DD 29: Duckable Chess Course

Duckfest Digest 29

Duckable Chess Course 

After spending a couple of months analyzing my own games, working on a plan to guide me on my chess journey, the realization hit me that I needed to develop my own chess course. 

The idea of a Duckable Chess course sounds ludicrous at first. Don’t worry, my steady but marginal progress hasn’t gotten to my head yet. I’m not deluded into thinking the world is in need of a chess course created by me. Which is fine, because it won’t be a course for everyone. This chess course will be tailored specifically for only one student, me. 

As I went on brainstorming about this idea, it started making more and more sense. My own Duckable Chess Course, designed with one goal in mind: to make Duck able.

tldr;

This article is about my project to determine the most important weaknesses in my chess, what I learned by analyzing over 200 of my own games and how I struggled to turn my data into actionable insights. How it was difficult to determine what was (most) relevant and what was not. Eventually I discovered that the solution was asking the question: Would I include this position in a chess course where I was the target audience? That became a new mindset. Instead of believing I needed a plan, my perspective became: what should I add to the chess course that’s tailored for me? I don’t know if the end result will be an actual chess course or not. Maybe.

Identifying my weaknesses
In my progress update for Q3, I’ve written that I was in need of a new plan. In the first half of 2022, my main project was developing an opening repertoire. By the end of Q2 the project was more or less finished. It will remain a work in progress, but it didn’t make sense to keep it a priority. To keep developing as a player, what I needed was an updated plan. A plan that outlined my priorities and helped me focus on the most important areas of chess.

It was a lot harder than I thought to create this plan. I had quite a lot of data to work with, since I already had developed the habit of always analyzing my games afterwards. The problem was not to find mistakes in my play, it was an abundance of findings big and small. The problem was identifying the most important areas to focus on. 

One of the videos that inspired me was a video by Chess Vibes - How To Find Your Weaknesses In Chess. It wasn’t the idea that did it for me. It was the message that this was what had to be done. If I wanted to identify my biggest weakness I needed to go over my games and go find them. 

Collection of positions
I started reviewing my games (again) and started building my own catalog of positions of interest. 
One fundamental challenge that remained was the trade off between getting good reliable insights for each game/position (analyzing in depth) and getting reliable insights about my overall play (by analyzing many games). 
I discovered that it wasn’t easy to identify structural weaknesses in my play. There are many different kinds of mistakes one can make and it’s very time consuming to organize them. My solution was to accept that it was a hard problem to solve and simply put in all the effort that was required.

Image: screenshot of how I started collecting positions.

In theory, creating a database like this would allow for many interesting ways to work these positions. Early in the process I realized that I wanted a more modular design than what ChessVibes was doing. By creating separate columns for each type of position and for each difficulty level, I enabled the possibility of working positions in batches when I wanted to. 

• I could select all positions where I made a tactical blunder that are easy or standard difficulty and work my way up to the higher difficulties. 
• Alternatively, I could make a selection of endgame positions and study them collectively. 
• Or, I could generate a list of positional mistakes I made when playing the Caro-Kann Exchange variation.
• etc.

Categories
In the column type I decided on these categories
Puzzle (tactical)         Positions that offer a clear tactical solution
Puzzle (positional)    Positions that don’t offer a clear tactical solution but require a more positional long term move. Also used to mark positions where I need to make a better long term decision, regardless of the short term best move (like castling on the wrong side of the board).
Thematic position     Positions that are relevant to the opening, regardless of whether it's tactical or positional. For example, I could select a list of typical positions that can occur in the Queen’s Gambit Declined.
Repertoire           Positions that occur in the main line of an opening that I should practice/know.
Typical pattern         Positions that I should be familiar with because they are typical patterns.

These are the most important ones that I specifically wanted to find and identify. 

Hashtags
Over time I started using hashtags in the description field to help me retrieve them. Examples are #pin, #wrongside (of the board), #hangingpiece, #freepiece, #missedfork, etc. Because I added this later I haven’t used them enough to get statistically significant results. 

Image: Some positions and hashtags I started using.

