An ice maze puzzle is a grid where the puck slides in a chosen direction until it hits a wall, a block, or the board edge, so every move commits to one complete straight line. The rule looks simple at first, but it removes every shortcut that ordinary tile-by-tile puzzles allow: there is no braking mid-slide, no double-tap to halt, and no stopping on an arbitrary open square. The only thing the player controls is the direction, and the only thing the board reports back is where the puck finally came to rest. That single decision, which wall, block, or boundary becomes the next stopping square, is what every level is built around. In Lizely's Ice Sliding Maze this rule runs across five original seven-by-seven layouts called Split Shelf, Cold Stair, Blue Switchback, Frost Relay, and Long Thaw. Their walls, starts, goals, cracked stops, and intended routes were authored specifically for the tool and verified by an independent breadth-first search over reachable stopping states. The result is a deterministic route-planning game where scoring, solvability, and shortest-path lengths are all proven rather than implied.
How Sliding Movement Works on an Ice Maze Puzzle
Movement is the single strict rule that defines every ice maze puzzle. Press an arrow key on the keyboard or tap a button on the on-screen direction pad and the puck keeps gliding in that direction until the next step would leave the seven-by-seven board or enter a blocked square. The puck cannot be released early, tapped twice to brake, or parked on any open tile you choose mid-slide. That makes each input a complete commitment: once the puck starts moving, the only thing you can influence afterward is the next direction you press.
Several visible elements communicate the rule without hiding essential state. Solid dark blocks are impassable and act as the walls the puck will rest against. The puck itself is labeled ICE on its current tile. The destination is outlined and labeled STOP so the goal is always visible. The most recent traversed cells receive a restrained accent so you can see the slide you just committed to, and cracked stopping cells carry a visible cross so you know where a slide would silently reset the rink. A boundary or adjacent wall that prevents any movement is treated as an atomic no-op: the move counter does not increase, the last path does not change, and no false failure is recorded. This is a discrete recreational puzzle, not a simulation of friction, momentum, acceleration, temperature, or real ice, and the product does not estimate travel time or physical distance.
The Five Rinks and Their Shortest Solutions
Every ice maze puzzle shipped in the tool is fixed rather than randomized, which makes every route reproducible. Each grid is a seven-by-seven board, and each cleared rink awards exactly 200 points. Finishing all five creates an exact total of 1,000. The shortest solution lengths climb steadily from rink to rink, and each literal route was checked by a separate breadth-first search over reachable stopping squares. Adding the shortest solutions together, 3 + 4 + 5 + 6 + 7, gives 25 total slides for a clean run.
| Rink | Grid Size | Shortest Solution | Points Earned |
|---|---|---|---|
| Split Shelf | 7×7 | 3 slides | 200 |
| Cold Stair | 7×7 | 4 slides | 200 |
| Blue Switchback | 7×7 | 5 slides | 200 |
| Frost Relay | 7×7 | 6 slides | 200 |
| Long Thaw | 7×7 | 7 slides | 200 |
| All Five | 7×7 | 25 slides total | 1,000 |
The minimum move count grows because each rink adds new walls, tighter corridors, or longer detours. The shortest path on each rink is independently proven by a BFS oracle that rebuilds the stop-state graph from the literal wall coordinates, then confirms that the displayed solution is genuinely shortest. The oracle implements its own slide loop rather than calling the production movement function, so agreement is meaningful rather than circular.
Plan From One Stopping Square to the Next
Because the puck cannot stop halfway, every ice maze puzzle is a planning problem over a small set of reachable stopping squares. On each rink there are only a handful of legal stopping positions: every wall face, every block face, every boundary edge that the puck can run into. Each reachable stopping square is a node in a graph, and each legal arrow input creates an edge to the final square reached in that direction. The shortest path to STOP is the shortest path through that graph.
Practical planning follows three habits. First, identify which squares can be reached from the puck's current position in a single slide in any direction, and treat that short list as the only set of next-stops. Second, check whether STOP is among those next-stops; if it is, finishing the rink is at most one slide away. Third, scan each candidate slide for cracked stops, which are marked with a visible cross and which silently reset the rink to its start without adding to the move count, but which still lock in a crack signature that may eventually deadlock the run. The goal marked STOP counts only when the puck actually rests on that square, so passing across the goal during a longer slide does not clear the rink.
How to Play Ice Sliding Maze to 1,000 Points
The interface is identical on desktop and mobile. Keyboard players use the four arrow keys anywhere on the page, while touch and pointer players receive a matching four-button direction pad with at least 44 pixels of height per control. Both input paths call the same reducer, so the stopping, scoring, trap, and terminal rules behave exactly the same regardless of device. If you have finished the five ice grids and want a similar deterministic route-planner with verified shortest paths, the Number Connecting Puzzle guide walks through another 1,000-point route-planning game.
- Open Ice Sliding Maze in a desktop or mobile browser. No download, account, or network gameplay service is required.
- Choose your input path: the four arrow keys for keyboard play, or the four on-screen direction buttons for touch and pointer play.
- Read the rink carefully. Locate the ICE puck, the outlined STOP goal, the impassable dark blocks, and any cracked cells marked with a visible cross.
- List every square the puck can reach from its current position by sliding up, down, left, or right. That short list is your candidate set of next-stops.
- Pick the slide that moves the puck toward STOP without landing on a cracked cell, and commit to the full straight line.
- Repeat slide by slide, treating each new stopping square as the start of the next planning step, until the puck comes to rest on STOP.
- Confirm the 200-point award that appears when the rink clears, then continue to the next rink.
- After the fifth rink clears, the run locks at exactly 1,000 points. Further inputs are ignored until Restart.
Trap Routes and Cracked Stops to Avoid
The opening rink, Split Shelf, illustrates why planning beyond the next slide matters. Its goal begins on the same bottom row as the puck, so the obvious move looks like a long slide directly to the right. That slide does not reach STOP. It stops on one cracked square, which resets the puck to the start. Sliding up first and then right looks like a quick detour, but that combination also stops on a different cracked square, which resets again. These two error routes are independently reachable and do not intersect the normal winning route, so a player who keeps guessing left or right will never stumble onto the correct path by accident. The shortest route actually travels up, left, and down even though the goal sits on the same bottom row as the puck.
Cracked stops use a level-plus-coordinate signature for dedupe. Repeating the same crack does not add a second mistake, but reaching a second distinct crack in the same run closes the rink until Restart. The point of the trap is not to punish guessing but to make sure every stopping square is a deliberate commitment. Treat each crack as a hidden boundary that removes one option from the next-slide list and forces a rethink rather than another casual arrow press.
Scoring, Telemetry, and Restart Behavior
Every successful slide that ends with the puck resting on STOP awards exactly 200 points. Five cleared rinks therefore produce exactly 1,000, with no rounding, bonuses, or hidden multipliers. A telemetry panel reports cleared rinks, total committed slides, the current row and column, level moves, and crack count, so the run can be audited without leaving the board. Once the fifth rink is complete or a second distinct crack has been reached, gameplay is frozen until Restart so extra events cannot alter the final state.
The shared GameShell supplies Restart, optional local best-score storage, and the double-Escape boss screen. Wall and boundary collisions with zero displacement preserve the exact state, which means tapping an arrow against a wall never burns a move and never counts as a mistake. No player input leaves the browser, no account is required, and no network gameplay service is called. The product's value comes from transparent movement rules, deliberately authored route geometry, exact score progression, and the auditable solver that checks each level independently.