Domino Chain Reaction is a five-level, browser-based placement puzzle in which you toggle upright tiles on fixed 5-by-5 grids, push START, and carry an ordered cascade to the goal bell. The Domino Chain Reaction tool uses a strict tile budget instead of an unlimited construction sandbox. Each workbench contains one fixed START tile, one fixed goal bell, and visually identical open slots; neither endpoint consumes the placement budget. Place or remove dominoes before committing, but the board will not mark a correct route. Press Push the start and a deterministic breadth-first traversal follows connected tiles in the stable order up, right, down, and left. Every reached tile receives a number showing its place in the cascade and changes into a fallen state. The test ends when the goal is reached or no newly connected upright tile remains. Every solved workbench earns 200 points, all five create an exact 1,000-point finish, and play stays in your browser without an account, subscription, server request, or new download.
The route-planning task behind the cascade
This is an abstract grid-reaction puzzle rather than a simulation of physical domino timing, momentum, spacing, or friction. A falling tile can trigger only another tile directly above, right, below, or left of it. Touching at a corner has no effect, while an empty cell stops that branch. The central task is therefore route planning: decide which positions form one continuous, edge-connected chain before the push button tests your work.
The five boards retain the same 5-by-5 dimensions and rules. The first presents a short direct handoff. Later workbenches add downward travel, shelf-like turns, switchbacks, and a longer final cascade. Difficulty comes from spatial reasoning rather than changing controls. Before a push, the interface shows only your own placements: open slots share the same visible and accessible description, and no solution marker or future sequence appears in labels, hidden fields, or data attributes. You must infer the useful route from position, adjacency, and the available budget.
Plan from both fixed endpoints inward
Begin at both ends. First inspect which open slots share an edge with START. Then inspect which slots share an edge with the goal bell. Work inward from those two groups and estimate how many placements are required to bridge the remaining Manhattan gaps. Because the reaction advances one orthogonal grid edge at a time, two cells that look close but meet only at a corner cannot complete the connection.
Trace the proposed sequence cell by cell before pushing. A corner cell can redirect the chain through its own top, right, bottom, or left connections, but each change of direction still needs a shared edge. Do not fill every slot merely because pieces remain: each domino should either belong to the connected route or form a branch you understand and can afford. The stable up-right-down-left traversal order makes the push reproducible, while the limited inventory keeps each board a compact test of spatial reasoning.
Place, push, and complete each workbench
- Inspect the board. Locate the fixed START tile and goal bell, read the tile budget, and identify the available open slots. Their appearance is identical, so none is marked in advance as part of the solution. START and the goal do not count against the placement budget.
- Toggle legal placements. Tap or click an open slot to place an upright domino, and activate it again to remove that domino. Stay within the strict budget. Attempts to use closed spaces, fixed endpoints, out-of-range cells, or more tiles than allowed have no effect.
- Trace a continuous route. Work from START toward the bell and check every handoff. Consecutive dominoes must share a top, right, bottom, or left edge. A diagonal touch cannot propagate the reaction, and even one empty cell breaks the chain at that point.
- Review turns and branches. Make sure every corner changes direction through orthogonal neighbors. Remove unused tiles, and count any side branch against the same budget. Before pushing, retrace the intended sequence one cell at a time rather than judging only the overall shape.
- Push the start. Select Push the start, or press Enter while the board has keyboard focus. This is a real propagation test: the engine starts at START and performs deterministic breadth-first traversal using up, right, down, and left as its stable neighbor order.
- Read every observed result. Reached tiles fall and receive numbered fall-order labels. A successful test stops as soon as the goal is reached and awards 200 points, unlocking the next workbench. A failed test stops when no connected upright tile remains, so the numbered record shows exactly how far the cascade traveled.
- Correct a stopped layout. Use the final numbered positions to locate the break, then add or remove at least one open-slot tile before trying a genuinely different layout. Repeating the identical placement signature is deduplicated and does not consume another distinct failure. A second distinct layout that still stops short closes the run. If a corrected layout succeeds after one failure, its failure history is cleared before the next workbench.
Read a stopped cascade as a diagnostic
Begin at START and follow the numbers toward the intended route. When the numbered path ends, inspect the next planned cell: is it empty, connected only diagonally, or outside the current connected component? If the sequence travels into a side branch, separate those numbered branch tiles from the cells you still need for the goal. The Fall recorder repeats the observed sequence as compact position labels, but it never reveals a future route.
This makes every unsuccessful push useful evidence rather than a random verdict. Remove an unused branch, bridge a specific gap, or redirect the final connection, then trace the revised layout from both endpoints again. Because identical layouts are deduplicated, the next meaningful test must contain at least one actual placement change.
Score, layouts, and the two-failure boundary
Scoring is fixed: reaching the goal earns 200 points and unlocks the next workbench. A complete run is five successful workbenches × 200 points = 1,000 points. The fifth success freezes that exact result. Each workbench allows two distinct failed layouts, but the second genuinely different stopped layout ends the entire run.
| Game event | Exact result | Best response |
|---|---|---|
| Goal reached | 200 points are awarded and the next workbench unlocks. | Continue with Next workbench. |
| First distinct stopped layout | Its observed fall sequence is recorded as a failure. | Study the numbered endpoint and revise the placement. |
| Identical failed layout tested again | The placement signature is deduplicated. | Change at least one tile before making another distinct test. |
| Second distinct stopped layout | The run enters a terminal state. | Use Restart to begin again. |
| Success after one failed layout | The remaining failure history is cleared. | Proceed to the next workbench without carrying that failure forward. |
| Fifth successful workbench | The exact 1,000-point result is frozen. | Finish the run or restart from level one. |
Restart performs a complete reset to workbench one with zero score. It clears placed tiles, the fallen sequence, recorded layouts, any terminal state, and all prior failure history. It is therefore a clean new run rather than an undo of only the most recent placement.
Control the board on a phone, mouse, or keyboard
The complete board fits comfortably within a 390-pixel mobile viewport, and tile targets remain at least 44 pixels high. Touchscreen users can tap any open slot to place or remove a domino, while mouse users can click the same slots. For keyboard play, focus a grid cell and use the arrow keys to move the visible cursor. Press Space to toggle an available slot and Enter to push the start. When focus is on Restart, Push, or Next workbench, those controls retain their native button behavior instead of being intercepted by board shortcuts.
Closed spaces, fixed endpoints, out-of-range cells, budget overflow, solved levels, and terminal runs are atomic no-ops: they cannot partially change the score, placements, or observed sequence. The game runs entirely in your browser, and placement choices, pushes, fall sequences, scores, and mistakes are not uploaded for processing.
Run a final pre-push check
Before selecting Push the start, confirm that:
- The first placed domino shares an edge with the reachable path from START.
- The final placed domino shares an edge with a path into the goal.
- No intended transition relies on diagonal contact.
- No empty cell breaks the proposed route.
- Every turn and side branch has a purpose you understand.
- The layout remains within budget and differs from your last failed signature.
A successful run is built by placing fewer, better-connected tiles and using each numbered result to verify the next route. The push reveals the consequence of your layout; careful edge tracing is what makes that consequence predictable.