We’ve gotten this far without even mentioning two of the activities best correlated with chess improvement: playing games and analyzing them. There is a reason for that. Part of the point of this guide was to emphasize the importance of immersing yourself in good patterns, and playing games yourself - especially against similarly-skilled opponents - does that only very slowly. That said, however, the point of chess improvement is to positively impact your actual play, and you can’t assess something that you can’t see. You have to actually play the game in order to identify where you need work and prioritize it appropriately.

That last sentence encapsulates how someone serious about improvement should approach playing games. Playing the game is fun, and recreation is usually the ultimate goal, but if you have specific goals around improvement then playing and analyzing your own games has a singular role: to assess where you need work. The proof is in the pudding, as they say.

I used online games for this assessment. The primary purpose of these games and their review was to assess where my intuition was working and where it was not, because that’s what I wanted to work on. As I result, I wanted to identify three things:

1. positions I was reaching where my intuition was blank - I had no preconceived notion of what to do
2. positions I was reaching where my intuition was wrong - I an idea of what to do, but it was actually a bad idea
3. positions where my intuition was incomplete - I had an idea of what to do, and it wasn’t bad but there were other ideas I didn’t consider

Because my singular focus was on developing my intuition, I would use a variety of time controls, even bullet. Bullet has a terrible reputation for improving chess play, but if you don’t play it to win but rather to assess your own intuition, it’s a great time-maximizer. Just don’t care too much about your bullet rating (probably good advice generally).

Now, I’ve never been a very good blitz player, and my bullet play is actually a joke, but that’s not the point. The point was to assess lots of positions and what came to mind quickly in those positions. My first review of a game was ideally done without Stockfish - just identifying moments where I didn’t have an idea of what to do, to draw out (1)-type positions for myself. Then, I would run through the game with Stockfish, and do my best to draw out (2)-type and (3)-type positions.

I would then make a personal grade for each of the biggest problem positions in a game and make an addition to a running list of “Things I have to work on”. This was a list of broadly-categorized position types where my intuition was relatively weak. I kept this list in a spreadsheet, although you could use almost any notes application. I used headings like “Sozin Middlegames”, “Defending Isolani positions”, or “Rook Endings” (that was one I came back to many, many times). Critically, I would make sure I added the link to my just-played-game as an entry under the relevant heading, along with a grade indicating the degree of problem that game revealed. Sometimes I would add the same game link to multiple headings. In this way, I started building out data showing where my gaps in intuition were most problematic. After a given session of game-playing, I would take a look at the list of things I had to work on and develop a plan to work on the ones that were showing up most often and most seriously.

I had three ways I had figured out to work on one of my gaps in intuition: basic analysis, model game changes, and sparring. I will describe each of these methods as follows.

First, basic analysis. When I started to work on one of my heading groups, I always had a bunch of links to games where I had identified the problem, which was a big help. I would look through the games, analyzing them with Stockfish, and think about how I should have played each one. After I had gone through them, I would make a decision about whether something more than that review was required. Sometimes I decided it wasn’t - maybe I had a clear idea at that point of what to do and thought that next time I encountered that type of position it would go better. I would gray out the relevant game links in my spreadsheet (but not delete them), and move on to something else. Later games showing those same issues would go under the gray ones of course, and would often encourage me to take stronger measures.

Second, model game changes. The strongest measure I had at my disposal for addressing problematic gaps in intuition involved some modifications to the model games I was drilling. I could modify a custom model game to better explore the theme I was struggling with, or more commonly I would just modify the “talk track” that I would use alongside a game to help me remember the key points I was forgetting in games. If an issue was serious enough, I would create a new custom model game specifically oriented around exploring ideas I was consistently missing. All of these modifications to the custom model games were relatively “expensive” however. Every additional custom model game required real work to memorize, and modifications to a game disrupted the memorization process so I really tried to keep these changes rare.

Basic analysis was usually easy but light on impact, and model game changes were much more impactful but involved. Fortunately, I also developed a middle-tier solution for intuition development that I called “sparring”. That’s the final technique that I’m going to cover in the guide.