Most adult improvers have one of two problems with attacking. Either they never do it — they shuffle pieces, wait for the opponent to crack, and quietly lose on the clock — or they throw everything at the enemy king on move 15, get nothing, and hand over a lost endgame two pawns down. Both problems come from the same root cause: they treat attacking as a mood rather than a set of conditions.
An attack on the king is not a personality trait. It is a decision you are allowed to make only when the position gives you permission. In fifteen years of coaching club players, the single biggest jump I see comes when someone stops asking “do I feel like attacking?” and starts asking “has this position earned an attack yet?” This guide gives you a concrete test for that question and a three-step build-up method for when the answer is yes.
Why adult improvers misjudge attacks
Club players below 1800 rarely lose because they cannot calculate a sacrifice. They lose because they commit to an attack that was never sound in the first place. The engine later shows a cold truth: the attack was refuted by a central counter-strike, or it simply ran out of attackers three moves in.
The fix is not more tactics puzzles — though those help you execute. The fix is a pre-flight checklist that runs before you push the h-pawn. Attacking a well-defended king is expensive: you spend tempi, you loosen your own structure, and you often invest material. You only want to pay that price when the position guarantees a return.
The Green Light Test: earn the attack before you launch it
Before you commit a single pawn or piece to a kingside assault, run these three checks. If all three are green, attack with confidence. If even one is red, you are gambling — and the house usually wins.
1. Is the center stable or closed?
This is the one adults skip, and it is the one that punishes them hardest. The oldest principle in attacking chess is that a flank attack is best met by a counter-strike in the center. If the center is open or fluid while you charge down the wing, your opponent simply blows it open in the middle, and your king — not theirs — ends up in the crossfire.
So before you attack on the side, make sure the center cannot be opened against you. That usually means a closed pawn chain or a stable, locked pawn tension. If the center is still in flux, your first job is not the attack; it is resolving the center in your favor. Understanding when to release central tension is its own skill — I cover the timing in detail in my guide to pawn breaks and open positions.
2. Do you have more attackers than they have defenders?
Count. Literally count the pieces you can point at the enemy king within two or three moves, and count the pieces defending it. Attacks succeed on local superiority of force, not on the global material count. You can be down a pawn and still have a winning attack if you have four pieces aimed at a king defended by one knight and a lonely g-pawn.
A practical rule of thumb: you generally want at least three pieces committed to the attack, and you want more attackers than there are defenders in the danger zone. If the opponent has a knight on f6, a queen on e7, and a rook that can swing to the kingside, you do not have enough — bring another piece over first.
3. Do you have an entry point?
An attack needs a door. That door is one of three things: an open or half-open file for a rook, an unobstructed diagonal for a bishop or queen, or a weak square right next to the king (classically h7, g7, or f7). If the enemy king sits behind three unbroken pawns with no open lines nearby, you have no entry point yet — and your third step will be to manufacture one.
The Build-Up Method: three moves that turn pressure into a breakthrough
Green light on all three? Good. Now you build. Amateurs lunge; strong players build. Here is the sequence I teach, in order.
Step 1 — Bring your slowest piece
The most common reason attacks stall is that one piece never joins. Look at your position and find the piece doing the least work — often a rook still on its home square or a bishop stuck behind its own pawns. That piece is your priority. An attack with three pieces beats an attack with two, and the difference between them is usually the lazy rook you forgot to activate. Reroute your worst piece toward the king before you commit to anything sharp.
Step 2 — Open a line
Attacks feed on open lines. If your entry point is already open, use it. If it is not, you open it — with a pawn lever or, when the math works, a sacrifice. A pawn storm (pushing your h- and g-pawns up the board) is the standard tool when the kings are on opposite sides. A piece sacrifice on h7, g7, or f7 is the tool when you already have overwhelming force and only need to rip the pawn cover open. The key discipline: you open the line after Step 1, never before. Sacrificing into a king that still has enough defenders is how adults lose miniatures.
Step 3 — Hit the weakest square
Every castled position has a weakest square, and it is almost always the one your opponent cannot easily add a defender to. Direct every piece at that square. The famous h7 sacrifice works because after the pawn falls, the king is dragged into the open and the whole army converges on one point. Your job in Step 3 is convergence: stop trading, stop grabbing pawns elsewhere, and pile everything onto the target. Momentum matters more than material here — an extra pawn is worthless if it costs you the initiative. Calculating these breakthroughs cleanly is a trainable skill; if your calculation wobbles under pressure, work through my tactics and calculation system for adult improvers.
Same-side versus opposite-side castling
The Green Light Test applies to both, but the flavor changes. When both kings castle on the same side, you cannot throw your own pawns forward without exposing your own king — so attacks are carried by pieces and by exploiting existing weaknesses (a hole on f5, a bad bishop, a weakened h-file). Patience and maneuvering win here.
When kings castle on opposite sides, the game becomes a footrace: you storm their king with pawns while they storm yours. Speed is everything, and every tempo counts double. The rule flips from “don’t loosen your own king” to “whoever arrives first wins” — so you push pawns without fear and you never waste a move defending unless you absolutely must.
When NOT to attack
Half of attacking skill is refusing to attack when the light is red. If the center is loose, if you have only two attackers, or if there is no entry point, the correct move is to keep improving your position — or to shut your opponent’s own plan down first. Learning to sense and prevent the other side’s attack before it starts is the natural companion to this skill; I devote a whole guide to it in prophylaxis: how to stop your opponent’s plan. A player who attacks only on green lights and defends calmly on red ones will out-score a reckless attacker every single time.
A 20-minute drill to build attacking vision
Reading about attacks does not build the instinct; reps do. Here is a drill I give students. Open a database or your own games and find ten positions where one side has a safe king and the other side has a lead in development or space. For each one, spend two minutes answering only three questions out loud: Is the center stable? Who has more force near the king? Where is the entry point? Do not calculate variations yet — just score the Green Light Test. After ten positions you will start seeing the preconditions automatically over the board, which is the entire point. Only once the test is second nature do you add calculation on top.
Put it into your next game
Attacking is not about courage; it is about permission. Run the Green Light Test, build with your slowest piece first, open the line, and converge on the weakest square. Do that consistently and your attacks will stop fizzling and start finishing.
Want to know whether your natural style even leans toward attacking play in the first place? Take our free chess archetype report — it identifies your strengths as a player and tells you which skills will move your rating fastest. And if you want a structured, week-by-week plan built around your archetype, the MyChessPlan premium plan ($14.99) turns these ideas into a training schedule you can actually follow.
Frequently asked questions
How many pieces do I need to attack a castled king?
As a rule of thumb, commit at least three pieces to the attack, and make sure you have more attackers in the danger zone than the opponent has defenders. Attacks succeed on local superiority of force, so count the pieces near the king rather than the total material on the board.
Should I attack on the wing or in the center?
Attack on the wing only when the center is stable or closed. If the center is open or fluid, a flank attack invites a central counter-strike that hits your own king. Resolve the center first, then attack the side.
Is sacrificing on h7 always sound?
No. The classic h7 sacrifice only works when you already have overwhelming force converging on the king and can calculate the follow-up to a clear win or perpetual. Never sacrifice to open a line while the defender still has enough pieces to hold — open the line only after you have brought all your attackers over.
How do I get better at attacking specifically?
Train the preconditions before the tactics. Score positions with the Green Light Test (center, attacker count, entry point) until it becomes automatic, then layer calculation practice on top so you can execute the breakthroughs you spot.




