Free tools Windows power users keep installed
One-click scans. No signup required.
PRISM’s developer says all 240 levels in the game’s six-chapter catalog were built by starting with a solved light-routing board, scrambling it with piece rotations, then checking that the resulting puzzle has exactly one solved arrangement. A separate breadth-first search checks whether the displayed par—the intended minimum number of taps—matches the shortest solution. Those are checks in the author’s described pipeline, not a universal guarantee about puzzle generators.
Why generate a puzzle backwards?
PRISM is a light-routing puzzle: tap a piece to rotate it 90 degrees, and the light immediately follows its new path. A board is solved when every crystal is lit at the same time with exactly the color it requests.
As an Amazon Associate I earn from qualifying purchases.
Rather than build an arbitrary board and hope it can be solved, developer 김종현 describes constructing a known solved board first. The solution is therefore present from the start. The harder questions—whether another solution also exists and whether the displayed par is truly minimal—are checked separately.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsHow the reverse generator creates a level
- Place emitters and routing pieces. Trace the light through the proposed board.
- Place crystals at the actual landing points. Each crystal is assigned the color of the light that reaches it, creating a solved state by construction.
- Scramble the pieces with rotations. Rotate them backwards by the chosen target number of taps. Reversing those rotations gives a known path back to the solved board.
This guarantees that the generator has an intended route to a solution, but not that the route is the only one. A second arrangement of piece orientations might light all the crystals correctly too.
#1 Best Overall
How the author checks for exactly one solution
The generator enumerates combinations of rotatable-piece orientations and counts how many arrangements solve the board. It keeps a candidate only when that count is exactly one; zero or multiple solved arrangements mean rejection.
The author reports abandoning the count and discarding the candidate when the state space exceeds 200,000. In other words, a candidate too large for the configured uniqueness check is not accepted on the assumption that it is probably unique.
This check also catches a subtle design problem: an unreachable rotatable piece can vary between orientations without affecting the light or crystals. Those variations create additional solved states, so the board fails the exactly-one test. Uniqueness checking therefore helps ensure that the pieces matter to the puzzle, not just that the crystals can be lit.
How the displayed par is checked
The generator chooses a target par and scrambles the solved board by that many taps. To check that the target is the true minimum, it runs breadth-first search from the scrambled board and compares the shortest path it finds with the intended par. If they differ, the candidate is rejected.
Rank #3
That search is separate from the reverse construction: the known scramble supplies a solution of the target length, while breadth-first search tests whether a shorter route exists. The author describes this as a way to catch errors in the generator as well as to verify the par.
Why passing the solution checks is not enough
A candidate can be solvable, unique and correctly parred yet still be a poor fit for its chapter. The author says the generator also rejects boards for quality reasons, including being too easy, already solved, having too few crystals, never bending the light, or failing to demonstrate the chapter’s intended effect, such as dispersion or color mixing.
Later chapters become harder to fill because they use more pieces and leave fewer useful empty cells; crystal placement adds another constraint. The developer reports that generating Chapter VI’s 45 levels took on the order of a million tries. The generator prints rejection counts for its different gates, helping the author tune chapter settings rather than treating every rejected candidate as the same kind of failure.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWhat is checked in the committed level data
A generator that runs correctly does not prove that the files ultimately shipped are correct. The author says checks run on every commit against the final level data across all 240 levels. They test whether each level is solvable, whether applying the solver’s path clears the board, and whether the level has exactly one solution.
Best Value
The author also describes a rendering-related failure mode. After a change made beam endpoints stop slightly short of absorbing pieces, the generator’s reachability check—which relied on whether an endpoint coordinate was an integer—began rejecting candidates containing walls. The author says the check was changed to use position and travel direction instead of endpoint coordinates. This account illustrates why validating the committed data matters in addition to trusting the generation process.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The reported 240-level catalog
In his 2026 account, 김종현 reports 240 levels across six chapters. The level counts and average pars below are the developer’s figures for PRISM, not independent measurements or general results for puzzle generation.
| Chapter | Levels | Reported average par |
|---|---|---|
| Reflection | 14 | 2.6 |
| Splitting | 32 | 3.7 |
| Dispersion | 44 | 5.1 |
| Mixing | 52 | 5.7 |
| Filtering | 53 | 7.2 |
| Convergence | 45 | 8.3 |
The reported averages rise chapter by chapter, while the later chapters also impose tighter generation constraints. The figures describe this game’s catalog; they do not show that the same pipeline will produce the same difficulty curve for other games.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →What the method establishes—and what it does not
As described by the developer, reverse scrambling gives each accepted candidate a known solution path, orientation enumeration checks that exactly one solved arrangement exists within the tested state space, and breadth-first search checks that the stated par is shortest. Commit-time checks then examine the final level data.
These claims are limited to the author’s reported implementation and checks. The account does not establish a universal method for proving arbitrary puzzles unique, nor does it independently verify the generator code or reproduce its tests. The author’s closing thought captures the game’s design philosophy: “A level you cannot solve is not a bug, it is logic you have not seen yet.”
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
