Starting From the End

Published on 21 Mar 2020

As I wrote in the February 2020 update, I started studying the endgame a few weeks ago. My plan was to go through all of 100 Endgames You Must Know before writing about the endgame but, you know, plans are meant to change. So I’ll start from what changed, then I’ll share a mental model that’s helping me making sense of the vast theory behind endgames, and finally a quick update about where I’m with my endgame studies and where I want to go from there.

Plans always change even when they don’t

First of all I had to adjust my expectations about how much I would enjoy studying the endgame. For some reason, I expected to enjoy endgame studies very much. But in reality I’m finding it very difficult so I don’t look forward to sit down and study endgames as much I thought I would. The truth is that endgames are really hard. If even Magnus Carlsen may have his doubts about a position (see the incredible Magnus Carlsen Takes the 100 Endgames Test!), it’s probably normal that an amateur like me feels stupid while studying some position.

Once I got past the initial shock, it got easier as I started seeing a new way to approach endgame studies. I’m always looking for ways to break a complex problem down in smaller bits that are easier to manage so endgames theory has been no exception. Here is how I’m clustering endgame positions:

  • Solved by exact calculation
  • Known answers to “what’s the idea here?”
  • Known solved positions

More about this in a moment.

So, as you can see, the plan did not actually change but it feels to me like it did because I had to slow down a lot (I thought I could go through all of the book in a month or so for example). On the other hand, studying endgames is giving me some motivation. I started winning completely drawn positions now and I suppose the reason is because, at my level (1600/1700 on lichess), knowing the basic ideas behind an endgame makes already a big difference in.

A mental model for endgame theory

I’ll go over each category I just discussed but, before doing so, let me say that I’m not aiming for a perfect categorization. I don’t even know if this is a mental model I will want to keep just in a few months. The goal is to chart the endgame territory (this website is called Chess Atlas for a reason, isn’t it?) with a draft of the borders as a starting point. It’s also worth noticing that the categories I’m using overlap. The mental model is naturally forming in my head but that does not mean it’s the right model. It most likely isn’t. I literally just started studying the subject. So take the whole thing with a grain of salt.

Solved by exact calculation

We all know that endgames are hard because you need engine-like precision. Sometimes the difference between losing, drawing, and winning is literally one move. The good news is that there are a number of situations in which calculation can be simplified with technique (sometimes to the point you do not need to actually calculate anything). The concept of key squares is a good example, the rule of the square is another good one. These endgames become simple once you know the right technique. Just to clarify a little the category with an example, you do not need to count to know this is a won game:

The point is that there’s a rule that helps you calculate precisely the outcome of some positions. I believe I need to learn these systematically and practice them often.

What’s the idea here?

I read the paragraph title with @agadmator voice in my head (it’s a tribute of course!). Some endgames have a core idea that you need to know. This concept actually hit me even before I started studying endgames. I remember the first time I saw how to mate a king with two bishops I realized some endgames would be way too hard to figure out over the board. Some ideas just just need to know. My favorite example involves an under-promotion and once you have seen the idea you’re going to remember it (black to move):

The only non-losing move in this position is an under promotion to a knight. Now due the fact that a newcomer to the game have a materialist tendency, it’s hard to image they’d come up with the idea on under promotion. Especially if they run into such a position for the first time.

Known solved positions

This category is not so different from the previous one. The main difference is that here I’m thinking about known positions with a name. The VanĨura position is one example, the Philidor position is another. Another difference about this category is that it’s not always an idea you have to be familiar with, these known positions often require you to know an exact sequence.

As knowing what’s the idea here may not cut it for these positions, I believe these require constant training as well. Right now, I find that sometimes it’s hard for me to recognize a given position is indeed a known position and I don’t have a plan for that just yet.

What’s next?

This categorization is very personal and I’m not sure it will make it to the atlas. I just started studying endgames and the theory is just too vast. This mental model helps me approach the problem in a way that works for me. I’m sharing because it may also help someone out there. Now, as promised, a quick update about my studies.

First, I’m almost done with 100 Endgames You Must Know. I will write review about it in the reading list but I can already say the book leans toward the 5-star review. I’m enjoying it a lot and I’m sure I will go back to it often in the coming years. At the same time, I’m also reading the soviet chess primer and this incredible book (5 stars aren’t going to be enough) has a very solid chapter about the endgame and a great chapter about calculation techniques (and most of them are must-know for endgames). Once I’m done with these two books, I will spend some time reviewing the endgame theory I learned. I plan to write about most things I learned so that I achieve two things at the same time: I enrich the atlas and I figure out where I need to go deeper. Before even going through a review phase, I’m pretty convinced rook endings will require deeper study. I’m finding them particularly hard and I’m not sure why just yet.

There’s also another idea floating in my head at the moment: adding an interactive training section to the atlas. It comes from my professional background: I would love to find more reasons to do some chess programming, it’s a fascinating subject. I can’t say I will do this but your feedback may push me into looking for more interaction between the reader and the atlas.