The Three Phases of Every Dots and Boxes Game

Every dots and boxes game — from a casual 4×4 on a napkin to a fierce 5×6 tournament match — passes through the same three phases. The opening, the middlegame, and the endgame. Each phase has its own strategic character, its own decisions to make, and its own ways to lose.

Most intermediate players lose games because they misidentify the phase they are in and apply strategy that belongs to a different phase. This post is about recognizing phases, knowing what to do in each, and understanding how the transitions work.

Phase 1: The opening

The opening is the "safe phase." Every line you or your opponent draws does not complete any box and does not create any third sides. The grid is mostly empty, and there are plenty of moves that do not commit to anything.

What matters in the opening

Very little, honestly. In standard dots and boxes, opening moves are almost interchangeable. The first handful of lines can be placed anywhere — edge, interior, top, bottom — without much changing the eventual game.

What little does matter:

• Edge lines are slightly preferred because they preserve flexibility.
• Interior lines commit to specific chain structures earlier, which may be disadvantageous if you do not yet know what structure you want.
• Avoid creating early third sides (though this is easy when the board is mostly empty).

When does the opening end?

The opening ends when the first "unavoidable" decision arises. This usually happens around move 15-25 on a 5×5 or 5×6 grid, when safe moves start running out and someone is about to be forced to draw a third side.

You can tell the opening is ending when:

• You start having to scan carefully to find safe moves.
• You notice developing regions — clusters of drawn lines that will become chains.
• Both players are slowing down a bit, thinking more per move.

Up until this point, the game has been essentially automatic. After this point, decisions matter.

Common opening mistakes

• Spending too much time on opening moves. They do not matter much. Play them quickly.
• Trying to "set up" for specific chain structures early. The game is too open in the opening for these setups to be reliable.
• Mirroring the opponent. Mirror strategies fail (see our post on mirror strategies), and the opening is the easiest place to fall into the mirror habit.

Phase 2: The midgame

The midgame is where the structure of the endgame gets decided. This is the phase where skilled players separate from unskilled players, because good midgame decisions set up winning endgames.

What matters in the midgame

Almost everything. Specifically:

• Chain count parity. How many long chains and loops will exist at the end? Is that parity favorable to you?
• Region structure. Which regions will become long chains, which short chains, which loops?
• Who will be forced to open? Counting remaining safe moves tells you who runs out first.
• Forcing moves. Can you force the opponent into specific moves that help your structure?

The midgame's central tension

The defining tension of the midgame is between:

• Playing safe moves (filling in neutral lines to burn time) vs.
• Making structural moves (lines that change the shape of developing regions).

Too many safe moves and you run out of them, forcing you to be the one to open. Too many structural moves too early and you commit to structures that the opponent can exploit.

Strong players balance these. They play safe when structure is set; they play structural when they can influence a region they care about.

The chain rule in action

The chain rule governs the midgame. You want to steer the eventual chain count to match your parity preference. This often means:

• Merging regions to reduce chain count by one.
• Splitting regions to increase chain count by one.
• Preventing loops that would change the parity.
• Forcing loops that would change the parity.

Every midgame move is, at some level, a vote for or against specific parity outcomes.

When does the midgame end?

The midgame ends when no more "neutral" moves exist — every remaining move either completes a box, opens a box, or is forced by the structure. Typically around moves 35-45 in a 5×5 game.

You can tell the midgame is ending when:

• Every move you consider creates a third side somewhere.
• You cannot find any move that leaves the structure unchanged.
• The game is about to enter its capture-heavy phase.

Common midgame mistakes

• Ignoring parity. Playing "safe moves" randomly without thinking about chain counts.
• Playing too safely. Burning moves without influencing structure, and then finding yourself forced to open.
• Playing too aggressively. Committing to structural moves too early, before the opponent has committed.
• Missing the transition. Continuing to play "opening-style" moves when the midgame has already begun.

Phase 3: The endgame

The endgame is where the arithmetic becomes concrete. Boxes get captured, chains get taken, and the winner is decided.

