V0.129 · DAILY PERFECT MAZE

Daily Maze

One date-stable perfect maze with a mathematically unique route, local autosave and a separate Maze streak that never changes Today’s 6.

The same maze every time you reopen the date

Daily Maze maps the local calendar date to one frozen source and one symmetry transform. Reloading the date reconstructs the same passages, start, goal and verified route. There is no account-dependent random draw and no remote maze API required to decide which board you receive.

The 2026 selector mixes Easy 11×11, Standard 17×17 and Challenge 25×25. It is deterministic but does not claim that all 365 dates are different. Release QA measures the actual distinct source-transform count and records the result.

Daily Maze is a specialist routine outside Today’s 6. Solving it updates Maze-specific Daily history and streak only; the frozen six-game lineup and its stored day records remain unchanged.

Use the shortest-path number as a transparent benchmark

Every Daily board shows its verified shortest route length. Because the maze is a tree, that route is also the only simple route from start to goal. Your actual step count can therefore be compared directly with the optimum.

A clean 100% route means no detour. If you explore a dead end and return, the extra movement lowers route efficiency in an understandable way. Backtracks and dead-end entries are shown separately so a slower time does not hide how you navigated.

This creates a useful daily target: first finish the maze, then try to reduce unnecessary branches on future dates. The system does not combine time and movement into an unexplained score.

Daily interaction works on phone and desktop

Arrow keys and WASD move one open passage at a time. On touch, you can tap a neighboring cell, swipe a direction, or drag through open corridors. The Challenge 25×25 board is designed to stay inside the mobile viewport without horizontal overflow.

Breadcrumbs display your current route and Fog mode reduces distant visual information if you want a memory-heavy variation. These are display controls; switching them does not replace the maze or create a separate Daily answer.

Pause stops the timer. Restart clears the current attempt while keeping the same date-stable board. Print uses the same maze data and removes interactive controls from the page.

Autosave protects an unfinished Daily

Every movement writes the current position, route, visited cells, steps, backtracks, dead-end history and elapsed time to browser-local state. Close the page after several moves and reopen the same Daily date to continue.

Completion removes the active autosave and writes a Daily history record with movement, time, route efficiency, assistance status and source identity. Consecutive solved dates build the Maze specialist streak locally.

The old V0.128 storage value is not overwritten during migration. V0.129 creates a new store and preserves prior specialist statistics while Maze begins with an empty record.

Verified Route Hint is available without hiding the assistance

A hard maze can be frustrating when one junction choice leads to a long branch. Verified Route Hint highlights the next neighboring cell on the unique path from your current position to the goal. It does not auto-walk the route.

Using the hint marks the attempt Assisted and increments the Maze hint count. Clean best records remain reserved for runs completed without that route information. This lets the Daily remain approachable without pretending a hinted solve was identical to an unaided one.

Because the underlying graph is a tree, the hint remains exact from any cell, even if you are currently deep inside a wrong branch.

Release proof covers the whole Daily selection

The source generator and release verifier are separate. The verifier checks every source for reciprocal passages, connectivity and the tree edge count, then independently rebuilds the start-to-goal path. It repeats the contract across every frozen symmetry transform.

All 365 dates in 2026 are sampled deterministically and validated. The same date is selected twice during QA to confirm replay stability, and the exact number of distinct source-transform starts is measured rather than guessed.

Actual Chromium QA then tests the product layer: movement controls, pause, hint assistance, autosave and reload, populated V0.128→V0.129 migration, Daily completion, history and streak.

Practice turns a difficult Daily into a learning loop

If a Daily board produces many wrong branches, switch to Practice at the same size. Seeded Practice produces replayable mazes, so you can retry one board and compare step count without waiting for tomorrow.

Focus on junction decisions. Before entering a branch, scan its immediate structure, remember whether a neighboring corridor has already been tested and use breadcrumbs to avoid repeating exploration. Over time, the dead-end and backtrack counters should become more informative than raw speed alone.

Then return to Daily Maze for one stable benchmark. That combination gives the specialist repeat-use value without changing the platform’s original Today’s 6 commitment.

Daily route-review loop

After a finish, compare your actual step count with the verified shortest route rather than treating completion alone as the goal. A low efficiency percentage usually means the run included exploratory branches or deliberate backtracking. On the next day, try to identify junctions before committing to a branch, notice which corridors terminate quickly, and use the breadcrumb trail to separate explored space from the route that still matters. Because every Daily board is date-stable, reopening the same date restores the same maze instead of silently swapping in a different layout.

What the counters mean

Steps counts legal moves. Backtracks counts moves that return along your current route stack. Dead-end entries records first visits to non-goal cells with only one passage. Junctions visited records cells with three or more passages. Visited cells measures coverage, while route efficiency compares the independently verified shortest start-to-goal path with the moves you actually used. These are local puzzle-performance measurements only; they are not IQ, attention, health, percentile, or diagnostic scores.

Use help deliberately

The Verified Route Hint highlights the next neighboring cell on the unique current-to-goal route and marks that run Assisted. That separation lets you learn without pretending a hinted solve was a clean route. Fog and Breadcrumbs change presentation but do not alter the maze graph. Pause stops the timer, and local autosave is designed to let a Daily run survive an ordinary reload in the same browser.