GM Smirnov's Advice: The Four Levels of Knowledge

As mentioned, my new year's goals include going through GM Smirnov's "Grandmaster's Positional Understanding."  One lesson in particular has really opened my eyes.  It's not even about chess per se, but rather about training in general.  Essentially, there's a world of difference between knowing something and knowing something, and we need to train ourselves to get from the one to the other.

He specifically included this because some of his lessons seem simple or even obvious.  "Develop my pieces? Thanks, Igor, I've known this since I was rated 800."  Smirnov's point, though, is that many players (and not just beginners) know they should develop their pieces, yet they frequently play games and fall behind in development for whatever reason.  They chase material, they worry about pawn structure, they try to attack, etc etc.

They might "know" about developing their pieces, but it is not their first thought when they look at an opening position.  Or, alternatively, they don't know how to do it: they might develop their pieces to poor squares, forcing them to move them again, falling behind in time. Basically, until "development" is their first thought and they do it easily and effortlessly in any given position, they don't know it.

This lesson struck a chord with me.  I don't remember even reading it the first time I did GMPU, honestly.  I'm not going to reproduce the whole lesson here, for obvious reasons, but I am going to discuss the general themes, which resonate with much of my earlier thoughts and writings on chess.

GM Smirnov's Four Levels of Knowledge

This is the heart of Smirnov's lesson, and the four levels are as follows:

1. Simply knowing about something;
2. Applying it in practice;
3. Automating it as a skill; and
   
4. Incorporating it into your overall thinking system.

Most of us get stuck at Level 1.  "Yes, I know I should develop my pieces, that is obvious."  This is equivalent to, "Yes, I know I should eat healthy and avoid junk foods.  That is obvious."  If you then reach for a doughnut and pour sugar into your coffee, then you don't know it.  You have read about it, sure, and you can parrot back the advice, but if it's not affecting your actions then it's not helping you.  Simply knowing about it is not enough.

Levels 2 through 4 describe the journey from simply learning the concept exists to fully mastering it.  If I continue to use the health analogy, eating healthy is simple in theory (lots of vegetables, less processed sugars), but it can be challenging in practice.  You might need to cook foods you never have before.  Heck, you might have never prepared meals before, preferring to order out. Your pantry might be filled with chips and cookies, tasty temptations that stray you from the path.  Unconscious habits, like grabbing snacks at lunch with co-workers, need to be identified and modified. It takes considerably effort to go from, "I know I should be eating healthy," to making eating healthy your automatic, default thought when you get hungry.

If we are being honest with ourselves, most of our chess knowledge does not reach all the way to Level 4.  "Don't lose material" gets mastered pretty quickly, because that decides virtually every beginner game.  "Develop your pieces" seems deceptively easy, but beginners frequently violate this... and yet, these same beginners often think they are following all the opening principles!

Heck, for fun, I just checked my last five blitz losses, and two of them showed a lack of development.  In this game, I'm behind in development and blunder as a result:

and in this game I have much worse quality development and soon blunder:

I like to think that as a 2200-rated player, and the author of an opening course, I probably know about developing my pieces ... but it may not be as automatic as I would like to think.

Practical Example

Let's walk through what this framework might look like for learning a chess concept: Good Knight vs Bad Bishop. Note that, while this is covered in the course, these are my thoughts, not Smirnov's.  To get Smirnov's thoughts, you need to get the course.

Anyway, I'm sure you have heard about Good Knight vs Bad Bishop.  It looks something like this:

Taken from my own game: https://lichess.org/tB9dXXXr/white#48

The Bishop stares at its own pawns and has nothing to do, whereas the Knight has lots of juicy squares to jump around and attack things.  Black has a terrible time ahead.

Okay, so that's Level 1 down.  We know about it.  Level 2 is to put it in practice.  This is best done by getting a bunch of games decided by this theme and studying them.  Notice, what did the winning side do?  How did they reach such a dominating position?  What did they do after?  Was it the main factor of the win, or was something else more important?

Smirnov would want to primarily use active learning during this phase, so study the games.  Cover up the notation and play Guess-the-Move.  Predict the continuation.  Figure out the plan to fully breakthrough.  Check alternative continuations that didn't happen in the game.  Find a position that's completely objectively winning and then try to beat the computer.

All of this is applying our knowledge, and like in the health analogy above, we're going to learn it's not so easy.  We're going to reach great looking positions and not know how to breakthrough.  We might like Bishops more than Knights and struggle to realize its potential.  We might balk at trading our Bishop for a Knight in the first place.  In the course, Smirnov shows an example of sacrificing a pawn to establish a Good Knight vs Bad Bishop.  Can you look at an example position and confidently say it's worth a pawn?

This becomes important because books tend to be idealistic.  They show the perfect case of Good Knight vs Bad Bishop, usually when a World Champion faces off against some other lesser mortal. Literally everything goes right... but chess is a messy game.  We're going to play games where everything is not in our favour, where the imbalances aren't so stark, where the evaluation is not so clear.  The more positions we practice, especially as they start drifting from the perfect dream scenarios, the better we develop this skill. 

This entire example is conflating Levels 2 through 4... and if I'm being honest, I'm not completely sure how Level 4 differs from Level 3.  But I understand the general idea: we go from having to think about it (Levels 1-2) to doing it unconsciously and automatically (Levels 3-4).  Our goal is to look at a position and, if it involves this potential concept, to have it jump to our mind without thought or effort.  That's when we know we know it.

Conclusion

I find this entire process distinctively different from "normal" chess training.  Books and courses place a premium on Level 1 knowledge, and some provide additional examples.  I haven't seen any, though, that place such emphasis on training the skill, on working through it until the knowledge becomes part of you.

Smirnov recommends repeating his courses multiple times to get the most value out of them.  Obviously, any author would say that about their own work, but this makes sense given the above framework.  Learning about concepts is not enough; we need to drill them and study them and practice them until we unequivocally have it mastered.  It's quite early on, but I'm getting value out of repeating these exercises for a second time, and I'm very happy I'm doing it again.

Let me finish by quoting from Smirnov's conclusion in this lesson, as these words really hit home and inspired this post:

"This process of automating a certain skill takes from 3 to 6 weeks for different people. Within this time period, from 3 to 6 weeks, you need to train and sharpen it regularly: You need to play training games geared for this specific skill; solve appropriate puzzles where you need to show this skill; and then analyze your results.

You need to know what was good in your way of thinking; you need to isolate what was bad and why it was bad; make the needed changes; and the cycle goes on. Practice it over and over and over again. And when you do it for a few weeks… for a prolonged period of time, you get the certain SKILL automated. [...] You will progress. Some might be faster or while others may move forward relatively slower, BUT you will improve definitely."

I find this inspiring.  Shut up and do the work, and in 3-6 weeks you will make significant progress. That's what we are all after with our chess training.  That's certainly my goal.  Expect an updated GMPU review shortly.  The course is just as good as I remember it.  Or, if you don't want to wait, here's my affiliate link to get it: here.