Chess Ultralearning — The Best Way to Learn Chess in Under 100 Hours

I’ve always been fascinated by learning new things. When I finished school and university, I naively thought the bulk of my learning was done — it was the applying I had to do now.

How wrong I was … and thankfully so! In my career, I’ve observed that one of the main factors separating high achievers at work is these people are quick at learning new skills and new technologies (I work in the technology sector).

So when I picked up “Ultralearning: Accelerate Your Career, Master Hard Skills and Outsmart the Competition” by Scott H. Young, I was intrigued by it. “Ultralearning” is a brilliant book born from Scott’s desire to learn how to pass the exams for the prestigious computer science program at MIT in just one year (it’s a four-year course). He managed it. As a result, he became increasingly interested in people who had achieved seemingly impossible feats of learning in short amounts of time. Completely fluent in Spanish in three months, for example. He pulled together nine key principles common to ultralearning projects, and his book is the result.

So is it possible that I’ve been wasting a lot of time learning things in a very inefficient way — all through school and university? Might this give me an edge in my career going forward?

So I set myself a challenge. I’ve always been quite mathematical and been interested in chess. But I’ve never played it seriously … and never very well. Chess has a ton of online resources, a large online community, and your progress (or lack of progress) is easily quantifiable. As the perfect first ultralearning project I set myself a goal: Could I get to an 1100 Elo rating ( here’s an explanation of Elo) in under 100 hours? Read on to find out.

Getting Started

After a bit more research, I defined the aim as follows. Could I achieve an Elo rating of 1100 in a rapid game (30 minutes per side, long enough to think a reasonable amount of time about each move)? I spread the task over four months. So 25 hours per month over a four-month period — easily doable in evenings and weekends around a day job.

So what did I know about chess at the start of the project? The notation of different squares? No. Common tactical plays? No. Classic openings? Didn’t have clue. But I did know the following:

The basic rules (how to set up the board, how all the pieces can move, the fact that white moves first)
The relative value of all of the pieces (very roughly)
Basic possible checkmates.
The importance of developing pieces at the beginning of the game.

The first principle of any ultralearning project is what Scott calls metalearning— learning how to learn. This sounds super obvious, but be honest — how often have you dived into learning something using a resource that’s just easy to access rather than optimal?

Scott advises that 10% of project time should be used to identify the different things to learn and to identify bottlenecks. That’s a lot of time. This sounds obvious, but trust me — it’s not simple to do. So, for example, if you want to learn to speak a foreign language, it wouldn’t make sense to start learning grammar before vocab!

So what are the most important areas in chess? In no particular order, this is the list I came up with:

Opening — there are a variety of possible chess openings that have been extensively studied over the years. Your opening combination hugely depends on whether you’re white or black. A good chess player has a large number of chess openings memorized.
Strategy — evaluating positions and setting goals and long-term plans for future play
Tactics — a short-term sequence of moves that limit your opponent's room for maneuver, which often results in material gain. These opportunities are usually found in what is known as the middlegame (i.e., after the openings and before the endgame).
Endgame — the end of the game, when typically few pieces are left on the board. The line between middlegame and endgame is slightly blurred — however, typically pawns and the king become much more powerful in this stage of the game.

So what are the most important skills to learn to get to an Elo rating of 1100? After a fair amount of online research, the consensus seemed to be:

For beginner chess players, the game is won or lost in the middlegame. The best way to get better at this is through tactics, tactics, and more tactics.
Beyond learning a couple of basic openings, it was pointless learning positions/theory on openings in-depth
It was much more important to learn about endgames than openings. Studying endgames was not just helpful for the endgame itself but could also improve your play in the middlegame as well.
When playing games, play longer-form games with very little time pressure, and think very careful about each game. Blitz games are much harder because you don’t have much time at all to think.

I decided to use the chess website, chess.com, as my main resource. This has a wealth of lessons, tactics and games on demand. This was an easy choice — it is by far the most comprehensive chess resource out there.

And this was my first and biggest mistake in this chess ultralearning project (I told you metalearning was important). I was so blinded with the constant reference to tactics that I neglected the importance of getting better at strategy. I was convinced that the better I got at tactics, the more obvious the strategy would become — and thus was not important in getting to an Elo rating of 1100. More on this later.

Tactics

So I started my ultralearning practice with tactics, tactics, and yet more tactics on chess.com. The key underlying skill in chess is pattern recognition — the more examples you’ve seen and solved, the better at pattern recognition you become. This corresponds to the drill ultralearning principle — pick the skill that’s the biggest bottleneck in your improvement, and practice it again and again.

