Chess Puzzles for Beginners: How to Practice Tactics the Right Way
Chess puzzles are positions where one side has a winning combination. Thirty focused minutes daily moves most players from 900 to 1200 faster than any other activity. Here's which puzzles to solve, in what order, and how to actually learn from them.
Chess puzzles are positions where one side has a winning combination. A sequence of moves that wins material or delivers checkmate. Solving them builds the pattern recognition that lets you spot tactical opportunities in your own games. Thirty focused minutes of puzzle practice daily is the single fastest way to improve below 1600 ELO.
Most players don’t practice puzzles correctly. They skim through them too fast, skip the ones they miss, or jump straight to mixed puzzles before any pattern is automatic. The method matters as much as the volume.
Where to find puzzles
Lichess puzzles (free, unlimited): lichess.org/training. Open-source database, over 6 million puzzles as of mid-2026 and growing continuously, rated by difficulty, filterable by theme (fork, pin, skewer, etc.). Every puzzle is generated automatically from real games played on the site, positions where one side made a mistake that let the other find a forced tactical win, then verified by an engine, so you’re training on genuine mistakes real players made, not composed problems. The best free puzzle resource. You can practice by theme or mixed.
Chess.com puzzle trainer (free and paid): chess.com/puzzles. Similar to Lichess with a different rating system. Puzzle Rush (timed mode) is available on the free tier and good for speed training once patterns are established.
Puzzle books: Traditional printed collections like Laszlo Polgar’s Chess: 5334 Problems, Combinations and Games have thousands of positions organized by difficulty. Less convenient than digital but useful if you prefer working without a screen.
The right order: themed sets first
Don’t start with mixed puzzles. Start with themed sets, and know what each theme actually is before you drill it:
- Fork. One piece attacks two or more enemy pieces simultaneously. Knights are the classic forking piece because their L-shaped move reaches squares no other piece threatens from the same distance, but pawns, bishops, rooks, and queens can all fork too.
- Pin. A piece can’t move (or shouldn’t move) because doing so would expose a more valuable piece behind it to attack, most dangerously when the piece behind is the king, which makes the move illegal outright.
- Skewer. The mirror image of a pin: the more valuable piece is in front, forced to move, exposing the less valuable piece behind it to capture.
- Discovered attack. Moving one piece out of the way reveals an attack from a piece behind it. When the revealed attack is check, it’s a discovered check, and when the piece that moved also delivers its own check, it’s a double check, which can’t be blocked or captured away, only escaped by moving the king.
- Removing the defender. Capturing or deflecting the piece that’s guarding a square or another piece, so the real target becomes undefended.
- Zwischenzug (German for “in-between move”). Instead of playing the expected recapture or response, you insert a forcing move first, usually a check or a bigger threat, and deal with the original situation afterward on better terms.
- Back-rank mate. The king is trapped on its own first or eighth rank behind a wall of its own pawns, with no escape square, and a rook or queen delivers mate along that rank.
Solve 30–50 fork problems in a row. Then 30–50 pin problems. Then skewers. Then discovered attacks. Then back-rank mates.
Themed sets lock patterns faster than random practice because you know what you’re looking for. Once you’ve done fifty fork problems, you start seeing knight fork distances naturally in your games, the specific squares a knight on a given square attacks become automatic rather than something you calculate from scratch each time. That’s the goal. Random puzzles at the start just build confusion, because you’re simultaneously trying to figure out what kind of tactic might be there and calculate the actual moves.
After each theme feels automatic in isolation, you’re solving themed puzzles within 30–60 seconds reliably, switch to mixed practice.

How to actually learn from puzzles you miss
This is where most players waste training time: they look at the answer, go “oh, I see it,” and move on. That’s not learning: that’s pattern exposure without retention.
When you miss a puzzle:
-
Find the move you played and why it was wrong. Which square was undefended? Which piece was overloaded? Which threat did you miss?
-
Trace back two moves. What made the combination possible? Was there an undefended piece that set it up? A king on the wrong square?
-
Name the pattern. “This was a knight fork because two pieces were on squares the knight could reach simultaneously.” Naming it forces active processing rather than passive recognition.
-
Set a timer on the next puzzle, don’t dwell. Spending five minutes on a missed puzzle helps once. Spending fifteen is diminishing returns.
Time limits
At first, take as long as you need on each puzzle. Once you’re solving most themed puzzles correctly, add a time limit: 3 minutes per puzzle maximum. If you haven’t found it in 3 minutes, look at the answer and apply the steps above.
For mixed puzzles after patterns are established, aim for 2 minutes per puzzle. Chess.com’s Puzzle Rush (5-minute timed sessions) is good for testing speed.
Across a full session that works out to somewhere between 30 and 60 puzzles for most players. The count matters less than whether you actually processed each one, per the steps above.
Puzzle rating vs. game rating
Your puzzle rating and your game rating are different things. Puzzle ratings reflect performance on isolated tactical positions; game ratings reflect complete chess. Most players find their puzzle rating runs 100–200 points higher than their game rating, because in games you also have to find the tactics yourself (without knowing one exists), plus handle opening, strategy, and time pressure.
Don’t be discouraged if your puzzle rating is higher than your game rating: that’s normal.
Building from puzzles to games
The transfer from puzzle practice to actual games requires one additional step: learning to recognize when a tactic is available. Puzzles tell you a combination exists; games don’t.
After each game you play (especially losses), go back through the position and ask at each critical moment: “Was there a tactic here that I missed?” This habit, combined with daily puzzle practice, is how the pattern recognition transfers.
See our chess improvement guide for the full study plan. And our chess tactics guide for the specific patterns to prioritize.
Sources
- Lichess Puzzle Trainer: free, open-source
- Lichess open puzzle database (live puzzle count)
- Chess.com Puzzle Trainer
- Chernev, Irving. Logical Chess: Move by Move. Batsford. (affiliate)
- Silman, Jeremy. Silman’s Complete Endgame Course. Siles Press, 2007. (affiliate)
This page contains affiliate links. As an Amazon Associate, DiscussChess earns from qualifying purchases.
Sources
Further reading
- Logical Chess: Move by Move, Irving Chernev — ASIN verified via Open Library 2026-05-02. Game annotations that explain the tactical ideas in context, pairs with puzzle practice to build pattern recognition.
- Silman's Complete Endgame Course, Jeremy Silman — ASIN verified via Open Library 2026-05-02. After tactics, endgame study is the second fastest improvement path for beginners.