Revisiting the Game Analysis Process

Photo by Ryan McGuire. CC0 1.0

It's been a couple years since I wrote my first article on analyzing your games - 4 Steps to Analyzing Your Game for Improvement.

Since writing that article, I've analyzed a few of my games as well as discussed the topic with other players, including some chess coaches and masters.

In today's article, I wanted to give some practical advice for those seeking to improve their chess through analyzing their own games and revisit and refine a few of the points I made in my original article.

If you haven't read the original article, you may want to check it out for more context as I will try not to be too redundant.

In addition to this article, I also created a video series where I played a game online, and then subjected the game to the game analysis process. You can watch the original game with my comments here.

Self-Honesty 

One thing I've observed in myself is that when I'm "self-annotating" my games - the step of the process where I describe my thoughts, analysis, and feelings during the game. - is that sometimes I tend to paint my thoughts more positively then actually happened. Oversights become "misevaluations." Bad moves become "inaccuracies." Blundering material is described as "not getting enough compensation."

It is important when we want to improve to get an accurate assessment of where we are in the development process. Sometimes, this can be embarrassing because we have a certain expectation of ourselves. However, it takes humility and confidence to admit our faults and mistakes. We have to get over this hurdle though if we want to get to the next level.

In the first step of the analysis process, try to capture all of your thoughts. Sometimes, you will forget everything you thought about - that's okay! Even it is only a few moves that you are very clear about - the key is to try to figure out how you think. The fact that you forget what you were thinking of specific moves is insightful in itself!

What are Key Positions?

In the second step of the analysis process, we will try to identify a few key positions for further analysis. I describe in more detail the types of positions you will want to find in the original article, but there are a couple points I've been emphasizing more lately that I'd like to discuss with you.

The first is positions in which your evaluation during the game differed greatly from the actual evaluation of the position. For example, positions that you thought were drawn that were actually won or lost, and so forth. The reason why this is so important is that our ability to evaluate positions is one of the most important skills we can develop.

When we misjudge a chess position, it can be due to a couple of reasons:

• We lack specific knowledge about certain positions - e.g. We don't realize that a specific endgame structure is drawn, so we push for a win only to throw away the draw.
• Our emotions cause us to be overly optimistic or overly pessimistic of a position. We can look at the example of Kasparov's game two loss in his 1997 match against Deep Blue, where much analysis has shown that Kasparov may have resigned prematurely (although there is some debate over this).

This is one area that the chess engine can at least give us an approximation of a position's evaluation to compare to our own. I will discuss using chess engines in our game analysis next.

In the video below, I apply the first two steps of the analysis process to my game.

Using Chess Engines in Analysis

I've gone back and forth over using the chess engines in analyzing our games. I've currently settled on the view that chess engines can be very useful if used sparingly and in specific situations. I've written before about the limits of chess engines in helping us improve, but here I'll give a few tips on how you might use them to help you in game analysis.

• You can run through the game with the chess engine on but only look at the computer's evaluation. Here, you can see where there are big jumps in the evaluation. This tell you where the big mistakes where made. If your opponent made them, you can see if you exploited them. If you are the one with the blunders, you can find out why. Some programs create an evaluation graph and you can see the peaks and valleys of the graph to identify key positions.
• It's okay to come to your own conclusions, even if the engine disagrees with you. If after comparing your analysis and the engine's analysis, you think your move is better, then accept that. This may seem like odd advice, but the fact is that sometimes the computer engine is wrong...and even when it's not, the type of move it produces may be an exception due to the tactics of the position and may not be generally applicable in other cases. Also, it's important for you to develop your own voice when it comes to chess - although I encourage you to do this in an environment where you have access to good instruction or good books that can guide you.
• Do not accept chess engine analysis if you do not understand it. Remember that the engine will not be with you in your tournament games. If a suggestion by the computer is too complex for your understanding or goes against what you believe are good positional principles, then let it go. You can always revisit the position at a future point when you are more experienced.

I feel a certain joy when I analyze a position on my own and then the computer confirms my thoughts! You cannot have this happen if you first turn on the engine when analyzing your games. I encourage you to do your own work first - at least a little - and then ask the computer for some assistance. 

Remember that your goal is not to find the best move in the position in your analysis. This is a by-product of your work, but your real goal is to uncover holes in your understanding or thought process so that you can correct it! The chess engine should be a tool - like a calculator - to check your own analysis and understanding of the game. 

When you find disagreements between your moves and the engine's, then do not blindly accept the engine's answer and move on. Why did your moves differ and - accepting that the engine in this particular case has the stronger move - how can you think differently (or what do you need to study) to help you find such moves in the future.

Accepting Uncertainty

One aspect of analyzing my games that I've come to appreciate more is that I won't always come up with the answers - at least right away. Sometimes we come up with positions that we just aren't sure what to do with. 

This could be in the opening. You or your opponent may play something that's out of your "book" knowledge and not listed among the various resources - e.g. databases, books, etc. You can spend some time and try to figure out the best response, but if after a reasonable effort you are still confused, it is okay to let it go for now.

Similarly, your analysis (perhaps with assistance from a chess engine) might reveal two or three viable moves in a middlegame position. It's okay if you can't decide which one is the best. Let it "simmer" for a while - maybe even months or years.

Sometimes, our future experiences and studies will cause something to "click" in our minds and what was once confusing now becomes very clear. This is the power of our mind to connect related information over time.

By learning to accept uncertainty in our conclusions, we can avoid some of the frustration that results in trying to find "the best moves" in our post-mortem analysis. All of this being said, it's important to put some effort into finding some answers, as we may encounter a similar position in the future and our efforts in the present will aid us when that happens.

Below is a video of the third and fourth steps of the game analysis process:

Time Constraints and Practicality

Finally, I wanted to share a insights that I've come to realize in studying chess that encompasses the game analysis process in general.

• You won't often have time to do game analysis process in-depth. It's okay to focus on just a few areas in your games.
• You may tend to spend more time analyzing your wins than your losses - resist this urge and seek out your weaknesses (if you are looking to improve).
• Analyzing for your own improvement is different than analyzing and annotating for publication or for other people. It's okay to be sloppy or use phrases or jargon that only you understand!
• Celebrate the insights you garner from your analysis - even the smallest of discoveries!
• Organization is both overrated and underrated. Find a system that works for you. Check out my article on workflows for some ideas.
• Seek answers as if you will be seeing your position in a World Championship match, but accept that you may not have all the answers right away.
• Don't do step 3 (corrections) on the same day that you play the game...sleep on it to gain objectivity.
• Some analysis of a game is better than no analysis. Try to find just one thing that you can improve if you're short on time.

Conclusion

I hope you found these reflections on the game analysis process helpful in your own game analysis. The four step process I wrote about is just a template. As I've studied chess over the last few years - squeezing time in between three children, work, and home responsibilities - I've realized that sometimes we need to take short cuts and truncate elements of the process. 

However, I've also realized the importance of analyzing my games, so I wanted to encourage you to keep up these efforts in your own training.

Until next time, I wish you Better Chess!