In this case, I took this principle even further. The tactical puzzles on chess.com are helpfully split into different categories that represent different tactical motifs.

For example, there’s the pin (when a defending piece can’t move, as it would expose a more valuable piece behind it), a fork (when a single piece makes two direct attacks simultaneously, usually resulting in a material loss for the defender), and many others.

To speed up my ability to recognize patterns, I practiced tactics for each motif separately and then tested myself in a rated exam with all motifs present. I did this multiple times. Here I also used another ultralearning principle — retention — practicing a particular motif again and again, almost beyond what seems necessary in order to really cement the pattern recognition in my mind. I ended up achieving a tactics rating of 1510 on chess.com.

Endgame

In parallel to this, I also started working on the endgame. After exhaustive online research, the consensus seemed to be the best resource on endgames was “Silman’s Complete Endgame Course” by International Master Jeremy Silman.

What’s really great about this book is it details what techniques you need to know depending on your chess level. For beginners (which he defines at having an Elo Rating of 1100 and under), he runs through some basic checkmates, guidance about what combinations of pieces can lead to checkmate in the endgame, and some slightly more complex ideas about using the king/pawns effectively.

This highlights the importance of having effective resources on an ultralearning project — I could have spent a long time studying all the various intricacies of the endgame, but this wouldn’t have been relevant for the level I was trying to achieve. Therefore, my study of the endgame only took about six hours.

Strategy

Halfway through the project (50 hours down), I was feeling pretty confident. My work on tactics was improving a lot, and I felt I had the basic ideas of the endgame sorted.

However, a few casual games with friends with no formal chess training gave me a reality check. In some games, everything seemed to click and I won easily. In other games, I lost badly. Why the variation, I wondered.

Another of the ultralearning principles was required, experimentation. I also realized I was very bad at anticipating what my opponent was planning. Reflecting on this, one of the characteristics of tactical puzzles is you’re always presented with a situation in which you have an edge. So the skill is very much to find that magical combination of moves that drives home your advantage.

I gradually realized I had completely neglected a crucial part of the chess players tool kit — strategy. But where to start? There’s no end of strategy advice online, and much it of it, I didn’t feel was relevant to achieve an Elo rating of 1100. After much research, I discovered this blog post on chess.com which seemed to have a great overview of the key strategic ideas for beginners.

Therefore, in the third month, I worked my way through the various different ideas and videos contained in this post — including ideas about good positional play, pawn structure, making a plan, king safety, and other techniques. I formed a new appreciation for the difference between strategy and tactics. I found that tactics tend to present themselves when you’re in a good position strategically and less so when you’re in a poor position. Or in a succinct phrase: Tactics are the servant of strategy.

By far the most impactful exercise was playing against a computer without a queen (see here). The computer always plays the perfect move from any position. Believe me when I say winning this game is harder than it sounds! One of the benefits of this exercise is the instantaneous feedback you get (feedback is one of the other ultralearning principles). As soon as you make a bad move, you can see your score dive, which is always a precursor to the computer gaining material or checkmate. It took me a surprisingly long time before I could beat the computer consistently.

Playing Games

So at the start of the fourth month (75 hours down, 25 hours to go), I still had no idea how likely I was to beat my goal.

As a final sprint to the finish line, I spent the remaining time playing rapid chess games (30 minutes per side) to try and push my score above 1100. Chess.com has a fantastic analysis functionality — I tried to spend at least 15 minutes after the end of each game looking through what I did wrong, noting down the key mistakes and thinking about how to avoid them in the future.

And the Punch Line Is …

So did I get to an Elo rating of 1100? Not quite, but I was very close. 1093 was my final score. Some final reflections:

I really enjoyed my first ultraleaning project. There’s something immensely satisfying about learning something completely different from scratch. I’ve also developed a newfound appreciation for the game of chess. It’s easy to start learning but very hard to master — I can see how it can become endlessly fascinating.
I should’ve spent longer on the metalearning part of the project at the start. If I had realized the importance of strategy earlier on, I’d certainly have changed the order I learned things and perhaps learned more efficiently.
I was constantly reminded of the importance of focus (another ultralearning principle). Looking back, I achieved so little when I was distracted or tired that there was almost no point in continuing. If I had only done ultralearning when I was super focused, I might well have achieved a lot more.

I’d love to your feedback on what I have done — I may well have missed something obvious. Happy ultralearning everyone!

👉The Better Humans publication is a part of a network of personal development tools. For daily inspiration and insight, subscribe to our newsletter, and for your most important goals, find a personal coach.👈