Coachss
Learn
Pricing
Analyze my gamesAnalyze
Coachss|coachess.app
The MethodContactTermsPrivacyCookies

© SORTINO LABS S.R.L.

Home

The method

How Coachess teaches you chess.

Every app says it helps you learn. Most do not say how. Here is how Coachess does it, in plain English, with the names of the researchers who figured it out first so you can check we are not making it up.

8

evidence-based principles

Depth 30

Stockfish verification

Real games

no invented positions

Contents

01Lessons02Puzzles03Drills04Practice05Game review06Difficulty07Session arc08What we refuse to do09Our promise
01

Lessons

How a lesson is built.

A lesson is not a video you watch. It is a position you navigate.

A Coachess lesson is not a video. It is a position you navigate, step by step, with Boris next to you explaining why each move matters. You play the right move, you see the opponent's best reply, you understand the idea and the one after it.

Every line Boris teaches has been checked against Stockfish at depth 30. When a lesson shows a trap, you walk into it yourself so you feel the refutation, because reading 'do not play this' teaches nothing but playing it and watching it fall apart teaches forever. Arrows are color coded on purpose: orange is what we are doing, red is what the opponent is threatening against us. The colors are part of the language.

♗♞Orange = our planRed = their threat

The research

Chess pedagogy built on primary sources. No invented positions. If we teach a line, it came from a real game or from theory backed by master practice.

02

Puzzles

How we pick your puzzles.

The difficulty is a real number set by real players, not a guess we made.

Every puzzle you see on Coachess has been solved and failed by hundreds of thousands of players before you, from open community databases. Their wins and losses are what set each puzzle's difficulty. It is a real number, not a guess we made.

Your rating moves with the same math that grades world champions, so the number in your profile is not a vanity score invented for this app. We match puzzles to your rating so you are always solving something you can actually solve, with just enough teeth to learn something. There is no timer, because a rushed solve is not a good solve.

80012001600200024001450you

The research

Rating systems for chess started with Arpad Elo, whose design was first deployed by the US Chess Federation in 1960 and adopted by FIDE in 1970. Mark Glickman refined it into Glicko in 1995 and Glicko-2 in 2012, which is the math Coachess uses to grade your puzzles.

03

Drills

We drill, because you forget.

The gap is the point.

Your brain is built to forget. The lines you studied last week are mostly gone by Friday unless you meet them again.

Coachess shows you the same position at exactly the moment your memory is starting to slip, so the meeting reinforces the line instead of re-teaching it from scratch. A card you see today might come back tomorrow, then in three days, then in a week, then a month. The gap is the point.

memorytime →day 1day 4day 14day 60each review lifts the curve before it drops too far

The research

Hermann Ebbinghaus mapped how fast we forget in 1885. Piotr Wozniak turned that math into the SM2 algorithm in 1990. That algorithm is what schedules your drills.

04

Practice

We ask. We don't show.

Finding it once beats watching it ten times.

Most chess apps play the brilliant move for you and ask you to admire it. Watching the answer teaches you almost nothing.

Coachess flips the board and makes you find the move yourself. Finding it once, even if you struggled, beats watching it ten times. That struggle is what moves the line from a thing you read into a thing you know.

watchinglearn almost nothingplayingvs

The research

Henry Roediger and Jeffrey Karpicke ran the experiment in 2006. Groups that tested themselves remembered far more a week later than groups that re-read the same material multiple times.

05

Game review

Every game becomes a lesson.

A mistake you analyze yourself sticks. A mistake you are told about evaporates.

When you play a real game, the lessons are sitting there in the moves you actually made. Not in a curated puzzle from a database. In the moment you played Bf4 instead of Nf3, in the move you spent six minutes on and still got wrong. That is your homework.

Coachess loads your games from Chess.com or Lichess, runs them through Stockfish, and finds the inflection points where the position turned against you. Then Boris walks you through them. Not a list of engine numbers. A coach naming what you missed, asking you to find the move that was actually there, accepting any move close enough to the engine's pick.

