Air Hockey: How Block Density Determines Optimal Route Order

Air Hockey looks like a simple slide-until-you-stop puzzle, but the catch is that you must collect three keys in a fixed order before you can exit, even though all three appear at once. Choosing the wrong opening direction can cost ten seconds of correction; choosing the right one comes from reading block density before you move.

Your puck cannot stop in open space; it slides until it hits a wall or a block. So the only valid stopping cells are those next to a wall or block, which makes Air Hockey a pathfinding problem over those stopping points rather than over open squares. Open areas are essentially invisible to routing; the blocks are what matter. Where blocks cluster you have many possible stops and directions; where they are sparse you have few. The first thing to read on any new stage is the density distribution: is the rink evenly populated, or are there dense and sparse regions?

Once you have the density map, look at where each key sits. The key fact is whether each key is in a dense or sparse region. Keys in dense regions are reachable from many angles; keys in sparse regions can usually be reached from only one or two directions. A key in a sparse region is the binding constraint: the route must reach it from a specific approach, which dictates the rest of the path. Keys in dense regions are flexible and can usually be taken from whatever direction the constraint key allows.

Because keys are collected in order, the first key is special, you must reach it before anything else. Look at the blocks adjacent to key 1, identify which directions slide the puck to a stop on it, and check whether the puck can reach that stop from the start. If yes in one or two slides, key 1 is easy and the puzzle hinges on keys 2 and 3. If it takes three or more slides, key 1 is the constraint and you should plan the whole path from start through key 1 first.

An approach vector is the direction the puck travels when it hits a key. Each key has at most four, one per cardinal direction, but blocks usually limit how many are usable. Map the approach vectors for keys 2 and 3 during your scan: the route from key 1 to key 2 must end on a valid approach vector for key 2, and the route from key 2 to key 3 must end on one of key 3's. When approach vectors are tight, the intermediate stopping points are tight too.

The exit unlocks only after all three keys are collected, so before sliding, note where it is and which approach vectors reach it. Build your plan so that after key 3 you can slide straight to the exit without a long detour. When the three keys are roughly equidistant, the exit often sets the order: collect the key closest to the exit last and the farthest one first.

Every stage has a clean solution, and the time gap between the intended route and a brute-forced one is enormous: a player who finds the right route clears a dense stage in fifteen seconds, while random probing can burn a minute on the same board. So if you have spent more than fifteen seconds on a single key with no clear progress, reset and re-read from scratch. The reset reloads the same layout, the cost is fixed and small, and a fresh read with a clean mental model almost always beats pushing a flawed one. The puzzle is not getting harder as you struggle; your first read had a blind spot, and starting over is the fastest way to clear it. Our Air Hockey strategy guide covers the fundamentals this builds on.

Air Hockey plays three stages back to back, and they are not equally hard. The first is usually sparse, with blocks spread out and keys a few slides away. The second tightens. The third is typically dense, a tangle of possible stops with only a narrow band of correct ones. Because total time is the only scoring factor, the common mistake is over-planning the easy first stage and then rushing the hard third under time pressure. Flip that instinct: move briskly through the sparse stages and reserve your careful reading for the dense one, where a wrong slide costs the most.

The full sequence is short: read the density map, find the keys in sparse regions, map the approach vectors, plan the order from constraints to exit, then slide. The reading takes five to seven seconds and the sliding another fifteen to twenty. It is essentially logical reasoning under a timer, and you can drill it on the Air Hockey guide demo, which loads a fixed board so you build the habit without touching your competitive stats. Then bring it to today's board.

Wikipedia, Pathfinding.

Wikipedia, Constraint satisfaction problem.