I see joseki as being, at their core, a review aid.
When you review the game, you can evaluate your performance in each corner and ask yourself whether the result you gained there was lacking and whether you made mistakes.
If that is the case, then the first question reasonable to ask is “Did I deviate from joseki and when?” and the second “Was this the reason why my result was poor?”.
The third question “Is the joseki move superior?” and the fourth “If it is, why is that so?”
In this way, the review and learning process can be streamlined.
You should give yourself some credit, since this is actually a method used to determine whether moves are good (called tewari). The idea is that you take the end position and look at different ways of reaching it, or at the same position with stones removed to evaluate whether the players really played the best moves. Sometimes there are other ways of playing the same stones that show that parts of your supposed ‘joseki’ were really slack (and thus shouldn’t have been played).
For example, take this sequence of moves that results from a pincer position:
White invades the 3-3, and pushes upwards with (5), upon which Black descends to cut white:
Now Black can cut, to aim to capture either of the two groups of two white stones:
White captures the cutting stone, but black captures the two white stones and keeps the corner.
This is very similar to the 3-3 invasion with the double hane:
But then, afterwards, it’s like White played (A) (which is perhaps a bit slow, but played sometimes as a follow up) and Black responded with (B) (which is just a terrible response to A)
Thus, this shows that the first sequence was bad for Black, since it’s like Black played a bad move after the end result of a completely different joseki. Backwards reasoning!
the board has some symmetry (especially if you play F)
first and second line moves are very small
it’s unlikely that your opponent doesn’t play on any letter.
It can happen though.
Maybe because of a mistake or because of some strategy you can’t dig.
In that case just ask yourself these two questions in this order:
I still remember that when I first got my “joseki book” as a gift from my uncle, I was first pretty excited, Even though I was mostly just put stones on my board according to the pictures in the book since I was still too young to understand long sentences and recognize complicated words.
And the books would show how different follow-ups would become. Sometimes it would become life and death, or tsumego problems,
Later one when I am older and go to a school and able to read, I finally realized that it was like a dictionary with thousands of pages and several volumes (the book title is literally “Joseki Dictionary” part 1, 2, etc, but the Chinese word for a dictionary - 辭典 is a very complicated one, so when I first got the book I don’t recognize what it is yet). It is even organized with indexes and page numbers in the front to look up different variations.
It is essentially the book version of the online joseki dictionary. But I think it shows how inter-connected joseki is to the very foundation of life and death, requiring tesuji to solve the problem reaching certain ends, but most of all, linked to the direction of play, influence, strong/weak groups, and to the whole board position.
You would certainly be able to play without knowing them, but like writing a good poem, knowing prior works and having a dictionary to look up phrases and rhymes will certainly help you finish it more elegantly and faster.
I would say: many deviations from Joseki are suboptimal. Usually, a move that isn’t joseki is somewhat of a mistake. This is more true for some sequences than others. Often, a non-joseki move is only slightly suboptimal, or requires very accurate play to punish. At 20k, most non-joseki moves will have minimal impact on the game.
As you move up in rank, though, more opponents will be able to take advantage of those mistakes. If you play a sequence that gives your opponent a bigger corner than they should have, that’s free points. If you end up completely surrounded, that’s influence. If you end up with a bunch of weak groups, that’s a liability. Playing Joseki generally means that you’re avoiding those sorts of pitfalls.
If both players play Joseki, then the result is usually balanced. Joseki represent ways to avoid mistakes, not to punish your opponents.
Maybe none, locally, but knowing the joseki will give you a base of knowledge to work from and add to and help you avoid and punish mistakes. Globally, however, there may be advantage, so issues of which joseki to select become important. In some cases a sequence of moves may be joseki but completely wrong for one player (e.g. their resulting influence is nullified by opposing stones that are already on the board).
The answer is simpler than most of these: since joseki is perfect play, by definition, the advantage is that your play is perfect. If your opponent departs from the joseki, you are ahead. If you depart from the joseki, you are behind.
Go is not a game of chance: in any game position there is always a single set of perfect moves. Sometimes the set contains only one move, and is obvious. More often, it contains a large number of moves. The set of perfect moves is not a strategy and does not depend on any characteristic of your opponent, since it assumes perfect play by both sides (it is a mathematical concept). The set of perfect moves, in general, currently cannot be calculated in a reasonable time, even by the best algorithms.
If I would give a definition of what is s joseki, I’d say it is a collection of common sequences played after the initial fuseki, and before locally settled positions where their initial steps preserved as many variations as possible.
Think of them as commonly traveled paths, starting from some well-known fuseki positions, collected over the years. And every time a game played (or studied), the footprints associated with that sequence get a little deeper. But if a sequence gets traveled ended up losing the game, the footprints get a little dusted and shallower.
Paths with their initial starting sequence still get traversed frequently are actually those that still don’t lead to a clear winning or losing games yet (most variations are still “preserved” waiting to be explored). AIs make traveling these paths millions/billions of times faster and would be able to cross out or deepen some of the paths, even exploring previously not well-traveled paths and make them common. But even AIs are still exploring new routes and keep digging deeper and further (and higher).
Joseki are locally perfect play. In practice it’s perfect play as far as humanity knows for now. If a joseki is discovered to give advantage one side, it’s not joseki anymore. If a sequence that was previously thought to be undesirable for one side was discovered to be actually ok, it becomes joseki.
So take it as your opportunity to shine.