your-worst-matchup as fast as possible so you can reprioritize to something else.

I think more than a lot of the other teachers, Fiction is enamored with the idea that you change what you are doing based on what your opponent thinks in response to your actions. Whereas the other teachers will typically focus on what to do in response to things, Fiction will focus more directly on what the opponent wants to accomplish with their moves, and how those goals are in conflict with yours.

To be a bit more concrete, Fiction will regularly mention things in the context of "if you've ever played this matchup from the other side", and there's a lot less of a detached focus on "beating X thing" without context of why X thing was done in the first place. This tends to make his lessons pretty memorable, since you get some pandolfinisms which are easy to remember (e.g. "Falco is a lot like peach, he cannot run very fast so he has to defend himself with moves").

The other teachers obviously still care about what the opponent's goal is, but their approach on teaching is vaguely more what do I do focused, rather than Fiction's slightly more what do they want approach. I don't really know how to describe this nuance any better than this, other than that I typically come away from fiction lessons thinking more about the other character rather than my own.

Anybody who has spent even a vague amount of time in a social circle even mildly aware of who Druggedfox is probably knows that he loves telling players to "look at the screen". This short phrase encompasses virtually all of the Druggedfox school of thought for Melee. It's very common for Melee players to use things mindlessly that they believe are good, to not know what they are doing wrong, and to get stuck. For Druggedfox acolytes, there are only two things you have to do to never stop improving - perceive every single interaction as it happens, and understand every single interaction.

As a goal to aspire to, it's certainly very appealing. The platonic ideal of a high level Melee match is that both players jockey for favorable, unreactable mixups, and every single mixup played is a genuine and meaningful outplay.

This, if you go and try it, takes a lot of focus to actually do for every moment of a complete game. But you start doing it, and you notice that a lot of times your opponent does things which you can just react to and punish for free, and that every so often you will randomly, inexplicably do something which your opponent can just react to and punish for free. There is, in fact, so much of this usually left on the table at most levels of the game, where Druggedfox and his school of thought care about virtually nothing else. If you can capitalize on all of these things, and your punish game is decently strong, you are already among the best players on Earth.

It is worth noting that improved tournament performance as fast as possible is not really the aim of the druggedfox approach, which makes sense when you remember Druggedfox has never really been a player who focused much on performing well in tournament⁵. The aim of the druggedfox method is to turn you into a player who knows what to do, correctly, as often and quickly as possible, at any cost. Whereas other teachers will make concessions in order to make things easy to implement at first, students of Druggedfox are typically cast into the abyss until they either perish or emerge being able to react to things.

The way I've heard it described by some players is that while some coaches teach you to play chess, Druggedfox teaches you to solve chess puzzles. In a lot of ways, the way Druggedfox approaches teaching melee reminds me a lot of Michael de la Maza's book Rapid Chess Improvement, which I remember being a pretty big influence for me before I gave up playing chess in tournaments⁶. The thesis of this book is that virtually every game below expert-level is decided by short combinations, from three to six moves long (either missing these or finding them). So, if you are below this level, you should do literally nothing except for solve chess puzzles thousands of times, over and over, for hours every day until you virtually never miss these combinations in a game ever again.

The comparison to De la Maza's book isn't perfect (De la Maza famously didn't know anything about openings or long-term planning, he would just try to remain active and then wait for a tactical combination to randomly emerge), but it's fairly similar in my view. Since the vast majority of matches are decided by one player not paying attention, the most important skill to develop is learning to pay attention - everything else follows after.

Both New Jersey style and MIT style value simplicity, but the difference is that New Jersey style cares more about the simplicity of the implementation, compared to MIT style caring more about the simplicity of the interface.

Both New Jersey style and MIT style value correctness. New Jersey style is willing to concede correctness if it's more simple to do so, whereas MIT style does not tolerate incorrectness.

Both New Jersey style and MIT style value consistency. In MIT style, this is as important as correctness, and a design is allowed to become more complicated if it makes everything more consistent. In New Jersey style, it is considered better to ignore less common circumstances rather than introduce complexity.

Both New Jersey style and MIT style value completeness. MIT style attempts to cover as many situations as can reasonably be implemented, whereas New Jersey style is willing to sacrifice completeness if doing so preserves simplicity.

I find the idea of "worse is better" to be a useful analogy to how players teach and learn Melee. In particular, I think the different teachers I discuss in this essay have different priorities for their students along this NJ-MIT axis; There is a dramatically different focus among them for what I refer to as "getting a weapon into your hands".

