Organizing my Openings part 1 Practical Progress

Duckfest Digest 09

Organizing my Openings part 1: Practical Progress

Introduction

After 20 years of absence from the game of chess there were many things I needed to work on if I wanted to improve and reach the goal I had set for myself, to become a 1500 rated player. In fact there were so many things I needed to work on, my whole methodology became a mess. Somewhere down the line, I made the decision to develop an opening repertoire. In my article on this you’ll see my full reasoning.
Essentially there were three main reasons:

1. I was playing new openings
   I started playing openings I was totally unfamiliar with, so I needed to develop a better understanding of them, know more lines and get more experience in playing them
2. I made early game errors
   Early in the game I played inaccurate moves and sometimes made small mistakes, creating difficulties further in the game. My lack of progress in Blitz is a direct result of my lack of readily available moves.
3. I was looking for structure
   Additionally, in the first months of my chess improvement journey, I was also searching for the best way to manage and organize everything. Understanding of middle games principles might be more valuable than understanding openings. Working on an opening repertoire is much easier, especially from a practical standpoint.
4. I enjoy studying openings
   The three reasons listed above are my official reasons. The fourth reason might actually be the main driver, because that is the sole reason I have dedicated more time to building my opening repertoire than any other aspect of the game. And enjoying myself is ultimately why I am here on this site, playing, studying and learning.

About this series
There are plenty of other sources, many of the GMs, that provide insights on creating a good opening repertoire, with an emphasis on opening repertoire. They cover openings in depth, discuss strategic considerations and even provide cool tactical tricks. My focus will be on building an opening repertoire, with an emphasis on building it. No chess related advice, just me sharing my experience with the whole process.

Chapters
Organizing my Openings part 1 - Practical Progress (this article)
Organizing my Openings part 2 - Conceptual Considerations
Organizing my Openings part 3 - Repertoire Reliability
Organizing my Openings part 4 - Practical Playability
Organizing my Openings part 5 - Structural Struggles (postponed)

Repertoire Reality
When I started with this process, I wasn’t even that ambitious. All I needed was to create a collection of moves to play in a given position. The idea is simple: for any positions I want to know which move to play next time, instead of the move I had played earlier. My perspective at the time was that an opening repertoire essentially is a giant table of positions with a preferred move for me to play for each position. That’s what I started working on. I expected this to be challenging, requiring a lot of effort and dedication. But slowly but steadily I would get there, as the process was simple:

• Analyze a position
• Determine the best move
• Find a way to store the move into my brain (e. g. memorize it or practice it enough times)

There are a couple of things I undertook to find the best moves. I bought a GothamChess course (for the Caro-Kann), on the London system I consumed a lot of YouTube content. Sometimes a quick engine analysis was enough to eliminate mistakes and replace it with a better move.

Candidate moves
Fundamental to my approach is the concept of candidate moves. The idea is simple: in any given position, you identify the moves you can play and then make an assessment of which move is best.

I started to collect notes and log my findings, initially in PGNs as well as Google Documents. I haven’t yet mentioned my failed attempts with Google sheets, but I still have a huge collection of unfinished calculations on my laptop. What overwhelmed me was the immense amount of positions I had to organize. I didn’t expect this. While playing, even though every game was different, there were definitely patterns to discern. But the closer I looked, the more I realized that positions that look similar at first glance, can be different from each other under closer examination.
Basically, there were many candidate moves to evaluate for each position. For each candidate move, my opponent had multiple moves they could play. And in each one I had multiple candidate moves I could play, to which they could respond in multiple ways. Very soon, it became too much to manage. From a theoretical standpoint this is unavoidable and, given my reasonable grasp on mathematics, I should have not expected otherwise. Yet, there is something unintuitive about exponential growth.

Therefore, I decided I should be very selective. My intention was to prioritize the most important positions first, shifting later to positions I had not encountered yet but I would expect to encounter sooner or later. Many games were irrelevant, openings wise, and I wanted to exclude lines that were rarely played. Even though I tried to be selective, the sheer magnitude of possibilities was insane.

Chess Position Trainer
Very soon after I discovered Chess Position Trainer (CPT), I realized this was what I was looking for. CPT is based around positions, not games. This means notes and comments on any positions are shared, independent of opening or how you got there. Like global stats. Finally, I had one convenient place to dump all my notes on every single position I’ve ever looked at.

About CPT - In CPT you select candidate moves and sort them in order of relevance. Most of the time, you allow your opponent multiple candidate moves while you try to limit yourself to 1 preferred candidate move for each position. Transpositions are handled great. If you ever get to a position you have analyzed before in an entirely different opening, you will have all your previous notes available, even if you only analyzed the position from the opposite perspective (color).
You create multiple openings, depending on your goals and are allowed to arrange the move order of candidate moves separately for each opening. You can even create a different repertoire for Classic and Bullet. There is a downside. Since all notes are shared over all openings, you can’t make separate notes for each. You also can’t make separate notes for white or black. (I’ve tried to circumvent this by running multiple CPT instances, each with their own database. This is not recommended, as I’ve run into saving and backup issues).

