How to train tactics with chess puzzles

By Andrian Yarotskyi · 21 October 2026 · 5 min read

Puzzles are the most popular way to train chess, and for good reason: tactics decide most games at club level, and puzzles train exactly the pattern recognition that tactics need. But how you solve them matters as much as how many you solve. Rushing through hundreds of easy puzzles feels productive and changes little. This guide explains how hard your puzzles should be, how to solve them, and how to make the practice carry over into your games.

Why tactics matter so much

Below master level, most games are not decided by deep strategy. They are decided by tactics: a fork that wins a piece, a pin that cannot be broken, a back-rank mate. The player who sees them more often, and allows them less often, wins.

Tactics depend heavily on recognizing patterns. A strong player does not calculate every possibility; they see a familiar shape, such as a king and queen on the same diagonal or an overloaded defender, and know what to look for. Puzzles build up that library of shapes faster than games can, because every puzzle contains one.

How hard should puzzles be?

The right difficulty is one where you solve most, but not all, of them. Roughly two thirds to three quarters is a good target.

Too easy and you are only confirming what you already know. It feels good, but you are not adding new patterns.
Too hard and you start guessing. A guessed solution teaches nothing, and failing again and again is discouraging.

Rated puzzle modes take care of this automatically: each puzzle is chosen around your puzzle rating, which goes up when you solve and down when you miss. Your puzzle rating is not the same as your game rating, and it is often higher, because puzzles tell you a tactic is there. That is normal.

Solve, don't guess

The way you solve a puzzle decides how much it teaches you.

Calculate the whole line before you move. Do not play the first move and wait to see what happens. Work out the replies and your answers until the position is clearly won.
Check the opponent's best defense. Most failed puzzles fail on a defensive resource: an in-between move, a check, a capture. Look for those before you commit.
Treat it like a game. In a real game, nobody tells you a tactic exists. Look at the position the way you would over the board: what is loose, what is overloaded, where is the king?
Avoid hints when you can. Using a hint is sometimes the right call when you are stuck, but count it as a miss.

Slow and fast practice

Both kinds have their place:

Slow solving (no clock, full calculation) trains calculation and accuracy. This should be most of your practice.
Timed solving (as many as you can in a few minutes) trains speed of recognition, which matters in faster games. It is a good warm-up and a fun change, but it encourages guessing if it is all you do.

A balanced week might be mostly slow, rated puzzles, with a timed session or two.

Focus on your weak themes

Puzzles are usually tagged by theme: fork, pin, skewer, discovered attack, deflection, back-rank mate, mate in two, and so on. Two ways to use themes:

Train what your games show you miss. If your own reviews keep finding missed forks, filter for forks for a week.
Train what you fail in puzzles. If you keep missing deflections, that pattern is not in your library yet.

Your own game reviews are the best guide here. See how to learn from your own chess mistakes for how to sort your mistakes into themes.

Review the ones you miss

A missed puzzle is the most valuable puzzle. When you get one wrong:

Look at the solution and play it through to the end.
Find the defensive resource or the attacking idea you missed.
Ask what in the position should have told you to look for it: a loose piece, an exposed king, two pieces on one line.
Save it and try again in a few days.

Repeating a fixed set of puzzles until you solve them all quickly is the idea behind The Woodpecker Method by Axel Smith and Hans Tikkanen, a training book many players swear by. The principle applies to any set: the patterns you see again are the ones you keep.

How many puzzles, and how often?

Consistency beats volume. Fifteen to thirty minutes a day of focused, slow solving does more than a weekly two-hour binge. Stop when you notice yourself guessing or getting tired: tired solving builds bad habits.

A puzzle of the day is a good minimum on busy days: one position, solved properly, keeps the habit alive.

Carrying tactics into your games

Puzzles only help if you look for tactics in your games too. Two habits make the connection:

Ask "is there a tactic?" whenever the position has the warning signs: a loose piece, an exposed king, pieces on the same line, a piece that is defending two things.
Use the same checks for both sides. Before every move, look at checks, captures and threats for you and for your opponent. That is also the core of the blunder check.

How Neochess helps

Neochess has several ways to practice tactics, drawn from the Lichess puzzle database of several million positions:

Rated puzzles, chosen around your puzzle rating. A new player's rating starts at the rating of their latest game, so the first puzzles are already about right, and it adjusts as you solve.
A puzzle of the day, the same for everyone.
Puzzle Sprint, solving as many as you can against the clock.
Theme filters, to train the patterns your games show you miss.
Favorites, to save puzzles you want to try again.

Alongside puzzles, the mistakes from your own analyzed games become exercises of their own, which is the most direct tactics training there is: the patterns that actually cost you games.

The short version

Pick puzzles you solve about two thirds to three quarters of the time.
Calculate the whole line, including the best defense, before moving.
Solve mostly slowly; add timed practice for speed.
Train the themes your games show you miss.
Review every miss and try it again later.
Fifteen to thirty minutes a day beats a weekly marathon.

Solve puzzles at your level with Neochess