Status
After reviewing each position, the outcome (or status) of the positions were, broadly speaking, divided into 4 main categories.
1. Position is not solved. In this category are the positions that I don’t fully understand yet. This includes all positions where I have no idea what to do, either because no moves make sense (and they all look equally bad to me) or because many moves make sense (and they all look equally good to me). . 
2. Position is semi-solved. These are positions where I know the best move, with the help of an engine, but I don’t (really) understand why yet. Even when I can see why the best move is good, more calculation and analysis is needed before I can call the position solved. 
3. My mistake is solved, but the position is not. For these positions I fully understand what is wrong with the move I played, but I haven’t analyzed the position enough to know what would be the correct move to play. I’ve put these in a separate category, even though they are unsolved positions. The most valuable part was understanding my mistake and I don’t know if I want to invest more time beyond that.
4. Position is solved. I fully understand what the best move is and why. Ideally, investing time to analyze positions where I made a mistake or a blunder will lead to a better understanding of the game. Because I gained a new insight, I won’t make the same mistake again. Ideally. However, that’s usually not how it goes. Most positions in this category are solved because they are easy to solve, especially when presented as a puzzle. The challenge is not to find the best move, the challenge is how to make sure I can find the best move next time in an actual game.

(I added a few more categories, but to illustrate the principle classification these are the main categories)

Relevance
It turned out learning from my mistakes would require a lot more effort than I anticipated. Many positions I had identified, I had yet to solve. Other positions I had solved but needed more effort to really understand them and do better next time. In other cases I had only solved what I did wrong, not yet what would be the solution. And finally there were positions that I had solved but I needed to practice more to ensure I would do better next time. Basically, there was a lot to do and I needed to prioritize. 
The main considerations were:

1. How relevant is solving/understanding this position for me and

2. How much effort does it take me to understand this position better?

I can’t justify the time investment for every position. Should I invest time in understanding an obscure position, that I will never see again, when I currently have no understanding of the position whatsoever?  For many positions, the answer was to not investigate any further. There are two problems however. 
1. Without understanding a position, it’s impossible to tell whether there is something relevant to discover. In a way, they might be even more relevant than the ones I do understand.
2. There is something counterintuitive about dismissing a position before understanding it. It feels wrong to simply accept that I don’t understand a position and choose to keep it that way.

Yet, with hundreds of positions that require a deeper understanding, I needed a way to choose what to prioritize.

Chess Course

The questions that I kept asking myself were “Why is this position relevant for me?”, “What is the reason I am saving this position to learn from?, “What does this position teach me?”, etc. What finally resolved it for me was asking the question. “Would I add it to an opening course targeted at players like me?”.  This question was more concrete and practical and as a result it became easier to determine the relevance of a position. 

This was the perspective I needed. It may not seem like a profound insight. But for me it was. Instead of trying to identify weaknesses in my play, I started identifying things I needed to learn.  It became easier to accept that not all positions where I made a mistake needed follow up.

Would I use this position in a chess course? No. Other positions on the other hand were a perfect fit. Would I use this position in a chess course for me? Yes, in the tactics course in the chapter called “Discovered attacks”. Would I use this position in a chess course for me? Yes, in the tactics course in the chapter called “mating net”. Would I use this position in a chess course? Yes, in the endgame course in the chapter “Rook and Pawn endgames”.

Alignment with other resources
It made me realize that I could use the same terminology that is used in chess courses and in chess lessons, to categorize and subcategorize all positions I came across. By adopting the existing terminology, everything I did fitted within the same structure. Puzzles that I missed could be added to the collection of positions that I had played incorrectly in my own games. As well as the content of the chess.com lessons. Even the chess courses I did online and the other content I consumed online followed the same structure. 
Everything I learned and still had to learn was now organized in the same structure. All I had to do was add everything of interest to my own chess course.

What I learned (so far)
Let’s start with some numbers. Total games analyzed ~ 230. Position saved: ~875. On average, I have ~3.75 positions saved per game. Which intuitively made sense. If I could in time replace my worst 3-4 moves each game with a better move, my rating would skyrocket. That’s an attractive goal to strive for.

1. It’s all about tactics. 
This may not surprise anyone, but it did surprise me. I had considered tactics one of my stronger areas and this was confirmed by comparing my scores (for tactics and mates) with similar players using Insights. While I might be relatively good at tactics, in absolute numbers it’s still where I make most of my mistakes. That’s why I started doing puzzles and puzzle rushes again to keep improving tactics.

2. I have no weaknesses that stand out
Initially, this was a disappointing outcome. The ultimate goal of analyzing all these games and reviewing all my mistakes and blunders was to identify my biggest weaknesses. Instead of discovering big weaknesses I only found small(er) weaknesses. But I found a lot of them. A lot.

Rather than being discouraged by the (lack of) results, I find it rather liberating. I’ve discovered enough things I should improve on. And when I’m done with them I have plenty more I could do. One step at a time I’ll continue working on my chess. 

Thanks for reading. Next article I'll share some of my interesting discoveries.