12 minutes
Thoughts on chess improvement, after gaining 600 points in 6 months (1200-1800)
I played chess briefly with my friends in high school, where I had a rating of around 1100 online. I gave up the game for about 10 years, then picked it back up again recently, in May. I’ve tried a ton of different approaches in the past 6 months, switching things up whenever I stopped seeing results. I’ve improved from 1238 to 1802, and based on my win rate currently I think I’ve got another 50 points in me before I plateau again. From what I’ve seen online, this is solid improvement for an “adult improver”, so I figured I’d write up some thoughts/observations.
Here’s my lichess profile, hopefully I’m not below 1800 by the time you read this :D
Methods
Tactics, the obvious way
The first and often only response you’ll see online, when someone asks how to improve at chess, is to do tactics. I tried this many times when I’d try to get back into chess. I’d open up chess.com, get a premium membership, then grind puzzles every day, thinking I was doing exactly the right thing.
I didn’t see any improvement, so then I thought “I’m probably just not being diligent enough, I get some of them wrong. I’m not going to move until I’m absolutely certain I have the right move”. I’d spend over 30 minutes on one puzzle sometimes. My rating didn’t budge. This is unsurprising looking back. What I needed wasn’t calculation practice, but pattern recognition. So I wasted countless hours, all the while thinking I was going to be great at chess when I started playing again, because I had such a killer puzzle rating.
I firmly believe that the mantra of tactics, tactics, tactics is doing harm to people trying to get better at chess, because the most obvious application of this advice is to open up the puzzles tab of lichess or chess.com and get after it. I got dismayed by my progress many times with this approach.
Woodpecker method
I got back into chess this most recent time, by picking up this book. The idea is you do a set of puzzles, let’s say 300. Then you do the same set in half the time, then half again, etc. 7 sets is the standard, the last set you’re supposed to be able to complete all the tactics in a day. This sounded like a novel approach, so I did a few sets of the ~300 beginner problems. By the third set, I felt like I was just remembering the exact positions. Plus I didn’t love some of the puzzles, so I’d get stuck on one then get frustrated when the answer was one of those “and then white has a crushing position”, which is totally useless for a beginner who can’t identify when a position is “crushing”.
I also didn’t like the analog format here. I’d rather a site I can go to on my phone, and open up an analysis board when I don’t understand the solution.
Can’t say a whole lot about the effectiveness of this method, given that I didn’t see it through, but I’m glad I encountered it because it got me back into chess.
Puzzle storm
Puzzle storm is a game on Lichess, where you try to do as many puzzles as you can in about 3 minutes. You get bonus time for getting puzzles right, and they get harder as you go.
My absolute biggest improvement came from playing this obsessively for a couple weeks. My high score my first time was about 20, my high score now is 33. It made me so much quicker at assessing what is happening on the board. You start to spot hanging pieces and simple tactics immediately. At the rating I was at (~1400), that was pretty much all I needed to start winning consistently. I didn’t have to look around for hanging pieces and obvious pins, they just popped out at me. I’d attribute about 150 points of improvement to just those two weeks of puzzle storming.
Puzzle streak
Puzzle streak is another game on Lichess. You keep getting harder puzzles until you fail. This seems to result in a good ratio of easy to hard puzzles. I try to do a streak or two per day. My high score is 45, which I’m pretty proud of.
Reading annotated games
I bought a couple books, Logical Chess Move by Move, and The Art of Logical Thinking. Both are collections of master games, annotated with explanations. This is another method that is often recommended to beginners or those seeking to improve. Personally, I didn’t get anything out of these. I think this is mostly because I’m lazy though, if I had pushed myself to really think about each move, play along, try to figure out why alternatives wouldn’t have worked, etc., then maybe I’d have benefited from it.
Guess the move
Guess the move is a game on chesstempo, where you play through a master game and try to guess what to play. It will tell you if your move was the move that the grandmaster played, or if it’s a valid alternative. When I really take my time to think about the position as I play through, I find this useful. It’s definitely more of a supplementary thing though, I didn’t see a big improvement from this alone.
