I Gained 125 ELO in 30 Days with This One Cool Trick*

*Disciplined metacognitive strategy

Hello! For the next month, I’ll be out in beautiful Albuquerque, New Mexico for my wife’s work, which means I won’t be playing my regular games at the club. I’m going to take this opportunity to write about a few other topics unrelated to OTB chess, though it’s always possible that some past games will trickle back in.

Life in Albuquerque. It’s never too early to learn the advance variation of the Caro-Kann.

This week, I want to check in on the chess pact I recently made, influenced by Anne-Laure Le Cunff’s book Tiny Experiments. In a post called “Is Learning the Hard Way the Only Way to Learn,” I wrote:

There’s only one way to both cut down on the number of games I play and increase the number of meaningful games I play: I need to analyze every single game. And I don’t just mean, I’m going to run the Game Review and check the engine and nod and move on: I need to narrativize them, deriving some sort of meaning. I need to tell a story about each game, even if it’s just a sentence-long story.

I’ve been reading Anne-Laure Le Cunff’s fascinating book Tiny Experiments. One of her primary suggestions is to make “pacts” with yourself: contracts in which you pledge to do something for a set number of days. These pacts can’t be outcome-dependent, i.e. “I’m going to get to 1800 over the next 30 days” — that isn’t up to me. They need to be fully within my control.

So here’s a pact that I’m going to make, that you, dear read, can hold me to: for the next three months — so, until June 24, at least — I am going to analyze every single game I play according to Nate Solon’s OBIT method. I’ll compare the opening to my files and adjust them accordingly; I’ll look at my blunders according to the engine (and my mistakes, which I think are worth considering, too); I’ll isolate one element from the game that is “interesting,” meaning that I wasn’t sure what to do while playing (having to identify this in and of itself makes me think more consciously about how I’m playing); and, maybe most importantly, I’ll come up with a takeaway.

Now, it hasn’t been 90 days yet; it’s been just under 30. But the results have been so remarkable to me that I thought it was worth writing about. At the time I wrote that piece, I’d just reached a Chess.com online rapid rating of 1714, which was my highest to date. But twice before, I had broken 1700 only to fall back into the 1600s. Even worse, on July 14, 2024, I’d achieved a rating of 1682, only to find myself, just over two months later, at 1434.¹ So I was aware of the possibility — nay, the likelihood — that my time in the 1700s would be brief, and that I could fall precipitously once again. Hence the determination to see if a dedicated practice of self-analysis might be able to sustain my gains.

It did more than sustain them. Ever since I started analyzing all of my games, I’ve been on a tear. I’ve climbed as high as 1845, more than 125 points higher than my previous peak and representing a whole new stage in my chess performance so far.²

If you look at my lifetime Chess.com graph, there are a few different phases that emerge. From 2017-2019, I learned how to play chess beyond the basic grasp of the rules that I’d had previously. This got me into the low 1300s before I stopped playing for about a year and a half. In late 2020, in the heart of the pandemic, I picked the game up again, and I immediately jumped to around 1500 before settling more consistently in the 1400s, with a dip all the way into the mid-1300s in July 2023. Then, at the end of 2023, I jumped into the 1500s, and that’s the phase I’d been in ever since: a year spent pinging back and forth between 1550 and 1650, with some brief interludes above and below those points

My first leap, starting in 2017, coincided with my absorption of John Bartholomew’s Chess Fundamentals and Climbing the Rating Ladder series. I would credit these videos alone with making me a 1300-strength online rapid player. In 2020, I got a bit better mostly just by playing more; I got a Chess.com premium membership in December of 2020. However, this approach wasn’t enough to represent significant and permanent improvement. That came toward the end of 2023, when I started more seriously engaging with chess: I joined the ChessDojo, read Michael Stean’s Simple Chess, upped the number of puzzles I was doing, and started watching ChessBrah’s “Building Habits” series and Daniel Naroditsky’s first Speedrun. Soon after that, in April 2024, I began playing OTB.

All of this bought me 100-200 rating points, depending on the day — but eventually, I plateaued, even as my OTB rating (which has been between 1600-1650) seemed to suggest that I should be slightly better online.³ I couldn’t figure out what the problem was. Playing online, I felt like I just couldn’t see the board as well as I could OTB. Nothing meant as much; I seemed to be deriving less information from the positions I looked at.

I’d experimented with analyzing my online games before, but I’d never gotten the habit to stick. So when I made the pact to analyze every online rapid game, it was a big shift. I didn’t know whether it would have any material benefit, but I figured it wouldn’t hurt. I certainly did not anticipate it making such a significant difference.