♞??what you playedwhat was there

The research

Mikhail Botvinnik, world champion from 1948 and the founder of the Soviet school of chess, made every student analyze their tournament games in writing after each match, finding the inflection points themselves before consulting any reference. Modern cognitive science calls this error-driven learning. Janet Metcalfe's 2017 review in the Annual Review of Psychology gathered the evidence: errors you analyze are remembered far longer than errors you are told about.

06

Difficulty

We make it a little too hard.

The answer has to be earned for the answer to stick.

If every card is easy, your brain coasts and learns nothing. If every card is impossible, you give up and learn nothing.

We calibrate the sessions so you are always pushing a little past comfortable. The middle cards of every session are the hardest on purpose. The answer has to be earned for the answer to stick.

skillchallenge →too easy → boredthe sweet spottoo hard → quit

The research

Robert and Elizabeth Bjork named this principle "desirable difficulties" in 1994. Making the practice harder than the thing itself is exactly what makes the thing itself feel easier when you get there.

07

Session arc

A session has a shape.

Ten minutes of chess, not ten tasks checked off.

Every drill session is three acts. The first few cards warm you up with patterns you already know. The middle is real work, at the edge of your ability. The last card is a win you can carry with you.

This is why a ten-minute drill on Coachess feels like ten minutes of chess, not like checking off ten tasks. You are in the middle of something, and then you finish. The arc is what makes the rhythm.

ACT 1ACT 2ACT 3INTENSITYTIME →hardest cardwarm upreal workfinish strong

The research

Mihaly Csikszentmihalyi described this state as "flow" in 1990. Sustained attention only happens when the challenge and your skill meet in a narrow band, held for a stretch of time.

08

What we refuse to do

No coins. No tiers. No fake streaks.

The reward for playing chess on Coachess is playing better chess.

Other apps pay you in points and badges to keep you coming back. It works, until it doesn't. The day the points stop feeling good is the day you stop caring about chess too, because chess became the vehicle for the points.

Coachess has a streak counter because flow deserves to be recognized, and the Board Vision drill has a weekly leaderboard so you can see how your square recall stacks up. Those are real numbers about a real skill, scoreboards rather than tokens. What we refuse: coins, gems, experience bars, Bronze-Silver-Gold tiers, daily league promotions, fake achievements. The reward for playing chess on Coachess is playing better chess. If that is not enough for you, the fault is ours and we should fix it, but not by hiding it behind tokens.

1000coinsLVL 42XP bargold leaguejust the chess

The research

Edward Deci and Richard Ryan have been showing since the 1980s that external rewards crowd out the internal ones. Give a child gold stars for reading and they read less, not more, when the stars stop.

09

Our promise

Every lesson is verified. Every game is real.

If Boris cannot cite the source, the lesson does not exist.

Every line Boris teaches has been checked against Stockfish at depth 30 before it ever reaches you. If the engine disagrees, the lesson is rewritten or it does not ship.

Every instructive-game lesson replays a real historical game. Capablanca versus Marshall, New York 1918. Kasparov versus Topalov, Wijk aan Zee 1999. Ding versus Nepomniachtchi, 2023. The white player, the black player, the year, the event, the result. If Boris cannot cite the source, the lesson does not exist.

GAME STUDYCapablanca vs MarshallNew York 1918 · 1-0Verified against Stockfish depth 30

The research

Not a research citation. A promise. Chess is old and well documented. There is no reason for a coach, human or otherwise, to invent a position when thousands of real games already teach the same idea better.

♔♘

If you want to dig into any of this, the names are real. Google any of them, read the original papers, argue with them if you want. We built Coachess on top of their work, and we think you deserve to know whose shoulders we are standing on.

Written and maintained by the Coachess team. If we get something wrong or if the research has moved on, write to us and we will update this page.