Memorizing master games
Exactly what it sounds like; find a game with an opening you play, and memorize it. I combine this with guess the move, in that I’ll play through the game once, trying to guess the continuation, then I’ll memorize the game. I’ve only memorized ~5 games so far, since I’ve just started this method.
I actually think this is hugely beneficial, surprisingly. I would have considered it a waste of time before giving it a chance. The trick is that to memorize a game, you sort of have to understand it. It’s possible to just memorize moves like you’d memorize a list of random words, but it will be 10x harder than just understanding what’s going on.
If you understand what’s going on, you end up memorizing the game in a series of chunks, instead of a series of moves. For example one miniature that I’ve memorized is this game between Peter De Bortoli and Botond Smaraglay. I can recite move by move, but the way I remember it is roughly “Smith Morra gambit, knight development, bishop development, scare off bishop, threaten queen trap, knight blunder, queen trap”. Memorizing a couple king’s gambit games has definitely improved my king’s gambit play by giving me more ideas.
Visualization practice
It’s very easy to envision a variation wrong. You could forget that a piece will be gone afterwards, or that a pawn move will open up a diagonal, or that the enemy queen will be in position to fork you up. There are a series of visualization courses on Chessable where you are presented with a position and a series of moves, and you have to figure out the right tactic after those moves, without playing those moves. I don’t know how effective it is, as I mixed it into my training at the same time as a bunch of other stuff, but it seems to help, and it’s fun. I’ve found it easier to envision the board state after a few moves, since starting these courses.
I have some reservations about the actual course material. Visualization 1 has a weird amount of move repetition, so after 4 moves there might only be 2 pieces moved, which reduces the difficulty a lot. I bought Visualization 5, which takes real games instead, but then I felt the solutions were too hard to find, even with perfect visualization. Ideally I’d want the solutions to be simple and the visualization to be hard. So I’m open to other ways of practicing this skill.
Update: I’ve been pointed towards listudy, which has this sort of visualization exercises. You can also set your own ply depth, which is sick. Check it out.
Updated update: I’ve made my own tool which addresses the problems I’ve seen in other tools. Unlike listudy, which only has checkmate puzzles currently, there are a variety of puzzles. The puzzles are taken only from high-level (2000+ elo) games, so the moves before the tactic should make sense. If you’re going to be spending hours visualizing moves, it’s probably best not to be visualizing the chaos of a 1200-elo game. All the puzzles are easy (less than 1500 on lichess), so the focus is on the visualizing, and not intense calculation. You can set your own depth for each puzzle. You can also check out the puzzle on lichess, if you need to analyze a variation or something. In my biased opinion, I think it’s the best tool for this kind of practice.
Playing games
Personally I don’t feel like I improve much by playing. It’s the only way to tell whether my training is working, but I don’t consider it training in and of itself. Running the computer analysis and finding out that I missed a fork doesn’t help me find them in the future. Maybe once I reach a higher level where there’s more strategy. For now, where games are still decided by pretty simple tactics and blunders, I don’t see much point.
YouTube
Felt like I had to mention it. This isn’t studying. But Gotham Chess is highly entertaining and I will happily waste 2 hours watching guess the elo.
Coaching
I’ve tried a couple coaches, and I can’t say they’ve been a huge part of my improvement. It did help keep me accountable in the beginning when my motivation would waver, but I just think there’s not a whole lot for a coach to tell someone my level. What are they supposed to tell you about your games if you’re below 1900? “Hey you probably shouldn’t hang your knight on move 12, also don’t fall for basic forks”? Everything I practiced with a coach I could have done on my own.
