Chess

Two 15-minute slots a day · ~2.5 hrs/week · built around a Sun–Thu work week

Today

The weekly loop

DayMidday — 15 minNight — 15 min
SunPuzzlesGame
MonPuzzlesReview Sunday's game
TuePuzzlesGame
WedPuzzlesReview Tuesday's game
ThuPuzzlesGame
FriOptional: review Thursday + one video
SatOff

Midday — puzzles, always

Ten to fifteen puzzles on lichess.org/training, solved slowly. Puzzles suit a work break because they're interruptible — close the tab mid-puzzle and you lose nothing. They're also the highest value-per-minute activity that exists at your level.

Game nights — Sun / Tue / Thu

One game, 10|0 or 15|10. One. If you lose, no revenge game — that's how fifteen minutes becomes ninety and how you drill bad habits at speed. Close the laptop.

Review nights — Mon / Wed

Fifteen minutes on the previous night's game. Run Lichess's free analysis, find only the moves flagged as blunders or mistakes, and for each one cover the engine line, look yourself first, then answer one question: what was I not looking at? One sentence in a notes file. Three blunders, three sentences, done.

The notes file is the actual secret. After a month you'll find you make the same three mistakes over and over. That list is what gets you to 1200 — not more games.

Why not play every night

A 15|10 game takes 20–35 minutes of real time, so it doesn't fit a 15-minute slot — and a game you rush at the end trains exactly the habit you're trying to break. More importantly, review is the bottleneck, not volume. Most beginners play five games a week and review zero.

One warning about chess before bed

Losing at 23:30 is genuinely activating; people lie awake replaying it. If that happens, flip the slots — game at midday if you can steal 25 minutes, puzzles at night. Puzzles are calm. Competitive games are not.

The move routine

At your level games are not decided by clever plans. They're decided by who leaves a piece hanging last. Run these three checks every single move — it's slow, boring, and worth several hundred rating points on its own.

  1. What did my opponent's last move do? Every move attacks or enables something. Find it before anything else.
  2. List my checks, captures and threats. Forcing moves first, always.
  3. After my move, is anything of mine undefended? Scan your own pieces before you release.

You physically cannot do this in a 3-minute game. That's why the plan says 10|0 or 15|10 and never bullet.

What puzzles actually are

Every puzzle is a real position from a real game where someone had a winning tactic available. You're told only "Find the best move for White" (or Black) — your side is the one at the bottom of the board.

The two things that surprise everyone

Puzzles are 2–4 moves deep, not one move. Play your move, the opponent's reply animates automatically, and you must find the next one too. Solving means seeing the whole line before you start.

It demands the single best move. A move that's good but not best is marked wrong. That's not unfair — in these positions something decisive exists, and "pretty good" means you missed it.

A puzzle, concretely

You're White. Black's king sits on g8, Black has a rook on c8, and you have a knight on d5. You play Ne7+ — the knight lands on e7 giving check, so Black must respond, and from e7 that same knight also attacks the rook on c8. Black moves the king, you take the rook. You're up a rook.

That's a fork, and it's roughly 20% of all puzzles you'll ever see. Once "knight on e7, enemy king on g8" is burned in, you spot it instantly forever. That burning-in is the entire point of doing this daily.

The solving loop

  1. Position loads. Note whose side you're on.
  2. Don't touch anything for 30 seconds. Look for undefended enemy pieces, the enemy king's escape squares, and any of your pieces that could give check.
  3. Find a candidate — then, before playing it, ask what does my opponent do in reply? Can't answer? Keep looking.
  4. Play it. Green = right, board replies, continue. Red = wrong, puzzle over.
  5. When wrong, do not click Next. View the solution, play it through, sit until you can say out loud why it works. The failures teach more than the solves.

About the rating number

Lichess starts you at 1500, which is wrong for a beginner, so it will fall fast — likely to 900–1200 in your first week or two. That's calibration, not failure. Puzzle rating and game rating are different scales entirely; a 1000-rated player commonly has a 1600 puzzle rating. Never compare them.

Settings worth touching

The seven patterns that cover most puzzles

ForkOne piece attacks two things at once.
PinA piece can't move — something more valuable sits behind it.
SkewerThe reverse: the valuable piece is in front, must move, exposing what's behind.
Discovered attackMove one piece; the piece behind it delivers the threat.
Removing the defenderCapture or chase off the piece holding everything together.
Back-rank mateThe king is trapped behind its own unmoved pawns.
Hanging pieceSomething is simply undefended. Take it.

Pace

10–15 puzzles in 15 minutes — about a minute each. If you're averaging 20 seconds you're clicking, not calculating. Fewer and slower beats more and faster, every time.

Before you start the loop

Puzzles train pattern recognition; they don't teach the rules. They quietly assume the following are automatic for you. This is about 90 minutes of work — one evening, not a semester.

1. Piece values — non-negotiable

Every puzzle's reward is "you win material." If you don't instinctively know a rook beats a bishop, the solution is meaningless.

Pawn
1
Knight
3
Bishop
3
Rook
5
Queen
9
King

Bishops are marginally better than knights in open positions. Trading a rook for a bishop or knight is called losing the exchange and is usually bad.

2. The rest of the mechanics

Lichess → LearnBasics covers all of it interactively in about 30 minutes.

3. Two checkmates and one endgame

You will win material and then need to convert it. Lichess → Practice:

4. Create the notes file

One line per mistake, forever. Then start on a Sunday.

There is no secret body of knowledge. Chess has about a page of rules. When a 1600 player instantly sees the right move, they aren't applying a theory you haven't read — they've seen that shape a few thousand times and it's simply familiar. Your gap isn't knowledge, it's pattern exposure, and you can't read your way to it. That's why puzzles come first.

What puzzles honestly don't teach

Openings — principles, not lines

Do not memorise openings yet. Four rules cover the first ten moves of any game:

  1. Control the centre with a pawn (e4 or d4).
  2. Develop knights and bishops toward the centre — every piece moves once before any piece moves twice.
  3. Castle early, by move ten.
  4. No early queen sorties, and no pawn-grabbing while undeveloped.

If you want one fixed repertoire so you're not thinking from scratch: 1.e4 as White; answer 1.e4 with 1…e5 and 1.d4 with 1…d5 as Black. Symmetric, principled, teaches real chess.

What this schedule gets you

Ratings are Chess.com rapid. Lichess runs roughly 300–400 points higher at these levels — never compare the two directly.

MilestoneRatingOn this schedule
Stop hanging pieces most games~600Week 4–8
Beat any friend who knows the rules~1000Month 4–6
Genuinely decent — plans exist1200–1400Month 12–18
Club strength~16003–5 years

Progress is not linear. The first three months are a rocket; then it flattens hard around 1000–1200, where "don't hang pieces" stops being enough. Most people quit there and conclude they lack talent. It's just the shape of the curve.

Protect the streak, not the session length. Five consistent 30-minute days beat one three-hour Sunday by a wide margin.

Resources

Reminders

Push notifications at your two slots, day-aware — midday says puzzles, nights say whether it's a game or a review.

Checking…

On iPhone, notifications only work if you add this page to your Home Screen first (Share → Add to Home Screen), then enable them from the installed app.