The philosophy of CPT is that you invest in creating your repertoire first and then later use the Training function to start training/memorizing the moves you have selected to be your preferred moves. The training was not important to me at first, I didn’t even expect to use it. Over time, I did start using it. It’s a valuable feature.

Researching my Repertoire
With Chess Position Trainer as my main tool, with Google Documents as my backup for things that didn’t fit the program I was starting to build my opening repertoire.

Caro-Kann
For the Caro-Kann, I was using Levy’s (Gotham Chess) course. He might have been the main reason I chose this opening in the first place. I had considered a couple of options. Earlier I had experimented with the PIRC but that wasn’t working for me. I needed something else. There was a general consensus that the Sicilian was to be left alone until a much higher rating, so that one fell off. The Caro-Kann had an instinctive appeal to me. It was introduced to me at an early age (probably somewhere between 9 and 11 years old), but at the time appeared too intimidating and complex. To take on the Caro-Kann, an opening that young me thought to be too difficult, had a certain appeal to me. What convinced me was that the opening seemed to match my rating, experience and skill level. After some research to validate my decision, I concluded this was the perfect match.

My decision to use his course was perfect, as it gave me a great kick-start. Levy provided structure to my opening repertoire and gave practical tips as well. He explained there were only 3 main variations in the Caro-Kann, which provided a good high level structure to organize my openings. While I later experienced more common variations, the premise that there are three ‘main’ variations still holds up.
The other benefit of using his course was that he provided specific playing tips. Instead of objectively going over each move, he was subjectively suggesting moves. This was huge for me as I was overwhelmed by all the possible lines I was facing.

My initial results were promising. The main position in the Caro-Kann is 1. e4 c6 2. d4 d5. My stats when playing the main variations are.

My plan to build an opening repertoire started paying off almost immediately. Over time my winrate stagnated. My understanding of the Caro-Kann improved over time, but at the same time I started facing better opponents. I will address that in one of my next blogs. For now, let’s celebrate that I’ve started building a repertoire and I’m doing well playing a legendary opening that was completely new to me 4 months prior to this.

London. Baby

My decision to play the London System with white was made very early after I got back into chess. I’m not sure exactly how, but I know Eric Rosen was an important influence. After some research, I had made my decision. My reasoning was simple: the London is solid while still allowing some great attacking opportunities. What really sold me were two other arguments:

• it’s more of a system than an opening, thus giving some freedom in move order and room to experiment. As a result, I could play the London immediately, without any opening knowledge other than the key concepts.
• The London System provided a solid structure but, especially at my rating level at the time, still very much allowed for a game of chess to be played. The option to delay castling, to maybe castle Queenside or maybe not castle at all, was exactly what I was looking for.

My experience with playing the London was great from the start. My winrate in the first two months was over 60%.
(There was one major drawback to playing the London when trying to build a repertoire. It’s not possible to distinguish main variations like you do in the Caro-Kann. Everything transposes into everything. Which I will address later, either in my article Conceptual Consideration or in my article Structural Struggles)

vs KID/PIRC, turning the tables
My performance, playing the London System, was very good from the start. In some lines, however, I wasn’t doing well. When my opponent played d5 on move 1, like a normal person, I’d have a winrate of 63%. Instead, if I faced a maniac that played Nf6 on the first move, my winrate would drop to 38.5%. Since d5 was much more prevalent, I had overlooked NF6 and didn’t notice how much it affected my overall winrate. My focus on building a better opening repertoire helped me spot early game weaknesses in my play. At some point, I noticed I didn’t do well against Nf6 and I needed to find out why.

(I’m almost laughing out loud as I write this. At the time I had absolutely no idea of the importance of Nf6.

I was losing to lines that started with Nf6, especially lines that were followed by g6 Bg7 and d6. Or in other move orders. I’ll call them KID/PIRC defense. Technically, they are not King’s Indian Defense until I play c4. Technically, they are also not PIRC if I haven’t played e4. Anyway, I’ll call them KID/PIRC defense.
This was one of the most significant discoveries in my opening repertoire. There were certain, fairly populair, openings that were completely crushing me if I didn’t adjust to them.

My current line looks more like this, with a 71% winrate over the last 6 months

Wrap-up
This is how I started building my opening repertoire. The foundation has been laid.

This is not the whole story. It sounds simple to just start collecting best moves but there is much more to it. I don’t even use the term ‘best move’ in my own notes. Because there is no such thing as a best move. What is a best move to a chess engine might for a human be a total minefield. The opposite is also possible, a move that’s borderline suicide against a chess engine might me devastating to a human opponent.
Stay tuned for more.