Very broadly speaking, I think that teachers like iBDW are more oriented towards getting their students to win games, compared to teachers like Druggedfox being more oriented towards getting their students to make good decisions. Students of both of these teachers often have a sort of low-key derision towards the other, and I think a lot of it can be explained by framing it as this "worse is better" discussion.

These two goals seem like they should be entirely aligned, but the way it ends up manifesting in their students is that the former tend to wield tools they don't fully understand (sometimes using them in bad spots), and the latter understand more than they are really capable of wielding at their current level (which makes them frustrated).

Personally speaking, I found the greatest improvements in the shortest time came from adjustments suggested in lessons from fiction or iBDW, even if the frameworks provided by those two were sometimes not entirely explained to me. Likewise, I found druggedfox to be the best choice when there were specific things I did not understand, since he tends to dive very deep compared to the other two about particular nuances to a situation.

I think the teacher you select should be based around the type of advice you are hoping to receive; if you want a lot of adjustments which sound like "braindead idiot nair is very strong vs sheik" then the former two would be a better teacher, but if you want to know why some nairs are bad and some are good then Druggedfox typically provides the additional nuance at the cost of being less targeted at immediately winning you as many games as quickly as possible.

It boils down to a difference of "know what to do" versus "know what I am doing" which in my experience as a competitor typically jump out at me as distinct problems when I need one or the other.

As far as games go, Melee has a very strong self-teaching culture, which leads to some funny quirks about the way largely self-taught elite-level players teach in their lessons. Reaching such a high level in such a competitive field requires literally thousands of hours of practice, most of which is laser focused on the worst-performing parts of yourself. Most fields do not have people this high-level offering lessons from 25-50 dollars.

I think an almost ubiquitous thing about melee teachers is an approach which bears a lot of similarity to how the teacher teaches themselves (which is understandable, since that approach let them reach such a high level). Most players, when looking at their own games, make a list of things they did that they thought were bad, and things they could do to change those things, and then go to work trying to implement those changes. It is a very normal, very healthy way to analyze your own videos, and the ability to identify things which need to be improved is the way you can continue improving over and over.

However, if you watch a video of a lower level player, there's a strong temptation to look at everything, explain why everything they did was bad, tell them to change everything, and then watch them drown in the despair of how everything they are doing is useless. The reason this doesn't perfectly translate from self-teaching to other-teaching becomes clearer when you think of it like a Rumsfeld speech.

1. Things you do, which you understand (known knowns)
2. Things you do, which you don't understand (known unknowns)
3. Things you don't do, which you understand (known unknowns)
4. Things you don't do, which you do not understand (unknown unknowns)

The purpose of self-analysis, generally speaking, is to A: take things from 2 and move them to either 1 (good) or 3 (bad), or B: take good things from 3 and move them to 1. You either develop understanding, or you implement stuff you don't do yet. The more understanding you develop, the more stuff naturally trickles in from bucket 4 into bucket 3.

The good thing about seeking out a strong player as a teacher is that they understand more than you, so they have better insight onto which things you should learn and implement. Maybe you've been struggling with how to beat a specific thing falco does, and none of the ideas in your bucket #3 seem like they can deal with it – the solution lives somewhere in bucket #4. A strong player with a much larger pool of things-understood might see it and immediately recognize that you should add X, or remove Y.

The bad thing about seeking out a strong player is that, sometimes, they understand everything in your bucket #2. And sometimes all of these things are so distracting that the teacher spends the whole time trying to address these things that they completely ignore the biggest advantage of getting a strong teacher in the first place - understanding things which are high-value that you do not understand.

This has mixed value for high level players, for whom this type of optimization is very very important, but it's often a mistake for players who don't know what they are doing in the first place; the teacher thinks they're giving the student a lot of stuff to work on, but the student comes away with a booster shot of disgust in their own play, alongside a generally-unaware sense of what order would be most effective to work on these things.

In other words, they leave realizing they are much worse than they thought, rather than being really excited to work on things. As a result, their improvement hinges more upon whether they can successfully navigate the emotional turmoil, which will inevitably lead to higher attrition rate.

In particular, I think that the melee community's predilection for calling everyone and every thing "bad" has some very specific, very damaging psychological effects which nobody ever acknowledges within our scene, specifically because everybody has this weird obsession with treating ourselves like emotionless machines.

In psychology there's a concept called self-discrepancy theory which proposes that individuals model their internal self-image into three selves: Actual self, ideal self, and ought self. These three self-concepts usually differ in some capacity, and the way they differ typically gives rise to the types of negative emotional experiences that people experience on a day-to-day basis.

There are two discrepancies that I want to highlight: distance between your actual self and your ideal self, and distance between your actual self and your "ought" self.