♞CHESS:
AN AI LESSON

MIDDLEGAME PLANNING · LESSON 1 OF 10

The Middlegame
Planning Checklist

A good plan begins with an honest reading of the position. Use this order so an attractive idea never distracts you from an urgent threat.

← Course contents

1. Deal with what is urgent

Before inventing a plan, ask three questions: Am I in check? What is my opponent threatening? Do either of us have a forcing move? Look at checks, captures and direct threats for both sides. This tactical scan protects every strategic decision that follows.

Do not treat every check as good. A check is only a candidate. Calculate the reply and the position that remains. Also look for loose pieces, overloaded defenders and back-rank problems. If your queen is attacked or mate is threatened, “improve my knight” can wait.

Next check king safety. A king with weak pawn cover, open lines or too few defenders may change the entire plan. Sometimes your best middlegame plan is simply to castle, create luft or exchange the opponent’s strongest attacker.

2. Compare the position

Once the position is tactically safe, compare five things: material, king safety, pawn structure, space and piece activity. You are not trying to give the position a perfect computer score. You are trying to discover what matters.

Material tells you whether simplifying may help. King safety tells you which attacks are realistic. Pawn structure reveals weak pawns, protected squares and likely pawn breaks. Space tells you which side can manoeuvre more easily. Piece activity tells you whose army is ready now.

Turn the comparison into one sentence: “Material is equal, both kings are safe, but I have more queenside space and Black has a weak backward c-pawn.” That sentence is much more useful than “I think I am better.”

3. Choose one useful target

A plan needs a target or a transformation. A target can be a weak pawn, exposed king, weak square, bad piece or open file. A transformation can be a pawn break, favourable exchange or transition into a better ending.

Choose something you can actually reach. A weakness on a7 is irrelevant if every one of your pieces is on the kingside. Ask how many moves are needed, what your opponent will do during that time, and whether attacking it improves your pieces even if the target survives.

Plans often work in stages: improve a rook, double on a file, force a defender into passivity, then attack the pawn. The first move may not threaten anything. It should make the next useful move easier.

4. Create two or three candidate moves

Build candidates from the position rather than from hope. Include a forcing move if one is sound, a move that improves your least useful piece, and a move that prepares the important pawn break or stops the opponent’s plan.

For every candidate, calculate the opponent’s strongest reply—not the reply you want. Then compare the resulting positions. A simple sentence helps: “I want to play ___ because ___. They may answer ___. Then I can ___.”

Finally, perform a blunder check before moving. What changed when your chosen piece moved? Did it uncover your king, abandon a defender or allow a tactical reply?

Worked example: strategy only after safety

Interactive board: find the right order

White wants active rooks, but must first notice the immediate tactical fact.

The point is not the exact move in an artificial diagram. The point is the order of thought: threat first, position second, plan third, blunder check last.

Your over-the-board checklist

  1. Safety: checks, captures, threats and king danger.
  2. Position: material, kings, pawns, space and activity.
  3. Target: what can I attack or improve?
  4. Worst piece: who is not helping?
  5. Candidates: compare two or three sensible moves.
  6. Calculation: give the opponent their best reply.
  7. Blunder check: what becomes loose after I move?

At first this feels slow. With practice, it becomes a short scan. Use it most carefully when the position is quiet, because quiet positions tempt us to move without a reason.