How to Analyse Your Games

Okay, that heading was a click bait. I can't tell you how you should do it, because analysing is an individual process where you'll have to develop your own method. However, I'm going to explain my own method, and hopefully it will inspire you to create your own method.

There are two kinds of analysis - the post mortem and the home analysis. 

Post Mortem Analysis

This what we do immediately after the game, where both players together play through the game, exchanging plans and ideas. It's primarily a social thing, but one thing I take home from the post mortem is the mood change moments (as explained below). 

Analysis Goals

My main reason for analysis is to address my worst weaknesses in chess - poor opening play and making too many mistakes. So my goals here are to improve my opening play and reduce the number of mistakes.

Analysis Tools

I use Chessbase, game databases, my repertoire books, Stockfish, and the Chess Cloud Database Query Interface as my primary tools. This is how I use them:

• ChessBase is a database management tool. It makes it easy to use various game databases, and also to organise, analyse and annotate your own games. It's the only commercial tool I use, and it's quite expensive too. If you think it's too much of investment, there are free open source alternatives to explore, such as Scid vs. PC. 
• Game databases are collections of games. There are plenty to choose from, both free and commercial. With a database tool, you can search your databases for players, openings, positions, and much more. I keep three  databases with my own games, one for OTB, one for CC, and one for blitz games.
  
• Repertoire books are a kind of databases which contains the lines and variations you play. I use two, one for White and one for Black.
• Stockfish is currently the best chess engine, and it's also open source and free. It works well with both ChessBase and Scid.
• The Chess Cloud Database Query Interface is a massive chess knowledge database, including an opening book and endgame tablebases. It's useful for investigating opening lines and for game preparations, and an essential tool for correspondence chess players.

Mood Change Analysis

Mood changes are those moments during the game where your assessment of the position radically changes. At the start, the mood is neutral, but if you come out on top from the opening, the mood changes to good. Then, after a bad move that will cost you material, it changes all the way to bad.

It would be helpful to mark those moments with symbols on your score sheet, but unfortunately that's illegal during the game. If your opponent complaints about it, you will be punished with a loss, just as Wesley So in the U.S. Championship.

What you can do, however, is to memorise those moments during the game, and then mark down your symbols during the post mortem. I use the symbols :) for good, :| for neutral, and :( for bad. 

This is not to be confused with evaluation symbols like ± (White is better) or = (equal position), we'll get back to those later.

The first thing I do when I get home after a game is to add it to my ChessBase OTB database, without any comments or lines. Then I add the mood change symbols, both my own, and those of my opponent that I noted during the post mortem.

The rest of the analysis work is saved for another day.

Opening Analysis

For this phase, I use the ChessBase Live Database, my repertoire book, and the Chess Cloud Database Query Interface (but no engines yet). The goal is to find the weak opening moves, figure out why I played them, and what to play instead. If I need to change or add things to my repertoire book, this is the time for that.

The results are some lines and annotations for the opening, to be used for future studies (or as material for a blog post).

Mistakes Analysis

This is the main part of my work, starting with finding the moves where the evaluation changes drastically. Now it's time to fire up Stockfish. I play through the game, and every time the Stockfish evaluation makes a jump, I mark that move by adding the Stockfish recommendation as a new line. After that, compare the evaluation change marks with the mood change mark. If they coincide, it's good. That means that I spotted the bad move immediately after I played it, so now I only have to learn how to spot them before I play them! 

When the marks differ, it's bad news. It means that my evaluation during the game was wrong. That usually happens in positions that you don't understand, and that's often caused by playing an opening that you don't know well enough. 

Next comes finding the mistakes that can't be explained by lack of skill and knowledge. What I look for is ghosts, delusions, missing the obvious and sense of urgency just to mention a few. For me, that's the most important part of my analysis, and that's what I spend to most time on.

Finally, I add the lines I should have played, the lines I considered during game, what I can recall from the post mortem, and other notes to make the game ready for archiving.

Analysing Blitz Games

A good thing with playing blitz online is that you can download the score sheets after the session. I save them in my blitz database, and run them through the opening analysis described above. That provides me with valuable statistics on which openings I play well, and which openings I should either study or abandon.

Chess Engines - A Word of Caution

Chess engines could be a great tool for analysis and study, but over-using it can be devastating (or at the very least a waste of time). As a club player, I know that Stockfish is about 2000 ELO points above my level, so in most cases I will not be able to fully understand the lines that Stockfish suggests, let alone reproducing them in future games.

You'll often see players bringing out their phones and start up Stockfish even before the begin the post mortem. They spend the time discussing how Stockfish would play instead of discussing their own ideas. I'm sure they find it entertaining, but they're missing out of a good learning opportunity.

Imagine that you were allowed to ask a grandmaster for advice just once during a game. If that advice is "play Bd2", would you take it? It's probably the best move, but if you don't understand the idea behind it or how to continue, what good will it do you?