Plan in Sequences, Not Placements

A piece that is awkward alone is often fine as the middle of a three-piece plan.

Most players think one piece at a time: where does this go? Strong players think in short sequences: where do these three go, together?

Why it changes the answer

Consider an S piece arriving on a flat surface. Alone, it has no good placement: every option leaves a hole or a spike.

Now suppose the next piece is an L. Place the S so it creates a one-cell overhang, then tuck the L underneath. Together they leave a clean surface. Individually, the S looked like a disaster.

That is the whole idea. The unit of planning is not a piece; it is a small group of pieces that resolve into a clean result.

The queue makes it possible

Five pieces of preview is enough to see two or three moves of interaction. That is not incidental: the preview length exists precisely so sequence planning is available.

How to start

Take the current piece and the next one, and ask: is there a placement for this piece that makes the next one better? Not "where does this go" but "where does this go given what follows".

Then extend to three. Rarely more; beyond three the board has usually changed enough that the plan does not survive.

The common sequences worth recognising

S or Z, then L or J. The awkward piece creates a step or overhang; the workhorse piece resolves it.

T, then anything. If a spin notch is available, the T goes there and the next piece continues the surface.

O onto a step. An O placed across a one-cell step creates an overhang and a hole. Almost always, the O should wait or go flat elsewhere.

The failure mode

Over-planning. A four-piece plan that depends on everything going right will be invalidated by the first surprise, and you will spend attention rebuilding it. Two or three pieces is the sweet spot: long enough to solve the awkward-piece problem, short enough to survive contact with the board.

Read next

• The Nine-Column Stack and the Open WellFill nine columns flat, leave one open, and wait for the long piece. The standard shape for a high score.
• Six-Three Stacking and Why Some Players Prefer ItSplitting the board into two flat regions instead of one gives you somewhere to put the awkward pieces.
• Read the Queue, Not the Falling PieceIf you are deciding where the current piece goes while it falls, you are already too slow for level ten.
• Use Hold on Purpose, Not as an Undo ButtonThree good reasons to hold, and one very common bad one.
• The NES Version and Its Beautiful FlawsNo hold, no kicks, a randomiser that can starve you, and a scoring table that made it a competitive sport anyway.
• Wall Kicks: Why Your Piece Moved When You Rotated ItA blocked rotation tries a list of nudges before giving up. Knowing that list turns a quirk into a tool.