There are 2 or 3 points, depending on how you look at it.
"Someone who knows exactly what Waldo looks like will find him faster than someone who knows there is something with a red and white stripe involved."
"You can’t recognize something fast if you don’t know exactly what it looks like."
First, there are many already established books which go over this. Nimzowitsch's My System, Silman's books, and John Nunn are a few. Those books are extremely boring to go over because of that slow incremental advance.
I would rather play a game, enjoy the game, and then run it through Stockfish to see what moves I could have made had I been a good Russian schoolboy and sat next to Yermolinsky in the front row.
Another slant to this is that you don't know where to start. It is arbitrary to focus on tactic 1 of 100 which has the same importance as 2-100.
The learning experience should have a snowball effect, and you can't do that with the "Waldo" approach. You should be making a Snowman, not looking for Waldo. I'll explain further.
The third idea is that puzzles are a way to learn how to move pieces. They aren't the true objective. We have different camps of people in this regard.
Camp 1 are the weightlifters. These are the people who do crossword puzzles or spend time watching Jeopardy to just know individual facts. If you are geeky in math, great, tactics won't be as frustrating as much as it is to others.
Camp 2 are the builders. These are the people who are engineers, artists, and managers/leaders, who put together ideas for a temporary purpose. The weightlifters are looking for absolute 100% muscle growth, nutrients in and muscle formation. The builders on the other hand assess what boundaries are to be met before working on something.
Camp 3 are the explorers. These are people who are looking for improvements. This could be a low beginner who has played other board games, or a high beginner who has played chess, or it could be a GM helping another GM prepare for a match. They are not interested in the final result initially, they are looking to find new ideas to explore and postpone judgment until they have exhausted their search.
You have to ask yourself where you lie? For myself, I am 50% Camp 2, 30% Camp 3, and 20% Camp 1.
If I were training for a tournament, I would not change a thing. I would embrace the noise. You have to learn to subtract from the confusion what sensible things to do.
The building snowman effect comes in to play when you decide how much of an opening you want to make (direct Italian or QGD transpositions for example). How much of a middle game do you want to explore (trade to get to endgame or not to focus on more middle game moves). This doesn't have to fit the weightlifters' agenda. It doesn't have to be Stockfish worthy. You are learning how to do things based on that position and not worrying about how wrong you were 10 moves ago. This is something puzzles don't provide. There are no scenic routes.
Ironically, that is what makes it boring for me because I don't like doing puzzles I don't know where the game came from. I like to see how the game came to that position. I would prefer to see commentators talk about how the positions they talk about also develop and not just comment on the position at hand.
To wrap up, do you want to continue looking for Waldo or do you want to make snowmen? How much of a gain in your life is solving a puzzle?
Thanks for the thoughtful reply. I completely agree that this approach won’t be for everyone, but for the subset of chess players who want to learn the basic building blocks of tactics as fast as possible I think it’s the right approach (and I think that’s a decent size subset). Of course all these tactics will have to be integrated with positional and strategic concepts. Like Uncle Bobby said, “tactics flow from a superior position.” It’s not like you need to do this at the exclusion of learning openings and strategy or just playing for fun and experimenting. It’s just when I’m learning, I want efficiency. I spend time practicing guitar and playing and efficient practice gives me more time to enjoy the fruits of that practice when I’m playing.
You said you wanted to embrace the noise and learn to subtract the confusion. You are speaking as a more advanced player (I’m assuming) so the challenge IS the noise. For a beginner the challenge is even recognizing the signal when there is no noise. Once you’ve learned that you can begin adding noise and enjoy that aspect.
I also vastly prefer to see how a position came to be vs an abstract puzzle. That’s part of the reason I want to make the tactical aspect as efficient as possible - cause I’d rather be doing something else. In fact that’s what I’m most interested in right now and have been looking for sources . I’ve looked at Logical Chess but I found following the annotations too tedious. If I can’t find good videos I’ll go back to that. John Bartholomew is pretty good. If you know any others that are good, I’d be interested in hearing.
I’ve only been studying chess for about a month but I’m very familiar with accelerated learning and how the brain learns patterns. I’ve been so frustrated with the tactics trainers I’ve tried thus far (Chess.com, Chesstempo, Chessable, and a few others) because they are incredibly inefficient (at least for beginners). I’m wondering if I’m missing something, perhaps a setup, or a resource that works with how we learn efficiently. I see some incorporate spaced repetition, but they repeat the wrong things.
The problem is these trainers all introduce noise before the signal is mastered. You need to separate the skill of knowing a pattern from finding it. Even the easy level on chesstempo has nearly a full chessboard, most of which is noise. You can’t recognize something fast if you don’t know exactly what it looks like. Someone who knows exactly what Waldo looks like will find him faster than someone who knows there is something with a red and white stripe involved. It’s critical to start with the barest bones of a pattern (the signal) and then gradually add either more signal or more noise. Let’s take knight forks for instance:
The first thing you have to master are the 6 (?) basic patterns that allow for a knife fork (diagonal touching, orthogonal 3 squares between, etc.) . You show them to the student and give them names. Then you flash patterns of two pieces, some forkable, some not, and just click yes or no (or the given name) You know you’ve mastered it when you get 100% correct in X seconds. After that’s mastered, you add some noise (other pieces) a little bit at a time, each requiring a mastery threshold, until you’ve got a nearly full chess board.
Next you remove the noise and add a knight to learn it’s spatial relationship to the fork pattern. Is it forkable on the next turn (there would be basic knight movement pattern training that would precede this, of course.)? Eventually you add enemy pieces that could foil the fork.
When you want to move on to intermediate forking (creating a fork rather than identifying one) you go back to two pieces on the board, where an enemy piece is one move away from a forkable pattern and have them move it into the forkable pattern. Now you’re learning what situations allow for attraction/deflection, etc. that create forks and how to avoid putting yourself in one. One of the reasons so many beginners say they can find forks in tactics trainers, but not in games is because they lack this skill.
You might think that the highly granular increments would make it take forever to learn anything, but this isn’t the case because these initial building blocks are learned very fast. You will learn the gestalt faster if you integrate individually mastered elements. I think someone could get to the same level of tactical expertise in one fifth the time using efficient methods vs the ones I’ve seen thus far.
I don’t think this is a monumental programming task because the basics are so simple. Does such a program exist or do I have to fund and create this damn thing myself? :)
EDIT: I apologize for the angry tone. Just finished some tactics training and was annoyed.