What matters in the endgame

• Opening order. Which region gets opened first? Second? Last?
• Double-cross execution. Are you taking all-but-two from long chains, or taking full chains?
• Counting scores. Who is ahead, by how much, and how much more is available?
• Identifying the last chain. The last chain is different — it is taken in full, not double-crossed.

The double-cross is central

As we have covered in detail elsewhere on the blog, the double-cross is the technique of taking all-but-two boxes from a long chain and sacrificing the final two. This converts a tactical win (receiving a chain) into a strategic win (forcing the opponent to open the next chain).

In the endgame, double-cross application is the single most important skill. Players who master it win most endgames against similarly-skilled opponents who do not.

The last chain

The final region on the board is special. Since there is nothing for the opponent to be forced into after it is resolved, the double-cross serves no purpose. The player receiving the last chain should take every box.

Miss this, and you give away 2 boxes for no reason. Correctly identifying the last chain — and switching from double-cross mode to take-everything mode — is a skill that comes with practice.

When does the endgame end?

The endgame ends when the board is full. Every possible line has been drawn. Count up the boxes and declare a winner.

Common endgame mistakes

• Double-crossing the last chain. As above. Gives away 2 boxes for nothing.
• Taking the last-but-one chain in full. Forces you to open the last chain for the opponent, losing the game.
• Miscounting chain sizes. Leading to wrong decisions about which region to open first.
• Emotional capture-taking. Grabbing 3 free boxes without thinking about the position afterward.

Transitions between phases

The transitions between phases are critical moments. They are where the character of the game shifts, and applying the wrong phase's strategy is how many games get lost.

The opening-to-midgame transition

Signaled by the first region that obviously threatens to become a chain. When you see developing structure, it is time to start thinking about chain counts, parity, and eventual opening order.

The most common mistake: continuing to play opening-style (any line, quickly) when the game has entered the midgame. You give up structural advantages that a more aware player would exploit.

The midgame-to-endgame transition

Signaled by the first forced capture — when someone is first forced to draw a third side and the opponent takes the box. Once that happens, the game is in endgame territory, and every move might be part of a chain-taking sequence.

The most common mistake: continuing to play "positional" moves when the endgame has started. You miss opportunities to double-cross or force specific opening orders.

Recognizing phases in Dot Clash

The same three-phase structure exists in Dot Clash, though the mechanics differ:

• Opening: early dot placements, mostly in corners and sides. Nobody has been captured yet.
• Midgame: regions are forming, potential enclosures are developing, players are scrambling for territory.
• Endgame: actual captures happen. Enclosures resolve. The score target is approached.

The strategic transitions are similar. Opening decisions barely matter; midgame decisions shape structure; endgame decisions decide winners.

Per-phase practice suggestions

To get better at each phase:

For openings: play many games and get the automatic-move feeling. Opening skill comes from playing, not studying. After 30 games, your opening play is fine.

For midgames: practice parity counting. In your next 20 games, count long chains + loops at move 15, 25, and 35. Compare to your target parity. Track whether your play shifted parity toward or away from the target.

For endgames: practice the double-cross. Apply it religiously to every long chain except the last. Use the first 10 games to simply execute correctly; after that, start thinking about which chain to force your opponent to open first.

The meta-insight

Every strategy game has phases. Every phase has its own rules. Players who recognize which phase they are in, and apply phase-appropriate thinking, play substantially better than players who use one strategy throughout.

This is true in dots and boxes, chess, Go, Dot Clash, and most games with enough depth to have phase structure. Learning to recognize phase transitions is one of the most universal strategy-game skills, and it transfers across games.

The summary

Three phases. Know which one you are in. Apply the right strategy for the phase. Most critically, recognize transitions and adjust.

Opening: play fast, play safe, do not think too hard. Midgame: count parity, shape structure, manage tempo. Endgame: apply double-cross religiously, identify the last chain, count carefully.

Get that phase-awareness into your habits, and your play jumps a level without learning a single new technique. It is the same skill you already have, applied in the right phase at the right time.