The two-mistake rule ends a mirror reflection puzzle run when a second distinct failed layout is recorded, so the fastest way to avoid common mistakes in mirror reflection puzzle routing is to memorize the four-direction turn mapping that the law of reflection produces on a 45-degree mirror. Each slash tile converts a rightward ray to upward, upward to rightward, leftward to downward, and downward to leftward, while each backslash tile supplies the complementary four turns. Empty cells do not redirect the ray at all, and the beam keeps traveling straight through them. The tracer stops on the target OUT edge, on a different boundary exit, or when the same cell-plus-incoming-direction would repeat, so a wrong-edge exit and a loop are reported as separate outcomes. Five fixed boards graduate from a single turn to a four-mirror route, so every new signal asks you to hold a slightly longer path in mind without adding random layouts or hidden rules. Players who treat slash and backslash as interchangeable, fill the board before tracing, or rebuild the entire layout after one wrong trace are the ones who burn both attempts before they ever read the highlighted path.
The Two-Mistake Rule That Ends Every Run
Every failed trace is checked against the current mirror layout, and only a second distinct wrong layout is what actually ends the game. The exact scoring rule on the Mirror Reflection Puzzle board reads like this: a wrong trace counts only when the placements on the board differ from any previously failed layout for that level, checking the same failed layout again does not consume another attempt, and a second different failed layout deadlocks the run. That distinction matters because it means you can leave a board in a clearly wrong state, press Enter many times, study the highlighted path, then keep adjusting without ever losing the run. The status line distinguishes between a ray that exits through the wrong edge and one that entered an enclosed loop, so you always know whether the mistake was a missing turn, a misplaced turn, or a closed ring of mirrors. Because the run only ends on the second different failed layout, most beginners actually lose two attempts for the same underlying reason without realizing that one of those attempts was free to repeat.
The Common Mistakes That Burn Both Attempts
Five recurring mistakes account for almost every collapsed run on the five-level puzzle. The first is treating slash and backslash mirrors as identical: a slash does not turn a ray the same way a backslash does, and swapping them across the entire board is the fastest way to waste an attempt. The second is forgetting that empty cells do not redirect the ray. A beam that enters an empty cell keeps moving in the same cardinal direction, so a missing mirror is not the same as a no-op tile. The third mistake is placing every mirror before tracing for the first time; the highlighted trace cannot tell you where the prediction diverged if you have already filled the board with speculative turns. The fourth mistake is rebuilding the entire layout after a single wrong trace instead of changing only the mirror that controls the divergence point, which often creates a brand-new second failed layout and triggers the end of the run. The fifth is treating wrong-edge exit and loop detection as the same failure; they are distinct outcomes, and the rule about empty cells, target exit, and repeated cell-plus-incoming-direction means each one points to a different fix.
Slash and Backslash Turn Mapping
The four-direction turns come straight from the law of reflection that OpenStax University Physics Volume 3 documents: the angle of reflection equals the angle of incidence when both are measured from the surface normal. On a 5-by-5 grid with 45-degree mirrors, that rule reduces to the eight cardinal mappings below.
| Incoming ray direction | Slash tile turns to | Backslash tile turns to |
|---|---|---|
| Right | Up | Down |
| Up | Right | Left |
| Left | Down | Up |
| Down | Left | Right |
Each row is a pair of opposite turns, and the slash and backslash columns are exact complements. Memorize the slash mapping, then derive the backslash mapping from it on the fly, and you will not have to re-check which tile belongs at each turn. The direction mapping is also what the tracer uses to draw outgoing arrows on visited cells, so the table is the same logic the engine uses internally to score a win.
Avoid Mistakes on the Five Fixed Boards
Follow this concrete sequence to keep both attempts alive until the final board reaches 1,000 points.
- Open the level and read the entry edge and the marked OUT edge before placing anything; the exit edge tells you the final ray direction and so fixes the orientation of the last mirror.
- Place only the mirror you can justify from the OUT edge and the entry path, leaving every other cell empty so the highlighted trace cannot hide behind speculative turns.
- Press Enter or choose Trace ray to run the beam cell by cell, and watch the outgoing direction painted on every highlighted cell for the first divergence between your prediction and the trace.
- If the ray exits through the wrong edge, identify the last mirror before the divergence and ask which direction the ray had to be traveling there; flip that mirror to the complementary tile and trace again.
- If the tracer reports a loop, find the smallest enclosed ring of mirrors on the board and remove or flip one of them so the ray can reach a fresh cell; the loop outcome means the same cell-plus-incoming-direction would have repeated.
- If the same wrong layout is produced again, that repeat does not consume a second attempt, so keep that layout on the board, read the divergence arrow carefully, then change exactly one mirror before tracing again.
- Once the OUT edge is reached, the level is recorded as solved, exactly 200 points are added, the next fixed board opens, and the run continues until the five boards are cleared for a perfect score of 1,000.
Read the Highlighted Trace After a Failed Attempt
After every trace, the board shows the outgoing direction on every visited cell, which turns a failed run into an inspectable artifact rather than a single "wrong" message. The first step in auditing a wrong trace is to find the highlighted cell whose outgoing arrow does not match the direction you predicted, then walk the ray one cell backward from there and ask which tile at that cell could have produced the recorded turn. If the outgoing arrow points through an empty cell toward the wrong edge, the missing mirror is the cell just before the divergence, not the empty cell. If the outgoing arrow points back into a previously visited cell, the failure is a closed ring of mirrors and the fix is to break that ring at the first cell where the predicted and recorded arrows disagree. Because the trace highlights every visited cell and records the outgoing direction, you can keep the same failing layout on the board as long as needed and change one mirror at a time without spending an attempt.
Solve Backward From the OUT Edge
A reliable way to avoid common mistakes in mirror reflection puzzle routing is to solve from the OUT edge toward the entry, because the exit fixes the final ray direction and so fixes the orientation of the last mirror. Ask which direction the ray must be traveling in the target exit cell, then place the mirror that converts that final direction into the direction the beam had when it arrived at that mirror. Continue walking one cell backward for each new mirror and use the slash and backslash table to pick the tile that produces the recorded turn. When the backward path meets the entry direction on a straight, mirror-free stretch, the route is complete and every cell outside the path can be left empty. This method is especially useful on the later boards because the four-mirror route is much easier to keep in mind when you only have to justify four turns, and the sparse board makes the highlighted trace trivial to read against your prediction. If you want a complementary read on placement logic in general, the mirror placement walkthrough covers the forward-solving side of the same boards.
Score, Restart, and Browser Behavior
Every solved level contributes exactly 200 points and all five levels together produce a perfect 1,000-point run; the score is exact, deterministic, and never hidden behind a streak multiplier or a timer. The five boards do not shuffle, so once you have solved a level the entry, exit, and required route are the same for every subsequent attempt, and Restart resets the run to level one, an empty placement map, zero mistakes, and zero points. Completed and failed states are frozen, which means extra clicks or key presses cannot alter a recorded result, and that is exactly why repeating the same failed layout is safe. Best score is stored only in local browser storage as an optional convenience, and the live run still works when that storage is unavailable. There are no timers, random seeds, network calls, third-party images, or paid features, so every mirror placement, every trace, and every score is reproducible in your own browser.
For a deeper look, see Hard Nonogram Puzzles: Why a 5 by 5 Can Still Beat You.