To re-state the reason why I started doing this: I had realized that I derived immense amounts of information-rich meaning from my OTB games, but that my online games basically disappeared from my mind as soon as I’d finished them. This struck me as an immense waste of a lot of time I was spending playing chess online; and because it has become eminently clear to me over the years that I am the kind of person who, given 20 minutes, would rather fit in an online game than pick up a chess book, I knew that if I didn’t find a way to turn these games into teachable experiences, I would stop improving.

Since I started the pact, I’ve analyzed 63 games — about two a day. For each one, I’ve come up with a takeaway that I’ve compiled in a spreadsheet. I believe that this experiment has been an unmitigated success, and not just because of the rating jump (though that provides the clearest evidence). Even more than that, I’ve had the noticeable experience of my chess vision improving over the course of the month.

By analyzing my games⁴, I force myself to determine why I screwed up, and that makes each one into a lesson in something. For example, this morning, I played three games in my mad dash for 1850, and all three of them came down to me not calculating accurately enough. Obviously, I need to put in some deep calculation time, and spend more time calculating in tricky positions.

Another example comes from a position I reached a few days ago:

There had been a pawn on b6, and I snapped off the bishop with my knight as soon as White played this move. If I hadn’t analyzed this game, I never would’ve thought about it again. If I merely ran the game review, I would’ve seen that it was a blunder, asked the engine to show the right move, saw that I should’ve pinned the bishop to the rook, and instantly forgotten.

But by analyzing it, I force myself to put the situation into words — and it very clearly struck me that the issue at hand here is keeping my eyes open for tactics in the endgame, when pieces are more likely to be undefended or loosely defended. It’s natural in an even-material endgame to get caught in the mindset of, “This is equal, it’s a draw,” but I could’ve easily won a piece here if I’d been attuned to the possibility. This is not a concept I had grasped previously, and by formulating it out of my own play, it becomes, if not a ready-at-hand idea, at least one that exists subconsciously, connected to an actual chess situation.

In Tiny Experiments, Le Cunff talks about metacognition, which she defines as “‘thinking about thinking’ or ‘knowing about knowing.’ It’s being aware of your own awareness so you can determine the best strategies for learning and problem-solving, as well as when to apply them.” Metacognition seems to be the most concise explanation for why this new approach is working so well for me.

Metacognition can help you maximize your potential to think, learn, and create, all while taking care of your mental health. Beyond the elevated self-awareness and consciousness you’ll experience by applying metacognitive strategies, scientists have investigated some of the many benefits of metacognition.

Learn better. Research shows that high-metacognition learners identify challenges much faster and change their tools and strategies to better achieve their learning goals. Metacognition can even compensate for IQ and lack of prior knowledge when it comes to solving new problems.

Make decisions faster. Monitoring and controlling your ongoing cognitive activity can make you aware of your cognitive biases and help avoid mistakes, or at least not reproduce the same mistakes twice. In addition, because of heightened awareness, metacognition leads to a reduction in response time, which reduces the time to solve a problem or complete a task.

Be more creative. According to Dr Markus Lång, all narrative works of art can be defined as metacognitive artifacts which are designed by the creator to anticipate and regulate the cognitive processes of the recipient. Intrinsically speaking, creativity is thinking about thinking.

Improve your mental health. Metacognition gives you the ability to understand your mental health and to adapt your strategies to cope with the source of any distress. As such, researchers defined metacognition as the process that “reinforces one’s subjective sense of being a self and allows for becoming aware that some of one’s thoughts and feelings are symptoms of an illness.”

Even the “improve your mental health” bit rings true. By instituting the “review every game” rule, I’ve hugely cut down on tilted sessions, in which I play game after game, getting more and more frustrated with each one.

So if you’re looking to improve at something — whether it’s chess or anything else — then I highly recommend implementing some sort of metacognitive practice. If you’ve tried anything similar in the past, let me know. And if you’re inspired to do so by this piece, definitely let me know!

1. My son was born on July 30. Turns out that playing on your phone at 1 am while your baby sleeps on your chest at is not a great way to boost your rating.

2. As of pub time, I’m at 1829. I decided to make a push for 1850 this morning and took two tough Ls.

3. The ChessDojo’s rating conversion chart equates a 1600-1675 USCF rating with a 1750-1850 Chess.com rapid rating.

4. To be clear: I’m spending five or ten minutes max analyzing these online games. I usually spend an hour or two analyzing my OTB games.