Openings
Second only to “tactics, tactics, tactics” in forums on chess improvement is “don’t study openings before XXXX elo”. There’s definitely some truth here, If I forgot all my openings right now, I wouldn’t drop that many points. That being said, it’s frustrating to drop pieces or pawns in the first 5 moves, or fall for common traps, and it’s so easy to remedy. I got a couple courses, My First Chess Opening Repertoire for White, and the same for black.
It took about 3 weeks of low-effort studying to complete the course at a depth of 5 moves. It got me into reasonable positions out of the opening, avoided common traps, and gave me a minute or two more each game to think about the middle/endgame. After learning the first 5 moves, I stopped studying openings, as you definitely hit diminishing returns, but I think everyone should have a repertoire at least a few moves deep.
I don’t love the openings recommended by the course, so I’m not saying to go and study those repertoires specifically, but it’s a convenient package, rather than having to figure out what mix of courses to buy to get a complete repertoire.
Endgames
This is another one where I’m going against the common advice. How often do <2000 players get into a P+K v K endgame? 1 in 50? I studied an endgame course for a couple weeks before realizing it wasn’t making any difference in my play. I’ll return to it later, when this becomes more important, for now it’s just not worth the time. If you’re around my level you’re better off doing easy tactics so you recognize the hanging knight on move 10, or avoid the fork on move 15. Then you won’t have to worry about king opposition on move 50.
Blunder preventer
This is a game on Aim Chess where you’re given two moves, and you have to figure out which one is the blunder. I started doing this recently when I got frustrated that I had just lost 3 games in a row to easy blunders. This was actually hugely beneficial, it’s like it flipped a switch in my brain and now I actually think tactically about the other side’s threats. Some of the solutions are iffy, as it’s not clear exactly what the blunder is sometimes, but besides that I think it’s an excellent method.
Board Visualization
Bit of an oddball one, I tried to get better at visualizing the board in my head for a long time. I’d just close my eyes and try to see the squares, what color they were, what other squares a piece on that square would attack, etc. Unfortunately I’m still terrible at this. Also pretty sure this gave me zero improvement.
Rough sketch of my progress
Calibrated at around 1230 elo. Did woodpecker method and played games, got to around 1300. Plateaued at 1300 for a while, hanging pieces like it was my job. Did more woodpecker method, some online tactics, climbed up to 1400. Discovered puzzle storm, jumped up to 1550 and had my longest winning streak (8 games). Learned some openings, up to 1600. Floundered for a while, struggled up to 1700 while doing a bunch of puzzle storm. Wavered around 1700 for a while, updated my training method to the list of things below, then shot up to 1800 in a week.
My recent matches have felt like a whole different game. Something clicked and now I’m very rarely hanging pieces, or falling for simple tactics. The winning strategy right now is just play moves until my opponent messes up, which happens almost every time. I think this will work until around 1900, at which point I might have to actually learn some strategy and make some plans.
Current training
Currently I’m trying to do the following every day:
- Review memorized games
- Memorize a new game if I’m ready for it
- 5 visualization problems
- One run of puzzle streak
- Play some games
- 5 problems in blunder preventer
- Chessable review
- King’s gambit course
- Smith morra
- My first opening repertoire, white & black
I never stick to anything for more than a few weeks, so I’m sure this will change soon.
If I were to start over, knowing what I know now about what works, I bet I could improve way faster. I would rip through an opening course in a couple weeks to get a repertoire, then do a ton of puzzle storm and puzzle streak. Once I was seeing basic tactics quickly, I’d mix in some blunder preventer exercises. Supplementary stuff like memorizing games and visualization I’d add in later.
If you play games, my PSN is mbuffett, always looking for fun people to play with.
If you're into chess, I've made a repertoire builder. It uses statistics from hundreds of millions of games at your level to find the gaps in your repertoire, and uses spaced repetition to quiz you on them.
Samar Haroon, my girlfriend, has started a podcast where she talks about the South Asian community, from the perspective of a psychotherapist. Go check it out!.
If you want to support me, you can buy me a coffee.
2451 Words
2021-10-06 19:32 +0000